I have a new Samsung LN40A450, and when working with the Avia disc trying to set the white level, I don't see the moving white bars even if I set the contrast to zero.

Any ideas as to what's going on and how to resolve it?

Greetings

Look to the DVD player and if it has image settings that could be set wrong.

regards

Something is clipping. As Michael suggests, it is likely the source. What is the source?

If this is regular Avia, this is also a more severe problem, because BOTH of the bars in Avia are below reference white and should ALWAYS be seen.

Avia II adds an additional white bar above reference white. Generally, I would say that this also should remain visible, but some people for reasons of maximizing contrast on digital displays may choose to clip peak whites partially or totally.

Greetings

Maximize contrast at the cost of losing detail and adding discoloration into the whites too. :( That's what John Q Public likes to see.

Now tell me why my image looks soft again?

Regards

Greetings

Just make sure you didn't pause the video at white level. i did the same mistake at the beginning.

regards,
seelan

This is the original Avia disc (not Avia II) played on a DVD player through S-Video (I know, I know - it's temporary).

I do not have the test section paused - but the way the disc is programmed, it's not like the DVD player shows a running counter on those test screens anyway, so it's kind of hard to tell if it's playing properly or not.

If your DVD player has it's own contrast adjustment, you may need to alter the setting. One of my Denon DVD players was clipping white at the default player contrast setting.

I have a new Samsung LN40A450, and when working with the Avia disc trying to set the white level, I don't see the moving white bars even if I set the contrast to zero.

Any ideas as to what's going on and how to resolve it?

Most modern digital displays can't have the white level set with ANY test pattern because they show every shade of white even beyond 100% white which is represented by a digital 235 for all 3 colors (red, green, and blue).

LCDs don't clip no matter how bright the backlight or Contrast control are set. So you won't get a color shift (but the color could have errors all the time.

That said, it is POSSIBLE your backlight is flooding the LCD panel with so much like, it is washing out top-end detail. So you should start by turning the Backlight control down as far as possible. Then see if the white bars are visible.

I haven't seen one of these Samsung LCDs yet, but there is a creeping problem with newer and newer LCD panels... they are so freakin' bright and the control for backlight is so limited, that you can't set the white level low enough for comfortable viewing in a dark room... which is really the whole point of home theater.

Now, back to white level... just about all the "white level" adjustment patterns on test/setup discs are geared to CRT displays and the instructions for using the patterns don't apply to digital displays. Even though consumer digital video is limited to digital values between 16 and 235, digital video works on a 0-255 range and there are few digital displays that do anything to eliminate visible detail in whites that are encoded at 236-255. So almost all digital displays will display all or most whiter-than-white values. And that means 99% white, 98% white, 97% white, etc. are also displayed and are visible.

It can be difficult to see these whiter-than-white and near-white steps when the display is in torch mode (maximum backlight, high Contrast) because there's just so much light your iris doesn't contract enough to be able to distinguish the white detail that is on the screen.

And as far as "setting white level" with a test/setup disc on modern digital displays... there's very little you can do by eye if the display is revealing all the steps OK. This is making the Contrast control useless for anything except setting the peak white level which requires a meter. You display a 100% white window pattern (though on an LCD it doesn't matter if you use a 100% white field), set backlight to minimum (usually, though occasionally there might be a display where the minimum setting is too low, not very common though) and use the Contrast control to get 30 fL or thereabouts. This is a comfortable luminance (brightness) level for viewing the display in a dark room. If you need more brightness for a room with lights on or daylight, you can crank up the backlight which is usually all that's needed for bright-room viewing... but a contrast tweak may help on some brands/models also.

Lacking a meter, it's hard to know how close you are to 30 fL for the peak white level, but I find if you adjust Contrast so that a 100% white window pattern doesn't make you squint and so a 100% white field (full screen) is bright but doesn't feel overpowering, you are probably fairly close to where you need to be (these evaluations need to be done in a dark room).

I hope this info helps you get some good results. Let us know.

Most modern digital displays can't have the white level set with ANY test pattern because they show every shade of white even beyond 100% white which is represented by a digital 235 for all 3 colors (red, green, and blue).

LCDs don't clip no matter how bright the backlight or Contrast control are set. So you won't get a color shift (but the color could have errors all the time.

Huh? Sure they do.

Greetings

I can show you some Westie LCD flat panel sets in the THX class that clip the heck out of white when the contrast is set higher than 60%.

To get contrast set right, the patterns should have white all the way to 255 ish ... not just 235. There is additional image detail there that just enhances the image. The difference between a man in a white T shirt and a white shirt with all the subtle gradations.

Regards

Aha! I switched out the S-Video cable for Component, and that did the trick.

I'm not exactly sure why, but I can now view the bars, so that's a good thing. :)

Aha! I switched out the S-Video cable for Component, and that did the trick.

I'm not exactly sure why, but I can now view the bars, so that's a good thing. :)

S-Video? DUDE are you stuck in the 1980s???? :confused:

So now you're on component - welcome to the 1990s! LOL! ;)

I think it's time to join the new millenium! HDMI cables are under $6 each at www.monoprice.com (http://www.monoprice.com) --- of course if you don't have any source components with HDMI outputs you'll need to take care of that first :D

Greetings

I can show you some Westie LCD flat panel sets in the THX class that clip the heck out of white when the contrast is set higher than 60%.

To get contrast set right, the patterns should have white all the way to 255 ish ... not just 235. There is additional image detail there that just enhances the image. The difference between a man in a white T shirt and a white shirt with all the subtle gradations.

Regards

That effect may LOOK like clipping, but it's not clipping. LCDs don't emit light, they filter light to produce different colors. The backlight provides all the light. Once an LCD pixel is completely transparent, the pixels for the other 2 colors would have to get more transparent than 100% transparent so they pass more light than the "weak" color, but that doesn't happen to LCDs. The luminance histogram with red, green and blue graphed separately shows that LCDs don't clip.

If the manufacturer lets the Contrast control crush any color or colors, that's a crush problem, not a clipping problem. That means for your Westinghouse example, let's say that a 60 Contrast setting produces a pixels that are 100% transparent when you send a digital 255, that means if you send digital 254, each pixel is a little less transparent than it was when receiving a digital 255. Now, say you increase the Contrast control to 61... now both the 254 level and the 255 level produce the same level of transparency (100% transparent). Then move to a Contrast setting of 62 and now you have digital values of 255, 254, and 253 all being 100% transparent. That's crushing whites not clipping because there are no colors getting brighter (more transparent) at settings of 61, 62 or higher, you're actually making lower digital values reach peak white (100% transparent) before they should.

I forgot the Westinghouse panels did that - so you could very well get some use out of a white PLUGE for setting one of those (or others that do something similar).

I've read and heard experts claim digital values of 236-255 are in consumer video all the time, but other experts say it never happens these days. It is certainly no big trick to limit digital values to 16-235 anymore... there didn't used to be tools back in the 1980s (for digital audio), but these days it's a simple thing. But you have to analyze the entire movie (if that's what you are digitizing) first to find the darkest and lightest points and assign those to 16 and 235 so everything else you digitize falls in between. You can crush whites just fine without it being a loss of 236-255 data. Just make all pixels equally transparent for all values from 225-235 and you lose the detail in white shirts without higher bits ever being involved. Personally, until I see actual measurements of digital values present in consumer digital video sources, I have to say I don't know if 236-255 are ever present in DVDs or high-def discs on any kind of routine basis.

I learned my lesson, though, I should never use never when talking about video displays, there's always going to be something out there that does something unnecessary/unexpected.

Greetings

Clip crush ... that's all semantics. Why not crush clip ...? :) Both cases you lose detail. One side because you are supposed to ... the other side ... not really.

You have too much faith in TV makers, Doug. :) They still do plenty wrong ... often on purpose.

MAjor brand players like the Samsungs and the Panasonics and the Sonys and the Pioneers and the Sharps are now regularly shipping products (I think LG and Toshiba are here too) that have a hard time clipping/crushing or discoloring. That's a good thing.

The second tier players are not quite there yet ...

The knowledge trickles down slowly ... but given where we have come from in the last 10 years ... it's rather amazing.

I would not have thunk this in 1998 when I bought my first HDTV RP set.

Regards

Doug, Mr.D who works in video production says:

http://www.avsforum.com/avs-vb/showthread.php?p=13406530#post13406530

I've never found any commercially available material that stays under 235. (except maybe Avia but I regard it as useless for this very reason) Its a white reference remember not a limit.

Video looks way too clipped anyway even if you maintain all the variation up to 255. I don't think I could watch a display that clipped at 235.

http://www.avsforum.com/avs-vb/showthread.php?p=13776980#post13776980

It may interest you to know I worked on a couple of 1080p TV stings the other day ( specifically "Grand Designs" and "The F Word" ) both were pastiches of films ( Rocky and Gladiator respectively) because of the creative grading on the Gladiator skit I was given a final graded foreground ( match moved greenscreen comp) and Rec.709 scans direct from a spirit.

Both had significant image detail below and past 16-235. ( and the Gladiator skit was deliberately blown out to match the look of the film).

Incidentally the 3rd party calibration software I use on my workstation specifically assumes that any video material will have range outside 16-235 and applies a LUT to correspond to this ( ie: it gives me a rec.709 video enviroment: with variable gamma selection from 2.2 to 2.8... I work 2.2 for more obvious disclosure and finally display at 2.5 as 2.2 is invariable too flat looking for clients expecting to see something that approximates their own broadcast video chain).

After taking all of this into consideration, I recalibrated my display protecting my grayscale and gamma all the way out to 255 and it's seemed to make a subtle but noticable improvement (XBR2 SXRD). Of course I've sacrificed quite a bit of contrast ratio to do so.

Greetings

Either you want it set up right ... or you don't. :)

Also remember that the human eye actually has less CR than the TV in most cases. High contrast ratio numbers that give you a headache to watch ... do not result in good images.

If you can live with discolored whites and soft images on the bright end lacking in detail ... by all means pump the contrast back up there. The scary thing is ... where do you stop then ...? If killing 5%-10% of the top end detail is okay to you ... why stop there? Go for 20% ... and get even bigger CR numbers ... :D

The train destined to image accuracy would have long since left the station.

Regards

Do any of the popular calibration discs have "whiter than white" window patterns? I know that Avia II does have window patterns above 100 (IRE in the Avia world) - I think 102.5, 105, and 107.5. It seems like it would be useful to include those and for software to allow folks to optionally do a 12 or 23 (0 - 100, 105, 109 percent) grayscale measure in order to check for RGB tracking above 235.

cheers,


--tom

Greetings

The AVS HD DVD and Blu test discs have all the white you could ask for. :)

DVE on the HD side gives you one block of WTW. Not enough really to do it right.

The Monster ISF disc gives you the man in the White shirt pattern ... and that is a full white test pattern.

Other good white contrast patterns to look for are white gradation patterns. The ones that smoothly start from black on one end and end up at white on the other end. If the white is at 255 ... then turning up the contrast higher will cause ... (you just have to lookat it yourself. The effect is interesting.)

Stacey Spears created this parting RED SEA test pattern as I call it for his HD test disc. It is really nice and easy to use ... but unfortunately, it is not for sale. :(

Regards

The AVS HD disc has a 105% and a 109%(254) white window pattern that you can use to check your grayscale above 100%. I mainly just make sure that red hadn't been clipped at 109% since my SXRD runs out of red at higher contrast levels.

The AVS HD disc has a 105% and a 109%(254) white window pattern that you can use to check your grayscale above 100%. I mainly just make sure that red hadn't been clipped at 109% since my SXRD runs out of red at higher contrast levels.

Thanks Michael and sperron - unfortunately, no Blu-Ray player yet. So, I've only been working with GetGray and Avia/Avia II. Interestingly on my Samsung LCD, I've got contrast at 98/100 with the backlight at 1. This actually improved RGB tracking at the lower end where there clearly is a blue rise - the higher the contrast setting the more that is pushed a bit right to left thus improving things at 20%-30% without causing more color error at the high end (according to the i1D) or dropping gamma. Getgray includes above 235 on its ramp patterns as well as a max full field white which are what I've used to visually look for any visible color shifting above 235. With my eyes though, grayscale errors are more noticeable in the low end - above a certain luminance level, the error has to be fairly large for me to perceive it.

cheers,


--tom

It seems like it would be useful to include those and for software to allow folks to optionally do a 12 (0 - 110 percent) grayscale measure in order to check for RGB tracking above 235.

I agree, this would be a good question to address to the HCFR guys. 236-254 is a visible part of the picture and it should be accounted for.

Edit: And I did address it in the HCFR 2.0 thread (http://www.avsforum.com/avs-vb/showthread.php?t=983943&page=17).

DVE Professional has window patterns ranging from interface level 001 to 254.

Below reference black: 001, 005, 010, 012, 014.
Above reference white: 240, 250, 254

Of course, it's pretty expensive only to get these patterns. ;)

S-Video? DUDE are you stuck in the 1980s????

So now you're on component - welcome to the 1990s! LOL!

I think it's time to join the new millenium! HDMI cables are under $6 each at www.monoprice.com (http://www.monoprice.com) --- of course if you don't have any source components with HDMI outputs you'll need to take care of that first




I did say earlier that my being on S-Video was only temporary. This was my first HDTV, my previous setup didn't use component because it didn't need it - and certain components (such as my Tivo) couldn't handle it.

As for HDMI, I refuse to use it due to HDCP. I don't want anyone else telling me what I can and can't do with my video stream.

High contrast ratio numbers that give you a headache to watch ... do not result in good images.



Well... yes and no... it depends on how they get the high contrast ratio. If it's a display that doesn't have black levels as dark as Pioneer Kuro displays and the high contrast ratio is achieved by extremely high peak white levels... that's torture. But
But I know what you are saying... today's typical black levels (around .02fL) with a contrast ratio higher than 1500:1 or so are headache-inducing. One of today's typical displays with .02 fL black levels would have to output 600 fL to achieve the 30,000:1 contrast ratios being advertised - it's really silly. 600 fL! That's HUGE. It's much easier to get high contrast ratios by making blacks blacker - but it's harder for today's display technologies to get blacker.

I'm looking forward to infinity contrast ratios - as long as you have full control over the peak white level. If the display produces true black at 0 fL, the contrast ratio will be infinity for every/any peak white level... even if the peak white level is1 fL the contrast ratio would still be infinity when the black level is really 0 fL. Once everybody can achieve 0 fL blacks will finally be able to do away with contrast ratio all together. At that point, the only measure needed will be peak white level.

Greetings

Clip crush ... that's all semantics. Why not crush clip ...? :) Both cases you lose detail.

It's the engineer in me - using the right description/formula/data can mean the difference between success and disaster. A damp road and black ice can look the same, but if you treat both the same way, you're headed for big trouble!

That effect may LOOK like clipping, but it's not clipping. LCDs don't emit light, they filter light to produce different colors. The backlight provides all the light. Once an LCD pixel is completely transparent, the pixels for the other 2 colors would have to get more transparent than 100% transparent so they pass more light than the "weak" color, but that doesn't happen to LCDs. The luminance histogram with red, green and blue graphed separately shows that LCDs don't clip.

You make no sense with this. It absolutely is clipping. When you run you out of output and can reach that through the adjustment range, that's clipping. Those values are eliminated. It can either be a hard clip across all three colors, or a more subtle clip that affects only one or two at a time, causing first a colorshift, then the clipping/obliteration of detail.

Your distinction about LCDs being transmissive is not at all relevant to whether the device is clipping. The values are clipped off, that's just as straightfoward as can be. They're gone. They're not crushed or compressed, they're clipped.

If the manufacturer lets the Contrast control crush any color or colors, that's a crush problem, not a clipping problem.

No, crushing is where the delta between levels decreases and obscures their visibility. What we're talking about here is clipping, not crushing. These values are completely excoriated from the image. And it's exactly what it looks like. You crank white level through the roof and at a certain point the detail at the brightest portion of the luma range vanishes because it is being clipped off. And it isn't any kind of soft-clipping either. It looks terrible.

That means for your Westinghouse example, let's say that a 60 Contrast setting produces a pixels that are 100% transparent when you send a digital 255, that means if you send digital 254, each pixel is a little less transparent than it was when receiving a digital 255. Now, say you increase the Contrast control to 61... now both the 254 level and the 255 level produce the same level of transparency (100% transparent). Then move to a Contrast setting of 62 and now you have digital values of 255, 254, and 253 all being 100% transparent. That's crushing whites not clipping because there are no colors getting brighter (more transparent) at settings of 61, 62 or higher, you're actually making lower digital values reach peak white (100% transparent) before they should.

No, that's called clipping. All the values above a certain point are simply clipped off the video content altogether. They all get clipped back to the lowest value reproducible, in effect.

I forgot the Westinghouse panels did that - so you could very well get some use out of a white PLUGE for setting one of those (or others that do something similar).

I've read and heard experts claim digital values of 236-255 are in consumer video all the time, but other experts say it never happens these days. It is certainly no big trick to limit digital values to 16-235 anymore... there didn't used to be tools back in the 1980s (for digital audio), but these days it's a simple thing. But you have to analyze the entire movie (if that's what you are digitizing) first to find the darkest and lightest points and assign those to 16 and 235 so everything else you digitize falls in between. You can crush whites just fine without it being a loss of 236-255 data. Just make all pixels equally transparent for all values from 225-235 and you lose the detail in white shirts without higher bits ever being involved. Personally, until I see actual measurements of digital values present in consumer digital video sources, I have to say I don't know if 236-255 are ever present in DVDs or high-def discs on any kind of routine basis.

I learned my lesson, though, I should never use never when talking about video displays, there's always going to be something out there that does something unnecessary/unexpected.

Also remember that the human eye actually has less CR than the TV in most cases.

I'm not sure where this impression comes from, but this is most certainly not the case. Standard displays don't even come close to matching the capabilities of the human eye.

I know this is something that has been taught for some time by the ISF, and that's quite unfortunate, but it is not at all true. Darrin Perrigo has spent a great deal of time combating this misunderstanding and confusion, which it appears arises from a misapplication and misunderstanding of CSF limitations.

Do any of the popular calibration discs have "whiter than white" window patterns? I know that Avia II does have window patterns above 100 (IRE in the Avia world) - I think 102.5, 105, and 107.5. It seems like it would be useful to include those and for software to allow folks to optionally do a 12 or 23 (0 - 100, 105, 109 percent) grayscale measure in order to check for RGB tracking above 235.

cheers,


--tom

Yes. Avia II has many patterns that go both below black and above white. Among the most useful of these is the "2.5IRE" black window pattern, and the xxdeep ramps pattern (also extremely useful to observe banding and clipping on a digital). It also has a steps pattern from 0-255, and 1-254 which may be of interest. Many of the other patterns also have moving black and white bars which include a bar that is below black, and above white, respectively. And also there's small black and white crossing ramps/checkers there too which are helpful.

DVE has the crossed ramps pattern which goes above and below, the dots mark reference white and black.

Get grays ramps go all the way too.

You make no sense with this. It absolutely is clipping. When you run you out of output and can reach that through the adjustment range, that's clipping. Those values are eliminated. It can either be a hard clip across all three colors, or a more subtle clip that affects only one or two at a time, causing first a colorshift, then the clipping/obliteration of detail.

Your distinction about LCDs being transmissive is not at all relevant to whether the device is clipping. The values are clipped off, that's just as straightfoward as can be. They're gone. They're not crushed or compressed, they're clipped.

Your description of Clipping in the 3rd sentence is not Clipping, it's just the maximum luminance of the display in question. When you increase Contrast, at some point the display won't get any brighter - that's normal/typical. If it happens for all 3 colors at the same time, it is not clipping and may or may not cause crush depending on whether the display is still responding in a linear fashion. Clipping is an luminance limiting issue affecting one (or sometimes 2) colors before the other color(s) stopped increasing output level.

For example, you could start with a low Contrast control setting and all 3 colors would reach the same peak luminance level. Increase the Contrast control up to a certain point, and usually 1 color (occasionally 2 colors) would no longer be able to achieve the same luminance level. But this happens only in plasma and CRT displays, not LCD or DLP. It's like having a 3-channel audio amplifier with 2 channels having 100 watts, and 1 channel having 50 watts... the 50 watt channel will clip before the other channels.

In an LCD, 100% white is always the level of the backlight minus the transmission losses through the LCD panels. All the pixels can do is reach 100% transparent. If the pixels are 100% transparent when they SHOULD be 95% transparent, that's not clipping, that's just bad design and it will cause crush in the highlights. It causes a complete loss of detail without a color shift.

Crushed blacks happen when multiple digital values produce the same or very similar luminance level. For example, crushed blacks happen if digital values 20, 21, 22, 23, 24, 25 & 26 all produce the same dark gray shade instead of slightly different luminance levels. This is a simplified example, 27, 28, 29, 30, 31, 32 might all produce a luminance level 1% higher than the previous group of steps, etc. Crushed whites happen if 220-235 all show the same luminance level. Crushed whites happen without a color shift. Crushed blacks happen without a color shift.

Clipping is completely different and does not produce the same luminance level for all 3 colors in PDPs and CRTs as it's just about impossible for the red phosphors, green phosphors, and blue phosphors to all peter out at the same luminance level. Calibrators have to set a peak level that keeps all 3 colors within a linear operating range to avoid clipping.

Example - let's say a PDP has the red gamma curve flatlined at 95% so red never achieves the same luminance level as blue and green. ANd blue and green continue producing the light they need to produce to reach 100%. The 90% white step (assuming a reasonably well-calibrated display) will be neutral because all 3 colors are able to achieve the same luminance level. But at 100%, red luminace will trail green and blue so 100% white will have a cyan color shift. (this assumes the red problem is not fixable with the Gain control of course)

An LCD (like the Westinghouse in the example) that has the Contrast control set well beyond 60 causing all pixels in all colors to be 100% transparent for digital values from 220-255 to detail will be lost, but there will be no color shift because there is no clipping. For clipping to exist in an LCD, one panel would have to be less transparent than the other 2 panels so that it could not achieve the same luminace level for any given backlight level. Let's say the Red panel behaved just like the green and blue panels up to 95%, but then the red panel never got any more transparent, but the blue and green panels would continue to get more transparent all the way to 100% transparent. That would be clipping in an LCD, but that never happens in the real world. As has been pointed out by Michael, most LCDs DO NOT clip - if you run into it, more than likely it will be in lower-tier products.


No, crushing is where the delta between levels decreases and obscures their visibility.

You are making my own arguement with this statemet. This is what happens for highlights and shadows when there is crush.


What we're talking about here is clipping, not crushing. These values are completely excoriated from the image.

Not true. Clipping simply means 1 color (maybe 2 colors in rare cases) can't produce a luminance level equal to the other colors. You will still have the steps in 2 colors, but one color stops getting brighter at some point. You end up with no steps in that 1 color and a color shift because the other 2 colors continue to get brighter.


And it's exactly what it looks like. You crank white level through the roof and at a certain point the detail at the brightest portion of the luma range vanishes because it is being clipped off. And it isn't any kind of soft-clipping either. It looks terrible.

Only for 1 color at a time (2 colors in rare cases), the colors that continue to get brighter are NOT limited and still produce gradation steps as digital values increase.


No, that's called clipping. All the values above a certain point are simply clipped off the video content altogether. They all get clipped back to the lowest value reproducible, in effect.

Displays that clip won't all cut off at the same point (it would be a miracle if they did, at least) - one color will give up first, another color will give up next, and the "strongest" color will give up last. Clipping causes color shifts because it doesn't happen equally in every channel at the same time.

Crush is a condition where all the channels stay together, but they are "wrong" in that there's not enough luminance difference between digital levels and steps will blend together without a luminance shifts for each digital value.

Crush can happen in shadows or highlights.

Clipping can only happen at the bright end of the spectrum when the pixels produce the light and one color doesn't achieve the same peak luminance as the other colors (and can't be brought back in line with the Gain control).

Transmissive pixels can't clip - every 100% transparent pixel will pass the same amount of light in a transmissive display like LCD.

I can't help if these terms have been used interchangeably over the years - but they are not the same thing. It's like wheel and rim - people use them interchangeably all the time but they are NOT the same thing from an engineering/technical point of view.

one color will give up first, another color will give up next, and the "strongest" color will give up last. Clipping causes color shifts because it doesn't happen equally in every channel at the same time.

I would have to think that SXRD is a transmissive technology similar to LCD. When turning up contrast it is possible for one primary in the grayscale to stop increasing while the others continue to go up. The behavior causes a color-shift in the near whites similar to your multiple descriptions of clipping. As far as I can tell you seem to be stating that the behavior I described is not technically defined clipping due to the process by which different displays produce grays. From a practical standpoint for an end-user calibration the distinction seems rather trivial to me, because regardless of display technology the end-result can still be an unintended color-shift in the grayscale when turning up contrast as one primary can no longer match the others.

Your description of Clipping in the 3rd sentence is not Clipping, it's just the maximum luminance of the display in question. When you increase Contrast, at some point the display won't get any brighter - that's normal/typical. If it happens for all 3 colors at the same time, it is not clipping and may or may not cause crush depending on whether the display is still responding in a linear fashion. Clipping is an luminance limiting issue affecting one (or sometimes 2) colors before the other color(s) stopped increasing output level.

I've never heard or read of clipping being described in this way. Clipping is literally where the top of the waveform is clipped off. That can occur across all three, or less than that. The waveform maxes out, and no values above the clip point exist, because you've reached the maximum plateu. That is exactly what is happening, but in the digital domain, when you put your white level too high on many (or most) digital displays. You will reach a certain light output, and then there will be a hard stop, and no values above that will be presented at all at any different luminance. They all get clipped back to the same level.

Crushing, however, I use and agree with Greg Roger's use, which describes a decrease in the delta between codes, but not their complete elimination. This is particularly relevant with dynamic gamma circuitry or other gamma tweaks where the values are still present, but detail visibility may be reduced (but not eliminated) because the luminance delta between codes (that are still present) has decreased.

With crushing, values are still represented and recreated, with clipping they are clipped off (literally) altogether.

However, due to confusion of crushing, more akin to luminance compression, many people use crushing and clipping interchangeably, which is unfortunate. Clipping is unambiguous.


For example, you could start with a low Contrast control setting and all 3 colors would reach the same peak luminance level. Increase the Contrast control up to a certain point, and usually 1 color (occasionally 2 colors) would no longer be able to achieve the same luminance level. But this happens only in plasma and CRT displays, not LCD or DLP. It's like having a 3-channel audio amplifier with 2 channels having 100 watts, and 1 channel having 50 watts... the 50 watt channel will clip before the other channels.

That most certainly does occur with LCDs and DLPs. That's why on some displays you have to pay attention first to the colorshift, where one RGB channel clips before the others. You need to keep your white level just below that point.

In an LCD, 100% white is always the level of the backlight minus the transmission losses through the LCD panels. All the pixels can do is reach 100% transparent. If the pixels are 100% transparent when they SHOULD be 95% transparent, that's not clipping, that's just bad design and it will cause crush in the highlights. It causes a complete loss of detail without a color shift.

That is exactly what clipping is.

Crushed blacks happen when multiple digital values produce the same or very similar luminance level. For example, crushed blacks happen if digital values 20, 21, 22, 23, 24, 25 & 26 all produce the same dark gray shade instead of slightly different luminance levels. This is a simplified example, 27, 28, 29, 30, 31, 32 might all produce a luminance level 1% higher than the previous group of steps, etc. Crushed whites happen if 220-235 all show the same luminance level. Crushed whites happen without a color shift. Crushed blacks happen without a color shift.

Again, that's called clipping. I retain the "crush" terminology for situations where the delta between levels is decreased, but not eliminated. Clipping, they are are eliminated entirely, which is exactly what you are describing. All values get clipped back to a single value, and are reproduced at exactly the same luminance, leaving no detail whatsoever. Crushing, on the other hand, is the reduction in detail visibility between small changes in luminance due to tue decrease in that delta.

Clipping is completely different and does not produce the same luminance level for all 3 colors in PDPs and CRTs as it's just about impossible for the red phosphors, green phosphors, and blue phosphors to all peter out at the same luminance level. Calibrators have to set a peak level that keeps all 3 colors within a linear operating range to avoid clipping.

You are correct that CRTs do not inherently hard-clip at all. In fact, I wouldn't use the term clipping for what CRTs do at all. However, you can certainly clip the signal coming into the CRT. Clipping has to do with what is happening to the signal, not so much with the display, however in digital displays the two aspects are connected, which is why you do get a hard clip that you don't really get with an analog CRT (usually).

Example - let's say a PDP has the red gamma curve flatlined at 95% so red never achieves the same luminance level as blue and green. ANd blue and green continue producing the light they need to produce to reach 100%. The 90% white step (assuming a reasonably well-calibrated display) will be neutral because all 3 colors are able to achieve the same luminance level. But at 100%, red luminace will trail green and blue so 100% white will have a cyan color shift. (this assumes the red problem is not fixable with the Gain control of course)

Correct, and this colorshift is caused by clipping on the red channel.

An LCD (like the Westinghouse in the example) that has the Contrast control set well beyond 60 causing all pixels in all colors to be 100% transparent for digital values from 220-255 to detail will be lost, but there will be no color shift because there is no clipping.

That IS clipping. That is unambiguously driving a signal so high that the top is totally clipped off.

If that's not what clipping is, I don't know how you're defining clipping, but it's a very strange definition indeed.

For clipping to exist in an LCD, one panel would have to be less transparent than the other 2 panels so that it could not achieve the same luminace level for any given backlight level. Let's say the Red panel behaved just like the green and blue panels up to 95%, but then the red panel never got any more transparent, but the blue and green panels would continue to get more transparent all the way to 100% transparent. That would be clipping in an LCD, but that never happens in the real world. As has been pointed out by Michael, most LCDs DO NOT clip - if you run into it, more than likely it will be in lower-tier products.

It seems to me that you understand clipping to occur ONLY when one color runs out but not others, causing a colorshift. This is a mistaken understanding of clipping, and I'm not sure where this definition or use comes from. I've never encountered this definition before in any text or in professional use.


You are making my own arguement with this statemet. This is what happens for highlights and shadows when there is crush.

Again, I differentiate between a clip, where all the values above (or below) a certain point are wiped out back to one value. Clipping terminology comes from the analog domain when you clip the extremes of a waveform, and the same thing happens digitally. This eliminates totally the delta between levels. Crushing I would reserve more when that delta is still present, but in diminished degree. I think in fairness, I wouldn't necessarily object to using crushing to describe two adjacent levels being crushed together within the video range, but not at the extremes where clipping is very much the most accurate way to describe it.


Not true. Clipping simply means 1 color (maybe 2 colors in rare cases) can't produce a luminance level equal to the other colors. You will still have the steps in 2 colors, but one color stops getting brighter at some point. You end up with no steps in that 1 color and a color shift because the other 2 colors continue to get brighter.

You're the only person I've encountered who has described it that way. In other words, it's not clipping unless only SOME of the signal is being clipped? If all three are being clipped, then it's only crushed, but not clipped? Makes no sense.


Only for 1 color at a time (2 colors in rare cases), the colors that continue to get brighter are NOT limited and still produce gradation steps as digital values increase.

Displays that clip won't all cut off at the same point (it would be a miracle if they did, at least) - one color will give up first, another color will give up next, and the "strongest" color will give up last. Clipping causes color shifts because it doesn't happen equally in every channel at the same time.

Um, well some do. It depends on the design. You can also just clip the signal, and that will clip off at the same point for all RGB together at once. Or you can clip in the component domain, and you're clipping Y', so it's also the same thing.

Crush is a condition where all the channels stay together, but they are "wrong" in that there's not enough luminance difference between digital levels and steps will blend together without a luminance shifts for each digital value.

Your use of this terminology is very odd, and that's the problem I have with it. Nobody uses these terms in this manner, in any literature.

If you literally are clipping the peaks of a waveform off, then I don't see how you can regard that as something other than clipping.


Crush can happen in shadows or highlights.

Clipping can only happen at the bright end of the spectrum when the pixels produce the light and one color doesn't achieve the same peak luminance as the other colors (and can't be brought back in line with the Gain control).

Transmissive pixels can't clip - every 100% transparent pixel will pass the same amount of light in a transmissive display like LCD.

They most certainly can, and they most certainly do. That's exactly what is happening when they max out. All values above the maximum are simply clipped off and reproduced at that maximum. That is called clipping. That's what clipping is, by definition. Whether one, two, or three of the signals are clipping results in different visible artifacts. However, there isn't some threshold where if one or two signals are being clipped it's called "clipping" and if all three are being clipped then suddenly it's not clipping anymore, but something else. That is nonsensical.


I can't help if these terms have been used interchangeably over the years - but they are not the same thing. It's like wheel and rim - people use them interchangeably all the time but they are NOT the same thing from an engineering/technical point of view.

And that's what I'm telling you.

For example, I try to use the terminology as wielded by Darin and Greg here:

http://archive2.avsforum.com/avs-vb/showthread.php?p=6668611&&#post6668611

I should have said that I think you were right in the first post. Yes, I think these are very similar. However, it partially depends on what we mean by "crushing". One way is to be technical and say that there can't be any visible difference between levels (like clipping). I take a somewhat different approach and feel that blacks are crushed to some degree if the differences between their levels are smaller than they should be to the point that we can't see as much detail as the material called for. In the case I mentioned above they don't have to clip all the 50%, 60%, and 70% levels to 50%. They could display the 50% at 45%, the 60% at 48%, and the 70% at 50% for that one frame. The best things to do get complicated when considering running video at speed though. So, I think it is fair to say there is a certain amount of crushing going on with the brightness compression, but it helps to make sure people know it isn't exactly the same as clipping (which sometimes gets referred to as crushing).

--Darin

and
Darin has explained exactly why I didn't call it brightness crushing, or white crushing. There's too much confusion about whether "crushing" means clipping or compression. In this case, it's compression, not clipping, as Darin points out.

__________________
Greg Rogers
AccuPel
Widescreen Review

http://archive2.avsforum.com/avs-vb/showthread.php?p=6411889&&#post6411889

I used the term brightness compression to describe the process the Auto Iris must use to avoid brightness clipping. Compression is used in various video and audio processes to overcome limitations of bit rate, storage capacity, bandwidth, noise floor, etc. In this case it is being used to overcome the black level (contrast ratio) limitation of fixed-pixel projectors. So compression, or brightness compression in this case, is not inherently bad. But like most compression systems it's a tradeoff at some level because its effects aren't completely masked.

When bright signal levels are compressed closer together you must lose contrast within those signals. You can no longer differentiate between high brightness levels as well as you could have in a linear system. But if you gain much better contrast in dark scenes, it may be a good tradeoff to give up some contrast in the bright image areas. However, if in some scenes too many bright levels are compressed too closely together, the effect is nearly indistinguishable from clipping below the maximum signal level. So it's a matter of degree, and personal judgment, as to where beneficial brightness compression ends and objectional compression artifacts begin. That is true in almost all compression systems I can think of, including MPEG, Dolby Digital Surround Sound, etc. etc.

White crushing sounds like something bad, so I would assume "they" mean clipping or brightness compression that is so severe it beomes objectionable because you can no longer see sufficient detail in a bright area of an image. But its a judgement call where useful brightness compression ends, and objectionable brightness compression artifacts begin. I don't have any problem with using white crush to mean severe, objectionable brightness compression artifacts, but I'd rather use the term clipping if that occurs, because it is more descriptive of what would actually be happening.

__________________
Greg Rogers
AccuPel
Widescreen Review

Greetings

And we call things with high color temperatures "cool" ... :D

Regards

haha, yeah, and black level is brightness.

I would have to think that SXRD is a transmissive technology similar to LCD. When turning up contrast it is possible for one primary in the grayscale to stop increasing while the others continue to go up. The behavior causes a color-shift in the near whites similar to your multiple descriptions of clipping. As far as I can tell you seem to be stating that the behavior I described is not technically defined clipping due to the process by which different displays produce grays. From a practical standpoint for an end-user calibration the distinction seems rather trivial to me, because regardless of display technology the end-result can still be an unintended color-shift in the grayscale when turning up contrast as one primary can no longer match the others.

I own an SXRD RPTV... it is LCD mounted to a mirror (that's on a chip) so the same "rules" apply. It never color-shifts as you increase the contrast setting and it never crushes whites. It retains every step of detail at every contrast setting. You can run it right up to max contrast with the most wide-open iris setting and if the gains are adjusted properly, there's no color shift at the top end nor any loss of "steps" (crush). This is typical for properly designed LCD, LCoS, and DLP displays because these all have external light sources. The panel or panels are merely valves for the light if you want to think of them that way.

I own an SXRD RPTV... it is LCD mounted to a mirror (that's on a chip) so the same "rules" apply. It never color-shifts as you increase the contrast setting and it never crushes whites. It retains every step of detail at every contrast setting. You can run it right up to max contrast with the most wide-open iris setting and if the gains are adjusted properly, there's no color shift at the top end nor any loss of "steps" (crush). This is typical for properly designed LCD, LCoS, and DLP displays because these all have external light sources. The panel or panels are merely valves for the light if you want to think of them that way.

Almost every single LCD, LCOS, and DLP display I've ever seen most certainly does colorshift and/or clip when you blow white level through the roof.

It has nothing to do with the fact that they "all have external light sources."

I don't know where you came up with these "rules," but they're certainly not the norm on the VAST majority of displays.

Iris setting has no impact on this at all, unless it's a DI.

I've never heard or read of clipping being described in this way. Clipping is literally where the top of the waveform is clipped off.

I would happily throw crush under the bus and never use it again - as posted in other messages in this thread, brightness compression is MUCH better. It doesn't take clipping to cause brightness compression. But I still stick with the differentiator of clipping being color shifts because it is difficult, maybe impossible, to cause the primaries to all clip at the same luminance level.

And I still stick with my assertion that because LCDs are transmissive, they cannot clip but they can be excellent at brightness compression in a poorly designed display.

LCDs produce images from a light source that is on all the time. It doesn't change -- except new-tech LED backlights, but let's not confuse this discussion with that detail.

LCD pixels act like resistors act to an electrical signal. If you have a 12 volt electrical signal, you can produce hundreds of different voltages from it with resistors - just make each resistor a slightly different value to get a different voltage.

LCD pixels "resist" light more or less by how transparent or dark they become. At 0% (black) they only get so dark and leak some light because the backlight is on all the time. At 100% all LCD pixels achieve the same degree of transparency.

There is NOTHING the LCD pixels can clip because they don't produce light, they simply resist light. All you can have is a poor design that allows a 254 level to become a 100% transparent pixel or even worse... when all pixels from say 210 to 255 become 100% transparent or for the steps to be compressed because of poor design, you have no highlight detail because every detail in the image is at the same white level as the backlight.

When a pixel emits/produces/generates light it CAN clip - there CAN be 1 color that is weaker than the others. The red phosphor may be less efficient, for example.

http://s167.photobucket.com/albums/u147/maxdb/?action=view&current=red-clipping.jpg

In the image (I'm hoping this appears in the final version of the message, if not it can be seen at http://s167.photobucket.com/albums/u147/maxdb/?action=view&current=red-clipping.jpg ), Red peters out before green and blue, never achieving the same peak luminance level - this is really clipping. Red can't get any brighter - red is not compressing, it just can't get any brighter.

This clipping does NOT cause a loss of detail like crush does. Blue and green continue to provide differentiation between steps from 93% to 100% so steps and detail are all still visible when clipping exists.

This clipping (as illustrated in the image) causes a color shift in whites that starts around 93% and gets worse as you go to 100% white.

All 3 colors compressing at the same time is different than clipping. In fact what may have happened (for example) is that the poor design of the display has let 90% achieve the same luminance level as 100% - this may or may not be the peak luminance for the display.

When you compress brightness, you don't get a color shift, you just lose the steps/detail. It's clearly not too hard to make digital video displays that always have the 254 level darker than the 255 level and 253 slightly darker than 255. Many displays in all categories never compress brightness... never. But plasmas may clip red (or some other color).

brightness compression = loss of detail due to the panel not maintaining distinct/appropriate "levels" for each digital value from 0-155 or 16-235.

Clip=limiting of maximum luminance for 1 or 2 colors while the other color(s) continue to increase in luminance level. You do NOT get brightness compression from clipping. Pixels have to generate/create/produce light for clipping to be possible.

If all 3 colors produce the same luminance level at 90% and 100% - that's not clipping, that's just a poorly designed display with a Contrast control that is allowed to force the pixel levels at 90% to be 100% transparent - that's just bad design, not clipping. But it IS a form of brightness compression because you WILL lose all the detail between 90% and 100% without a color shift. Turning down the contrast control pull the 90% luminance level down where it is supposed to be (and if you turn Contrast down ENOUGH, the 100% peak white level will come down also).

Now... wiping out the use of the term "crush" is the next challenge.

Going by your resistor framework. Aren't bulbs and backlights going to have a certain spectral output that's not necessarily D65? If you want to try to calibrate grayscale, then doesn't one color become limiting factor in calibration and going any higher with contrast would cause a color shift toward the bulb or backlight's output?

So, would this be termed "white crush" or "brightness compression" or something else?
I am thinking of the details missing on the white shirt and the colours missing
on the guy's forehead in the second pic.

I would happily throw crush under the bus and never use it again - as posted in other messages in this thread, brightness compression is MUCH better. It doesn't take clipping to cause brightness compression. But I still stick with the differentiator of clipping being color shifts because it is difficult, maybe impossible, to cause the primaries to all clip at the same luminance level.

And I still stick with my assertion that because LCDs are transmissive, they cannot clip but they can be excellent at brightness compression in a poorly designed display.

That's fine, but that doesn't make it factual. I can walk you up to any LCD display and crank the white level through the roof and gee-willikers! There it goes into clipping. I can do it on this very LCD monitor. It clips, the whites get blown out, it looks terrible.

LCDs produce images from a light source that is on all the time. It doesn't change -- except new-tech LED backlights, but let's not confuse this discussion with that detail.

LCD pixels act like resistors act to an electrical signal. If you have a 12 volt electrical signal, you can produce hundreds of different voltages from it with resistors - just make each resistor a slightly different value to get a different voltage.

LCD pixels "resist" light more or less by how transparent or dark they become. At 0% (black) they only get so dark and leak some light because the backlight is on all the time. At 100% all LCD pixels achieve the same degree of transparency.

Okay, but how that's relevant to signal clipping I don't know. You can clip off the signal in the display, or you can do it before the display. The result is the same: clipped video. It can clip one or two of the primaries, or it can clip everything equally. Or heck it could clip chroma too for that mattter. Obviously where the clipping is occurring will have some bearing on what's being clipped and why. But it's still being clipped, whether it is part of the signal, or all of the signal. If you have a waveform, the peaks of which have literally been clipped off, that's called clipping. In this case, it's occurring in the digital domain. It's the same thing.


There is NOTHING the LCD pixels can clip because they don't produce light, they simply resist light. All you can have is a poor design that allows a 254 level to become a 100% transparent pixel or even worse... when all pixels from say 210 to 255 become 100% transparent or for the steps to be compressed because of poor design, you have no highlight detail because every detail in the image is at the same white level as the backlight.

What you just described IS clipping. That is what clipping IS. I mean, it literally is an identity. Clipping=clipping. You're claiming that in some cases, clipping=/= clipping because an LCD is involved, or because too much clipping is happening, therefore it should be called something else?

When a pixel emits/produces/generates light it CAN clip - there CAN be 1 color that is weaker than the others. The red phosphor may be less efficient, for example.

The nature of the display itself is completely irrelevant. If a portion of the signal is being clipped off, THAT IS CLIPPING. Whether the display element is reflecting light, transmitting light, emitting light, magix pixie dust, whatever, that has no relevance to whether the SIGNAL, as it is being reproduced, is being clipped off at the extremities for some reason that may or may not have anything to do with the display.


http://s167.photobucket.com/albums/u147/maxdb/?action=view&current=red-clipping.jpg

In the image (I'm hoping this appears in the final version of the message, if not it can be seen at http://s167.photobucket.com/albums/u147/maxdb/?action=view&current=red-clipping.jpg ), Red peters out before green and blue, never achieving the same peak luminance level - this is really clipping. Red can't get any brighter - red is not compressing, it just can't get any brighter.

That's called clipping.

And if it were red and green BOTH doing that, that's called clipping.

And if it were red, green, and blue ALL doing that, that's called clipping.


This clipping does NOT cause a loss of detail like crush does. Blue and green continue to provide differentiation between steps from 93% to 100% so steps and detail are all still visible when clipping exists.

I object to your bizarre choice of terminology, but you are correct. The visual effect of what is being clipped varies. However, certainly the effect on the particular signal that is being clipped is the same. If you viewed only the red channel, all detail above the clip point would be totally gone.

If you clipped luma, all detail above a particular level of white/gray would be completely gone. Likewise if you clipped R G and B all at the same time.

I fail to see where you get the idea that somehow clipping part of the signal is somehow fundamentally different than clipping the 'white' part of the signal. Yes it's obviously visually different, but it's still clipping.

Trying to claim that because not everything is being clipped equally that therefore it's something other than clipping because it looks different to you is a completely ridiculous proposition. That is like claiming that because all four of my tires are flat, I don't have a flat tire because it's very different than just having one flat tire.


This clipping (as illustrated in the image) causes a color shift in whites that starts around 93% and gets worse as you go to 100% white.

All 3 colors compressing at the same time is different than clipping. In fact what may have happened (for example) is that the poor design of the display has let 90% achieve the same luminance level as 100% - this may or may not be the peak luminance for the display.

How and why is that at ALL different? Whether you clip part of the signal, or clip all of the signal equally, or clip all of the signal but at different points, or clip at both ends, or clip at just one end, or whatever does not matter to whether or not it's clipping. If the signal is by definition being clipped, how is that not clipping!? I completely fail to grasp your point of view here.

When you compress brightness, you don't get a color shift, you just lose the steps/detail. It's clearly not too hard to make digital video displays that always have the 254 level darker than the 255 level and 253 slightly darker than 255. Many displays in all categories never compress brightness... never. But plasmas may clip red (or some other color).

So the visual impact differs. How is that relevant?

I mean, there's nobody that anywhere goes by this terminology, where you picked up this definition or got to this concept I do not know.

I could claim that if you clip the signal near black, that looks totally different than if you clip the signal near white (and of course it does look different), therefore one is clipping, but one isn't. One must be called something else, so I'll call it banana jello. That's the way you're attempting to differentiate between things, which you are also doing completely on your own and by using words that are pretty well defined and technically understood. If you want to make up new definitions for things, please at least make up new words or use words that aren't already defined in ways that totally differ from your usage of them.

brightness compression = loss of detail due to the panel not maintaining distinct/appropriate "levels" for each digital value from 0-155 or 16-235.

I agree with that.

Clip=limiting of maximum luminance for 1 or 2 colors while the other color(s) continue to increase in luminance level. You do NOT get brightness compression from clipping. Pixels have to generate/create/produce light for clipping to be possible.

I totally and vehemently disagree with that. You've come up with, totally on your own, a unique definition for clipping. Never have I EVER seen the definition limited in this way.

This is what clipping looks like:
http://i81.photobucket.com/albums/j230/sparky3489/Clipping.jpg

Link: http://i81.photobucket.com/albums/j230/sparky3489/Clipping.jpg

I just pulled this random image off the web, it's just a random clipped waveform. Now, simply because we have multiple signals that represent a full-color image doesn't mean that therefore the definition of clipping those signals changes to something else entirely. If you clip ALL of the signals, that's still clipping. I totally fail to see why, and of what use, is your bizarre choice to define clipping EVERYTHING in ONE particular way (at the same level such that there is no visible colorshift) as something entirely different for no good reason.

If all 3 colors produce the same luminance level at 90% and 100% - that's not clipping, that's just a poorly designed display with a Contrast control that is allowed to force the pixel levels at 90% to be 100% transparent - that's just bad design, not clipping. But it IS a form of brightness compression because you WILL lose all the detail between 90% and 100% without a color shift. Turning down the contrast control pull the 90% luminance level down where it is supposed to be (and if you turn Contrast down ENOUGH, the 100% peak white level will come down also).

Well, again, you're saying that up is down. Or at the very least, you're saying that in very particular situations, say on a tuesday afternoon, up is actually something else for a little while. Just because you decided it is, despite the fact that it goes completely against the normal and established definition of "up."

You're doing the same thing here. According to you, clipping signals sometimes isn't clipping if you clip the signals in a particular way, as defined by you, because you decided it should be that way and because you wanted it to be called something else for some reason that evades me.

Now... wiping out the use of the term "crush" is the next challenge.

And I wouldn't really mind that, particularly.

So, would this be termed "white crush" or "brightness compression" or something else?
I am thinking of the details missing on the white shirt and the colours missing
on the guy's forehead in the second pic.

That's called clipping. The whites are totally blown out. Above a certain value, there are no higher values, and everything gets clipped off above that point.

Some people might call that "white crush," but 'crush' is such a vague term it can be used for other things too.

I would not call it brightness compression. I agree with Greg R's use of the term brightness compression for when the signal is maintained, but the slope changes, which decreases the delta between levels, but does not eliminate those levels. This would decrease the visibility of fine details that are close in levels of luminance, and may even be severe enough to obscure them completely. I suppose if you compressed things enough such that the slope became very close to zero, it would in effect be the same as clipping. Indeed, if the slope is zero, that would be clipping.

Now, there are significant color differences in those two pictures two, but that's not what you were asking about I presume.

That's called clipping. The whites are totally blown out. Above a certain value, there are no higher values, and everything gets clipped off above that point.

Some people might call that "white crush," but 'crush' is such a vague term it can be used for other things too.

Well, Ok, I have called that "white crush" but I am ok with calling it clipping.


Now, there are significant color differences in those two pictures two, but that's not what you were asking about I presume.
No. :) but yes, the colours and most things are very much different.
The pictures are actually from two different editions of The Godfather.
Believe it or not, but the second one, with excessive clipping, is supposedly the new carefully restored edition! :confused:
I actually hope those screencaps are wrong since I cannot understand why they would destroy so much visible information.
For more caps, see http://www.zonadvd.com/modules.php?name=Sections&op=viewarticle&artid=843&page=2

Going by your resistor framework. Aren't bulbs and backlights going to have a certain spectral output that's not necessarily D65? If you want to try to calibrate grayscale, then doesn't one color become limiting factor in calibration and going any higher with contrast would cause a color shift toward the bulb or backlight's output?

In a cheap product perhaps, but I don't get calls for inexpensive displays very often. In well designed displays, the cuts/gains and the intial design of the display control how much 'headroom' there is in the Contrast setting for all 3 channels to go up/down without color shifting. I'm working with a new Samsung plasma that does not clip and there are no color shifts of conseqence when you run the Contrast control from the normal 30 fL for a darkened room to the max setting which is in the 50 fL range (white window pattern). The Sony SXRD XBR2 doesn't color shift or compress brightness as you increase the contrast control... but if you change from power save mode to standard mode, there is a small change in the color temp of the lamp that can be tweaked with cuts/gains. There's a low-end Sharp LCD upstairs (made by Sansui) that doesn't have brightness compression or color shift as you change the Contrast control. Changing the backlight control does shift color a little, but it is consistent for any contrast setting and can be "tuned" with a cut/gain adjustment.

If the real issue here is poorly designed lower-cost products vs those that are designed better and tend to be more accurate than the lower-cost products - so be it... as I said, I don't spend a lot of time with "low-end" or "lower cost" brands. I'm also not talking about anything but the last 2 or 3 generations of displays - yes, before that, things were progressively worse, but that's HISTORY, not the current state of the art.

The Sony SXRD XBR2 doesn't color shift or compress brightness as you increase the contrast control...

I own an XBR2 and it most certainly does color shift (I use an eye-one pro). The XBR2 runs out of red at higher contrast settings. You can run the contrast on the set fairly high if you are only measuring a 100% window, but if you measure a level 254 window it runs out at a much lower contrast setting. I actually use the point where red runs out at level 254 as my maximum contrast setting. I'm considering installing a CC20R filter to see if it will help me get more contrast since I have a lot of green and blue headroom while red has no headroom left. I'll most likely try it out since the filter only costs $50.

That's fine, but that doesn't make it factual.

Well, it doesn't make it common usage, but it's still factual.

To clip, you have to reach a physical limitation of a device... in the amplifier world, the limitation is the power supply and/or output devices in combination with the load. In a video display, the physical limitation is the inability of the display to produce a color of light, maybe even 2 colors of light.

I'm saying the physical design of LCDs, LCoS and even DLP puts them in the light valve category and as such, the valves are to light as resistors are to an electrical voltage/signal. A resistor cannot cause clipping. A light valve cannot cause clipping because it is not producing light, it is merely "throttling" light. The light that is controlled by light valves is what it is - different spectrums from different sources. The design of the device and the operation of the controls is supposed to take the light source and it's pecularities into account when the product is designed.

Bottom line, when the red phosphor reaches it's luminance limit before blue and green, that is "officially" clipping.

When a poorly designed light valve display produces a color shift at a high contrast setting - that's just a poor design, it is not clipping in the sense that the light source for red has reached a physical limit or nor is it a limit in the power supply that prevents red from getting brighter (which would also be clipping) because light valve displays don't have power supply limitations as plasma and CRT displays do. Their light source is on all the time and should be designed for within the product.

Brightness compression is a whole different thing and is 100% poor design and 100% NOT related to clipping when you apply engineering terminology to the use of the term clipping rather than let clipping fall into general use to describe things that are NOT clipping - but merely poor design choices.

I am making a distinction between a physical/power/physics limitation (clipping) and a poor design that may produce something that MIGHT look similar, but it's not clipping because no physical limitation is involved - it's just a matter of allowing light valves to operate outside of a linear range which is easy to prevent - and is prevented in "better" light valve displays.

In a cheap product perhaps, but I don't get calls for inexpensive displays very often. In well designed displays, the cuts/gains and the intial design of the display control how much 'headroom' there is in the Contrast setting for all 3 channels to go up/down without color shifting. I'm working with a new Samsung plasma that does not clip and there are no color shifts of conseqence when you run the Contrast control from the normal 30 fL for a darkened room to the max setting which is in the 50 fL range (white window pattern). The Sony SXRD XBR2 doesn't color shift or compress brightness as you increase the contrast control... but if you change from power save mode to standard mode, there is a small change in the color temp of the lamp that can be tweaked with cuts/gains. There's a low-end Sharp LCD upstairs (made by Sansui) that doesn't have brightness compression or color shift as you change the Contrast control. Changing the backlight control does shift color a little, but it is consistent for any contrast setting and can be "tuned" with a cut/gain adjustment.

If the real issue here is poorly designed lower-cost products vs those that are designed better and tend to be more accurate than the lower-cost products - so be it... as I said, I don't spend a lot of time with "low-end" or "lower cost" brands. I'm also not talking about anything but the last 2 or 3 generations of displays - yes, before that, things were progressively worse, but that's HISTORY, not the current state of the art.

I'd love to have products as you postulate, but I've never seen them. Every digital display I've encountered you can push into clipping pretty much. This includes professional/commercial displays, and certainly displays of way higher caliber than a consumer xbr2 which is nice, but certainly not what I'd consider 'high end' anyway. And they colorshift too. I don't know about samsung plasmas, but their LCD displays absolutely do clip.

Certainly the JVC and Sony projectors will, which are all LCOS. And DLPs sure will too, regardless of price, even if you blow past 6-figure pricetags. I've never seen one that didn't. Ever.

You could easily design one that didn't by limiting the adjustment range for white level, but that would be a very limited design and a poor engineering choice to make since it cripples the flexibility of the unit for the consumer.

Well, it doesn't make it common usage, but it's still factual.

To clip, you have to reach a physical limitation of a device... in the amplifier world, the limitation is the power supply and/or output devices in combination with the load. In a video display, the physical limitation is the inability of the display to produce a color of light, maybe even 2 colors of light.

But for some reason, you invent the concept that as soon as it's 3 colors of light, then it's somehow not clipping?

Again, I've gone through this in pretty excruciating detail, but you've not really even made an attempt to explain why or how you make this distinction. Nor do you provide any references for this bizarre use of terminology that flies in the face of pretty much anyone's understanding of signal clipping and every technical definition I've ever read of the term.

I'm saying the physical design of LCDs, LCoS and even DLP puts them in the light valve category and as such, the valves are to light as resistors are to an electrical voltage/signal. A resistor cannot cause clipping. A light valve cannot cause clipping because it is not producing light, it is merely "throttling" light. The light that is controlled by light valves is what it is - different spectrums from different sources. The design of the device and the operation of the controls is supposed to take the light source and it's pecularities into account when the product is designed.

I'm saying that this assertion is unfounded, and moreover logically absurd.

In other words, these displays can cause the signal literally to be clipped off, but because of some bizarre way you interpret the nature of the light production, somehow that's not clipping.

Sorry, that just doesn't fly.

Bottom line, when the red phosphor reaches it's luminance limit before blue and green, that is "officially" clipping.

And how is that at all different then running out of red on an LCD or LCOS or DLP, or a plasma display? It's not!

When a poorly designed light valve display produces a color shift at a high contrast setting - that's just a poor design, it is not clipping in the sense that the light source for red has reached a physical limit or nor is it a limit in the power supply that prevents red from getting brighter (which would also be clipping) because light valve displays don't have power supply limitations as plasma and CRT displays do. Their light source is on all the time and should be designed for within the product.

SO WHAT!? Who cares if the backlight or the bulb is on the whole time? The SIGNAL IS BEING CLIPPED! It results in an image that APPEARS "clipped." How on earth can you maintain that it isn't clipping simply because you invent some concept that because a display handles light in some different way than another display, somehow it's just totally different and completely immune from clipping. And "because I think so" is not a valid argument.

Brightness compression is a whole different thing and is 100% poor design and 100% NOT related to clipping when you apply engineering terminology to the use of the term clipping rather than let clipping fall into general use to describe things that are NOT clipping - but merely poor design choices.

I've tried to avoid talking about brightness compression because we can't even get to the must crude and basic understanding of clipping.

I am making a distinction between a physical/power/physics limitation (clipping) and a poor design that may produce something that MIGHT look similar, but it's not clipping because no physical limitation is involved - it's just a matter of allowing light valves to operate outside of a linear range which is easy to prevent - and is prevented in "better" light valve displays.

And you're just making things up. Ever post you make your explanation changes, now it's because you think there's something unique about "light valves" as you choose to characterize them.

I mean, I simply fail to get why you can't grasp what clipping is. Whether you clip one signal, or two signals, or three signals, or 8 million signals isn't relevant. IT'S CLIPPING. I mean, look at that picture of the scope I posted. That's TEXTBOOK clipping. That's what's happening in any digital display if you crank the white level too high (or the black level too low). The signal CLIPS. And as a consequence, you see an image that APPEARS CLIPPED. That's clipping.

Clipping is clipping. If you want to call it lime jello or 'magic light valve pixie dust' you can do that, but it's still completely ridiculous.

This is getting silly. I mean, we have a signal that is clipped. By definition. By any common sense. But somehow, because it's being displayed on a CERTAIN kind of display, that somehow ameliorates that or otherwise transforms it into not being clipped anymore because you feel like definining it in some other way to the exclusion of clipping.

If the signal is clipped, the image is clipped. It's THAT simple. clipping=clipping. I mean, this isn't some metaphysical question about whether things really exist, or whether a chair really is a chair. I don't see why this is so difficult.

I'd love to have products as you postulate, but I've never seen them. Every digital display I've encountered you can push into clipping pretty much. This includes professional/commercial displays, and certainly displays of way higher caliber than a consumer xbr2 which is nice, but certainly not what I'd consider 'high end' anyway. And they colorshift too. I don't know about samsung plasmas, but their LCD displays absolutely do clip.

If they are light valve products, they don't technically clip. Through design issues, they could have luminance compression in 1 channel or unequal luminance across the 3 channels, but they aren't clipping in the technical/engineering use of the term.


Certainly the JVC and Sony projectors will, which are all LCOS. And DLPs sure will too, regardless of price, even if you blow past 6-figure pricetags. I've never seen one that didn't. Ever.

Then it's luminance compression in 1 or more colors and not clipping. A light valve doesn't clip, it simply manages how much light gets through from the source. It can do this accurately or inaccurately, it can't clip.


You could easily design one that didn't by limiting the adjustment range for white level, but that would be a very limited design and a poor engineering choice to make since it cripples the flexibility of the unit for the consumer.

There's no reason in any light valve display for any pixel to respond to the 255 digital level with anything other than 100% transparency (or 100% reflectance in the case of DLPs). There's nothing advantageous to using any other scheme. Light sources may or may not be spectrally consistent, but properly adjusted gains/cuts should eliminate color shifts in light valve displays if all 3 chips/panels are working the same way (ie 100% transparent for the 255 level, slightly less transparent for the 254 level, etc.). If there is a single chip/panel there is the vagarity of the color wheel in relation to speed, color fidelity of the filters - but the single chip (LCD/LCoS/DLP) can't clip.

You can have luminance compression in any color or brightess compression (all 3 colors equally compressed), but you can't have clipping in a light valve display.

But for some reason, you invent the concept that as soon as it's 3 colors of light, then it's somehow not clipping?

Again, I've gone through this in pretty excruciating detail, but you've not really even made an attempt to explain why or how you make this distinction. Nor do you provide any references for this bizarre use of terminology that flies in the face of pretty much anyone's understanding of signal clipping and every technical definition I've ever read of the term.

I am changing the approach to the arguement in hopes of getting through, but I am not changing the bottom line statement that light valve displays do not clip... only a plasma or CRT can clip. I have never tried to explain this to somebody who didn't "get" it on the first attempt before.

You are making my point here all by yourself. There is no signal clipping in light valve displays. When they are being driven correctly, a 0 digital level produces the least transparency or least reflectance of light. When they are sent a 255 digital level, the LCD/LCoS pixels are 100% transparent and the DLP pixel is reflecting the maximum amount of light. They CANNOT clip, period as there is no signal TO CLIP you have digital levels setting transparency or reflectance levels for each pixel. That can be done wrong, but is is NOT clipping. Period.

I make this distinction between compression and clipping because for 34 years I worked on professional imaging systems, often in engineering and product development roles. We had to be precise about performance and causes/solutions for problems. Film is a light valve also - film cannot clip, but it can produce the appearance of clipping.


I'm saying that this assertion is unfounded, and moreover logically absurd.

It's absurd when an engineering concept becomes so perverted that things that are not clipping are called clipping because it's convenient or in general use... albeit incorrect general use. Sorry if it shakes your belief system, but this is reality. Your belief system re. clipping is just convenient.


In other words, these displays can cause the signal literally to be clipped off, but because of some bizarre way you interpret the nature of the light production, somehow that's not clipping.

Sorry, that just doesn't fly.

Light valve displays do NOT "clip the signal" period. It is impossible. It cannot exist and until you "get" that point, you'll never understand the concept. There is nothing bizarre about my interpretation of the nature of light, it is accurate from a physics/engineering point of view.

In light valve displays (using the LCD example in this case for simplicity), the pixels ALWAYS respond to digital 0 with 0 transparency and the 255 digital level is supposed to produce 100% transparency. Between 0% transparency and 100% transparency, there should be 256 distinct and measurable steps. If the display's controls alter that in any way... it is NOT CLIPPING because the digital signal is NOT CLIPPED, period. It is purely a matter of assigning an inappropriate transparency level to any given digital value. NO SIGNAL IS BEING CLIPPED. It's like using the wrong color matrix - the wrong look-up table.

In a plasma where a phosphor is producing light, it is possible for the display to send a digital 254 and for the phosphor to be producing the right amount of light for the 254 level. If the display then sends the 255 level and the phosphor is maxed out and does not get any brighter... that's clipping. The signal... the digital 255 cannot be reproduced by the phosphor. In this clipping the maximum light output of the pixel is LESS than the peak level because the peak cannot be achieved.

In compression, the 254 pixel becomes 100% transparent... that makes the display show the 254 level as the same as the 255 level. You do not lose peak luminance, you make the peak luminance higher for lower digital values. It is vey fudamentally a different condition than clipping.

In clipping, the display is saying "get brighter" and the phosphor is saying "I can't". In compression, the brightness is there, but the spacing of the 256 steps from 0-100% transparent (in LCD light valves) is wrong. These are not the same condition and even though it may be common practice to use the same word to describe both conditions, that dosn't make the common use accurate.


And how is that at all different then running out of red on an LCD or LCOS or DLP, or a plasma display? It's not!

If the light valve display is weak in red, it is a design problem, not clipping - for there to be clipping, the signal has to be compromised. Weak red in a light valve can only be caused by 2 things - the red LCD pixels are not as transparent as the blue or green pixels (no excuse for this other than exceptionally poor design) or the amount of red light is not in proper proportion to the amount of blue and green light which is also a design problem and NOT clipping. It means the lamp or backlight is deficient in red or that the filtering in the red light path is removing too much red light or the mirror in the red light path is not as efficient at reflecting red light as it is at reflecting green or blue light. But it is NOT clipping. It is a problem that is always there, designed into the product.


SO WHAT!? Who cares if the backlight or the bulb is on the whole time? The SIGNAL IS BEING CLIPPED!

No, the SIGNAL IS NOT BEING CLIPPED - it simply means 255 red is not as bright as 255 green or 255 blue. That is NOT clipping.


It results in an image that APPEARS "clipped."

Maybe maybe not. It could look like brightness compression which doesn't have quite the same appearance.


How on earth can you maintain that it isn't clipping simply because you invent some concept that because a display handles light in some different way than another display, somehow it's just totally different and completely immune from clipping. And "because I think so" is not a valid argument.

For something I "invented" you sure keep using it over and over again - you keep saying "the signal is clipped" when it is clear, the signal is NOT being clipped - there is a completely different reason the display is not doing the right thing. And yes, I am saying a light valve display CANNOT clip because the pixels always do what the display tells them to do... 256 steps from 0%-100% transparency. If the display tells the pixels to do something different, that's a design problem. If the light source for 1 color is weak... that's not clipping. The LCD panel will still make 256 steps from 0%-100% transparency. Unless the display compresses the steps, the image will look normal. Compressing the steps is not the same thing as clipping because the display is NOT CLIPPING THE SIGNAL. The display is intentionally manipulating the signal.


I've tried to avoid talking about brightness compression because we can't even get to the must crude and basic understanding of clipping.

I've got clipping under control - you're trying to use it to describe 2 different things and that's where the problem lies.


And you're just making things up. Ever post you make your explanation changes, now it's because you think there's something unique about "light valves" as you choose to characterize them.

Yeah, that's it, totally non-scientific, just dreaming it up for the sake of getting into a forum arguement with you.


I mean, I simply fail to get why you can't grasp what clipping is.

I grasp exactly what clipping is. The problem is that you aren't grasping what clipping isn't. You want to assign the same name to everything because it's habit/convenient even if it is not accurate.


Whether you clip one signal, or two signals, or three signals, or 8 million signals isn't relevant. IT'S CLIPPING. I mean, look at that picture of the scope I posted. That's TEXTBOOK clipping.

And it has NOTHING to do with what happens in a light valve display. If you want to see a 'scope representation of what happens in a light valve display, use the same sine wave input to the scope and put a 256-step resistor array between the sine wave and the 'scope. Turn the resistor array from 0 to 256 and you will se NO CLIPPING. And that's the WHOLE POINT. Light valves are resistors for light - they cannot clip.


That's what's happening in any digital display if you crank the white level too high (or the black level too low). The signal CLIPS. And as a consequence, you see an image that APPEARS CLIPPED. That's clipping.

Actually, the signal in a digital video display NEVER clips. Ever. You can end up with the wrong amount of light, but it's NOT from clipping - unless you have a light-emitting pixel that runs out of light before you run out of digital levels.


Clipping is clipping. If you want to call it lime jello or 'magic light valve pixie dust' you can do that, but it's still completely ridiculous.

Paraphrasing Arthur C. Clarke... When any technology or concept is sufficiently advanced, it appears as magic to a less technically advanced society...


This is getting silly. I mean, we have a signal that is clipped.

You keep saying that, but refuse to understand that in a light valve display, you are NOT CLIPPING THE SIGNAL - the transmissive or reflective pixels ALWAYS do what the display tells them to do. If the display tells them to all be 100% transparent for every digital value from 100-255, that's NOT CLIPPING. Period. The pixels are doing what they are being told to do. In a plasma, if the display is sending 255 and the pixel can only achieve the luminance level that should be digital 240 because the pixel can't physically produce any more light than that.... THAT is clipping because the signal and the pixel are no longer doing the same thing - this is the only case where the signal is compromised in any way.


By definition. By any common sense. But somehow, because it's being displayed on a CERTAIN kind of display, that somehow ameliorates that or otherwise transforms it into not being clipped anymore because you feel like definining it in some other way to the exclusion of clipping.

But you feel it is OK to assign clipping to anything that "looks" like clipping whether it is actually caused by clipping or not. I choose to make the distinction between clipping and some other cause for a luminance problem that is not acually clipping. Passive devices do not clip... period. Light valves are passive devices. LCD, LCoS, and DLP are light valves.


If the signal is clipped, the image is clipped.

In a digital world, the signal never clips. It will always have 256 steps. It's what happens to the signal when it is converted to light. If the light comes from a self-illuminating pixel, there can be clipping. If the light comes from a passive light valve device, clipping is not possible, but you can experience compression effects or ligh path design issues that produce effects people like to call clipping but they have nothing to do with the signal and pixel illumination getting out of sync. The passive pixels in light valves will do whatever they are told to do within their 0%-100% transparency for digital steps from 0-255.


It's THAT simple. clipping=clipping. I mean, this isn't some metaphysical question about whether things really exist, or whether a chair really is a chair. I don't see why this is so difficult.

Except when you assign clipping to luminance compression, give up science and enginering in the bargain. It's convienient to have 1 name for everything, but in the night sky, everything that looks like a star, isn't necessarily a star. And everything that looks like clipping isn't necessarily (and probably isn't) actual clipping.

I'm done with this... the discussion was pretty clear after the first 2 or 3 exchanges. You're not going to convince me of anything - I know perfectly well the point you are arguing, and from an engineering/technical point of view it is not accurate. From a general-use point of view, I'm nuts, in spite of the fact that the underlying cause of many luminance problems is not often actually clipping.

I... the discussion was pretty clear after the first 2 or 3 exchanges. You're not going to convince me of anything - I know perfectly well the point you are arguing, and from an engineering/technical point of view it is not accurate.

From a general-use point of view, I'm nuts, in spite of the fact that the underlying cause of many luminance problems is not often actually clipping.Thank you Doug! Your detailed explanation is quite succinct;)

If they are light valve products, they don't technically clip. Through design issues, they could have luminance compression in 1 channel or unequal luminance across the 3 channels, but they aren't clipping in the technical/engineering use of the term.

I have explained to you several times why this is not correct. You have not at all explained why anybody should accept your bizarre definition of clipping. I have repeatedly asked you to explain how or why "light valve" products are somehow unique or immune to clipping the signal. And to the contrary, there are numerous persons, not just myself, who attest to the fact that nearly all LCD, LCOS displays will certainly clip if you max out the white level.


Then it's luminance compression in 1 or more colors and not clipping. A light valve doesn't clip, it simply manages how much light gets through from the source. It can do this accurately or inaccurately, it can't clip.

Why not? There is no difference. The signal is being clipped off. That means it's clipping.

Why is this so complicated for you to understand?


There's no reason in any light valve display for any pixel to respond to the 255 digital level with anything other than 100% transparency (or 100% reflectance in the case of DLPs).

If it's calibrate to that, then yes you're right (putting aside that a DLP is always 100% or 0% and nothing in between). But the point is that if you adjust white level one click higher, the display's maximum white point is now below 255, and all values between the current maximum and 255 are now being clipped off.

That is the TEXTBOOK, classic definition of clipping. It's a hard clip, unambiguously so. It's not even a soft clip or bloom on a CRT which is mroe ambiguous. It's a hard, digital clip.

There's nothing advantageous to using any other scheme. Light sources may or may not be spectrally consistent, but properly adjusted gains/cuts should eliminate color shifts in light valve displays if all 3 chips/panels are working the same way (ie 100% transparent for the 255 level, slightly less transparent for the 254 level, etc.).

They won't be working the same way if you're calibrated to D65, because the light source is almost never D65. That's why the colorshifting.

If there is a single chip/panel there is the vagarity of the color wheel in relation to speed, color fidelity of the filters - but the single chip (LCD/LCoS/DLP) can't clip.

Not only can it, it most certainly does.

Your claims to the contrary cause me to question whether you've ever even SEEN an LCD, LCOS, or DLP display at all. It's absolutely ridiculous.

You can have luminance compression in any color or brightess compression (all 3 colors equally compressed), but you can't have clipping in a light valve display.

And I'm telling you that you're living in a fantasy world.

I've failed in getting through to you despite repeated explanations, so in the hope others won't stumble across this thread and actually give any heed to any of your other ideas, I'll let your own words speak for their own, utterly ridiculous, selves:

There is no signal clipping in light valve displays.

They CANNOT clip, period as there is no signal TO CLIP you have digital levels setting transparency or reflectance levels for each pixel.

Light valve displays do NOT "clip the signal" period. It is impossible.

In light valve displays (using the LCD example in this case for simplicity), the pixels ALWAYS respond to digital 0 with 0 transparency and the 255 digital level is supposed to produce 100% transparency.

Actually, the signal in a digital video display NEVER clips. Ever.

In a digital world, the signal never clips.

From a general-use point of view, I'm nuts,


One of these statements is accurate...

If the display tells them to all be 100% transparent for every digital value from 100-255, that's NOT CLIPPING. Period. The pixels are doing what they are being told to do.

The whole point of communication is to convey an idea. Google returned http://www.twysted-pair.com/dictc.htm and also http://dictionary.reference.com/browse/clip Sure the electronics definition of clip doesn't necessarily exactly line up with the standard definition, but this being a calibration forum what does it really hurt to use the latter? If all levels of gray from 45% up are displayed as the same shade, the output doesn't match the intent of the input, all information brighter than 45% is effectively not displayed because the digital levels end up the same brightness, what is so wrong with calling that situation clipping? In effect the electronics definition is that the output doesn't match the input, so I really fail to see how it's such an incorrect use of terms to call no change in luminance clipping and to call a less-than-intended difference in luminance crush.

Greetings

It's not the hill I want to die on, but we'd better damn well know what to do to fix this what ever it's called because a whole bunch of my info is still missing from the picture.

Regards

Most analog amplifiers are designed to operate within the linear portion of a lazy s curve. Once the input level exceeds a value that no longer produces gain in the output clipping occurs. Sometimes zener diodes are included to introduce clipping and limit output to a fixed level.

The whole point of communication is to convey an idea. Google returned http://www.twysted-pair.com/dictc.htm and also http://dictionary.reference.com/browse/clip Sure the electronics definition of clip doesn't necessarily exactly line up with the standard definition, but this being a calibration forum what does it really hurt to use the latter? If all levels of gray from 45% up are displayed as the same shade, the output doesn't match the intent of the input, all information brighter than 45% is effectively not displayed because the digital levels end up the same brightness, what is so wrong with calling that situation clipping? In effect the electronics definition is that the output doesn't match the input, so I really fail to see how it's such an incorrect use of terms to call no change in luminance clipping and to call a less-than-intended difference in luminance crush.

Because there is clipping condition and a brightness or luminance compression condition. They don't look the same and shouldn't be referred to with the same term. And there's also a peak white level limit - every display has one and that's not clipping or comression, it just IS. If the TV's controls are so whacked that they continue to try to make the TV brigher when it has already achieved peak brightness, again, it's a design problem, not clipping.

Why not use the right terms for different conditions?

There are plenty of ways for the output to not match the input that are most assuredly NOT clipping - and this refers to audio and video.

I've failed in getting through to you despite repeated explanations, so in the hope others won't stumble across this thread and actually give any heed to any of your other ideas, I'll let your own words speak for their own, utterly ridiculous, selves:


I've been polite and gentlemanly trying to explain something that is clearly over your head. You have chosen to take the insult route -- and to ignore some very simple technical concepts even though I tried approach them in different ways to help you understand. You so clearly did not understand what I was saying, that you even accused me of changing my "story". Frankly, the concepts are not that difficult. But you want a 1-word solution that paints eveything with the same brush, even if it is wrong. Have at it.

I've been polite and gentlemanly trying to explain something that is clearly over your head. You have chosen to take the insult route -- and to ignore some very simple technical concepts even though I tried approach them in different ways to help you understand. You so clearly did not understand what I was saying, that you even accused me of changing my "story". Frankly, the concepts are not that difficult. But you want a 1-word solution that paints eveything with the same brush, even if it is wrong. Have at it.

You never answered any of my questions. You never explained how it is that a clipped signal isn't clipping. You never forwarded ANY supporting sources for your bizarre definition of clipping. And it appears you have no experience adjusting the displays in question.

Given that it is standard practice for nearly every calibrator and professional on earth to calibrate digital displays by observing clipping, a pheonomenon that you claim without any experience, example, logic, or supporting information simply cannot occur, it seems to me that your bizarre definition here rightly should be ignored.

Nearly ever single digital display I have ever calibrated has clip points at black and at white, that occur at the display's maximums. Likewise every other calibrator on this forum would tell you the same thing.

I have attempted repeatedly to explain to you the nature of clipping, and what this looks like on a digital display.

As a purported calibrator yourself, it stuns me that you are unaware even of how to set black and white level properly on a digital display, which requires observing display behavior in exactly a way that you believe cannot occur. What disastrous results this leads to for your clients I cannot imagine. If you do not even understand signal clipping, and the setting of basic white and black level, how you proceed to do more in-depth adjustments, I do not know.

Because there is clipping condition and a brightness or luminance compression condition. They don't look the same and shouldn't be referred to with the same term.

And I've already explained this to you, and I referenced already Greg Roger's definition of brightness compression, particularly how he uses it to describe the effects of dynamic gamma in a DI system. Brightness compression is not a hard clip, but the compression of levels. It is a reduction in slope of gamma, which decreases the visibility of fine luminance deltas.

Clipping occurs in any digital or analog circuitry when an absolute maximum of signal excursion is reached beyond which no values exist, yet were present in the signal before this point in the circuitry. Obviously in a digital signal, the nature of the "excursion" is different than an analog waveform, but the nature of clipping remains the same.

This can occur in the display, or previous to the display. And it can occur at different points in the display as well.

And there's also a peak white level limit - every display has one and that's not clipping or comression,

And those peaks, which every digital display has, is a clip point. At that adjustment level and beyond, you begin to incur irreversable clipping of the video signal. All levels beyond a certain value are clipped off the signal. They are not reproduced by the display, and the image is clipped. Detail is removed permanently from the image.

I've explained this a half-dozen times to you already in this thread. I fail to see how you have such difficulty understanding this.

Here is the definition of clipping by Extron:
Clipping - Cutting off the peaks (or excursions) of a signal. A form of distortion that occurs when the signal excursions exceed the limits of the circuit.


Here is the definition of clipping provided by Charles Poynton:

Clip (v) Forcing a signal to a certain maximum(or minimum) level, so as to avoid excursion above (or below) that level.

This is exactly what you yourself have described when you raise white level (or lower black level) beyond a digital display's maximum. Values above that point are eliminated, and the signal excursion reaches an extreme.

Further, Poynton goes into more detail describing clipping with regards to brightness and contrast controls. Specifically regarding contrast (white level) control, he states:

...Contrast can be set to whatever level is appropriate for comfortable viewing, provided that clipping and blooming are avoided. In the studio, the Contrast control can be used to achieve the standard luminance of white, typically 103cdm^-2.

Later on page 28 he provides a graphic, which illustrates signal clipping with the adjustment of the brightness control, which comes with the following caption:
At the minimum and maximum settings, I show clipping to the Rec. 601 footroom of -15/219 and headroom of 238/219.

On the same page he provides a similar illustration regarding the contrast control:
Contrast (or video level) control in video provides a gain factor between roughly 0.5 and 2.0 to R'G'B' components. The output signal clips if the result would fall outside the range allowed for coding use. here I show clipping to the Rec. 601 headroom limit.

This is exactly what occurs, whether it happens to the signal prior to its arrival at the display, or within the display. And if it happens within the display, it is the same whether the cause is poor design of the display that causes the signal always to be clipped, or whether it is clipping caused by the user's misadjustment of black or white level controls. The results will be the same in all of these situations in terms of portions of the video signals reaching a maximum beyond which no values are produced. That is what clipping is. It is extremely damaging to the quality of the reproduced image, for obvious reasons.

For this reason, and because digital displays behave differently (they clip rather than bloom) than CRTs, it is standard procedure to observe clipping or partial clipping (colorshifting) of the signal particularly at white to align the display's peak white level to the signal's peak white level. This is the exact description by Joe Kane, using the Reverse Gray Ramps and Steps on DVE:

...
In a solid-state display it will make it easy to spot any clipping that may take place at the upper end of the dynamic range. It's the pattern of choice when setting the contrast control on a digital display.

Further he goes into detail:

Unlike the analog CRT, digital displays do not go into blooming. As the contrast is turned up the brighter portions of the video signal hit a clipping circuit, the maximum capability to pass video. Any information that is going to get to the display must be below the clip point in level. ... If the contrast is turned up chances are that part of the video signal will be pushed into this clipping circuit. Whatever detail is in that portion of the video signal will be turned into a flat white with no detail. The contrast control must be turned down until the detail re-appears. Run the contrast control up to see if the steps start to blend together.

This is the normal practice to locate the peak white point on a digital display, by using the display's white clip point. Yet you maintain, in the face of universal procedure and common sense, that digital displays CAN NEVER CLIP!

Here is Guy Kuo's description of exactly the same procedure, describing the White Level Bars on Avia PRO:

Similar to black level bars, Avia PRO includes animated indicators for white level. These are particularly useful with digital displays and processors that may clip signal levels more severely than older analog systems. ... Use of Avia PRO's white level bars helps avoid the common problem of near white detail clipping on digital displays.

And yet you claim, unbelievably, that despite the normal definition of clipping, and despite the universal procedure to locate a digital display's white point by observing the display's clip point, that digital displays can't actually clip at all.

That is a ridiculous and unsustainable assertion, particularly in the face of the numerous examples I've provided throughout this thread, not just this post.

But again, perhaps you, having sat through a seminar or two, know more about video than Charles Poynton. Or Joe Kane. Or Guy Kuo. Or the other thousands of engineers the world over who use the standard methodology that they describe in aligning digital displays. Apparently they're all on drugs, and you know better, those displays aren't clipping at all because it's not even possible for them to clip!

It's your made-up definition against the entire established video engineering community.

So in ever so clear terms, IT IS CLIPPING.

And here is another of Guy Kuo's description of clipping on digital displays:

If you have a lamp based display like an LCD or DLP, you don't have to worry about avoiding damage by having too high a contrast, but you still have to avoid hurting image quality.

...

On lamp based displays, having contrast [too high] doesn't shorten display life but causes clipping of hilites or shifts in the color of white as one primary color or another runs out of dynamic range. Clipping of hilites means that things that are bright but not quite white become indisinguishable from white. Shifts in the color of white means the color of white changes. On DLP's that shift is often blue-green.

But again, according to you, clipping a signal isn't actually clipping. Why? "Because Doug Blackburn said so." If you're not already aware, that is not a valid argument. Most people outgrow that rudimentary level of reasoning in grade school.

"the video signal will be pushed into this clipping circuit"

That's actually what he's addressing - that the TV's electronic circuits aren't necessarily being pushed into electronic clipping. I find the point very silly when the base word clip means such things as to cut short, curtail, limit and when sounds as simple as 'too' have to be put in context in english. For how-to descriptions I really don't grasp how using the word clip to describe a limit is really too confusing. If it's important that a term needs to be defined precisely, then do so in the relevant discussion. As an example Avia and TomHuffman's guide use saturation to mean entirely different things, but they each define their terms in their discussion. Maybe it's just me but having saturation mean two things, or having different IRE values to represent black, seems no more confusing than having clip have more than one singular intended meaning.

You never answered any of my questions. You never explained how it is that a clipped signal isn't clipping. You never forwarded ANY supporting sources for your bizarre definition of clipping. And it appears you have no experience adjusting the displays in question.


I answered all your questions, but you chose to ignore the answers or they went right over your head.

You keep referring to "clipped signal" as a key indicator of clipping. More than once, I tried to explain to you that in the digital domain, THERE IS NO SIGNAL CLIPPING. PERIOD. EVER. You never lose numbers. You are essentially sending numbers over wires. In our case, the numbers are 0-255. None of them are EVER clipped when they exist as digital signals. But you keep saying signal signal signal and providing definitions of clipping that apply only to analog signals that have no application in anything digital.

If you insist on applying analog signal-level teminology to a luminance effect caused by setting pixels to inappopriate values (because of poorly designed controls on the display, not because any signal is EVER being clipped), more power to you. Would you say a 4 cylinder Toyota Camry clips at 105mph just because it can't attain 120mph? No. Then why is it OK to say one video display clips at 40 fL just because another display can achieve 50 fL? It's not. What if you have 2 displays, LCD panels, they both achieve 80 fL of peak white level. On 1 display the contrast setting is 100 to achieve 80 fL. Turn down the contrast to 99 and you read 79.9 fL. So this one has a nice linear contrast control. And you can see all the steps from 200-255. The other display produces 80 fL with the contrast control set to 50 and you see all the steps from 200-255. You reduce the contrast control to 49 and you measure 79.7 fL. You increase the Contrast setting to 100. You still measure 80 fL, but all the steps from 180 to 255 measure 80 fL.

You say the latter case is clipping. I say the latter case is just poorly designed controls in the video display. It's compression of the gray scale. There is NO SIGNAL BEING CLIPPED if you examine the electrical signal, it still contains every number from 0-255, the video display merely alters the lookup table that drives the pixels to various levels of transparency. Nothing is being clipped, but the pixels are being driven inappropriately. Yes, you lose visible steps. You say "that's clipping" and I say "It may look like clipping, but there is no clipping at the signal level that is causing that appearance, it is simply a poorly designed control." Every pixel in the "clipped" display is doing EXACTLY what the display is telling it to do. When you have a clipped electronic signal, the output device (say a loudspeaker) cannot reproduce the original analog signal because that signal has been permanently damaged by the clipping of the analog amplifier output.

You can say "Whatever, I'm still calling it clipping." and I would say "Whatever, as long as you understand, that's not really what you are seeing. You are only seeing something that LOOKS sort of like clipping might look if clipping could exist in the digital domain."
\

Given that it is standard practice for nearly every calibrator and professional on earth to calibrate digital displays by observing clipping,

They are observing something... usually luminance compression, not clipping. They and you call it clipping because some non-engineering person used the term and it has been repeated many times since then. Clipping is something that happens to an analog signal. In a digital display, you can have a condition where 0-15 are never displayed (display won't show blacker than black). Some people might try to call that clipping. But it is not clipping. It is merely the display setting up a lookup table that has digital 0 displayed with the same pixel transparency as digital 16. Digital 1 is also assigned the same transparency as digital 16. Nothing is clipped, all the numbers are still there, and all the pixels are doing exactly what the display is telling them to do.


a pheonomenon that you claim without any experience, example, logic, or supporting information simply cannot occur, it seems to me that your bizarre definition here rightly should be ignored.

As I said before, I have 34 years of experience with professional imaging equipment, the last 20 years were 100% digital imaging products in product development, engineering, and product support roles. Engineers often know what they are talking about and tend to buck the technically incorrect use of engineering concepts no matter how "popular" the incorrect usage may become.


Nearly ever single digital display I have ever calibrated has clip points at black and at white, that occur at the display's maximums. Likewise every other calibrator on this forum would tell you the same thing.


I have attempted repeatedly to explain to you the nature of clipping, and what this looks like on a digital display.

Explaining it wrong doesn't make it right. Explain it 1000 times and it will still be TECHNICALY wrong, even if it is in popular use. I have no problem understanding what you are describing. But you aren't describing clipping. You keep saying "the signal is clipped"... the signal the signal the signal. And I keep saying there is NO SIGNAL CLIPPING in what you are seeing. You are seeing the effects of a display's controls doing something wrong that is NOT clipping.

Now, if you completely give up your obsession with THE SIGNAL, and you say "I'm going to call this effect where there are missing or compressed measured LIGHT steps CLIPPING even though it has NOTHING to do with the SIGNAL." OK, you are hijacking an analog electrical signal term and applying it to luminance performance of a video display just for convenience. I can almost live with that if you just get over the fact that it has NOTHING TO DO WITH THE SIGNAL being clipped and everything to do with how the controls work (incorrectly if there's a visible problem) in the display.


As a purported calibrator yourself, it stuns me that you are unaware even of how to set black and white level properly on a digital display,

OK back to the insults you go - you know damn well this statemet is a crock.



which requires observing display behavior in exactly a way that you believe cannot occur.

This is not about belief. This is about what is or what is not clipping of signals as you keep focusing on. Setting black level and white level has nothing to do with clipping any electrical signal. It has to do with the luminance performance of the display in question. Period. Either the display is doing the right thing or it is not. [[[[ There is 1 exception to this that I have mentioned in previous posts - in plasma displays, there can be something that I can accept as being real clipping when 1 color cannot achieve the same luminance levels as the other colors - in that case, the display is trying to drive the color in question brighter (with higher digital numbers) but that color simply can't get any brighter]]]] This is not the same thing that is happening in LCDs where the 3 layers of pixels are all essentially identical (or where there are 3 LCD panels or LCoS chips that are identical).


What disastrous results this leads to for your clients I cannot imagine.

More insults. Are you able to keep this civil or not? Yes, my former employer was so disturbed by my performance that I was kept around for 34 years. Shocking. It's also shocking I would be a professional reviewer for a national magazine.


If you do not even understand signal clipping, and the setting of basic white and black level, how you proceed to do more in-depth adjustments, I do not know.

If you believe there is ANY electronic signal being compromised in what you are calling clipping, you are deluded and truly out of your element - and this final attempt to get you to understand that in digital products, there is no clipping of signals fails, you have no business arguing your point from your technological vacuum.


Clipping occurs in any digital or analog circuitry when an absolute maximum of signal excursion is reached beyond which no values exist, yet were present in the signal before this point in the circuitry.

Except signal excursion has nothing to do with clipping in the digital domain. If the signal starts out containing the values 0-255, it will have the values 0-255 right up to the instant a signal is applied to a pixel element. If you reduce the "signal excursion" of a digital signal at some point you lose the entire sigal, you don't lose high numbers first or low numbers first. In a digital signal, all numbers are "equal" they are merely encoded differently so they can be extracted accurately. The issue in digital displays is after the numbers are extracted, you will still have a range of numbers from 0-255. Something has to control the video display. Typically this is a lookup table that essentially says "for the digital value '255' send this voltage to this pixel" usually that value will make the pixel 100% transparent which is what you'd want for 255. Digital 254 should send a voltage that makes the pixel just a little less transparent (or flash the pixel less frequently). The Contrast control can change the top end of the lookup table, the brightness control can control the bottom end of the lookup table. In a properly designed display, nothing inappropriate will happen as you adjust contrast and brightness. In a poorly designed display, the lookup table could end up telling the video display that for digital values 200-255 send them all the same voltage (or frequency) to make them 100% transparent. Do you see that there is NO SIGNAL CLIPPING involved here? Yes, you get the wrong result and you lose steps that should be visible, but it has NOTHING to do with clipping any signal.


Obviously in a digital signal, the nature of the "excursion" is different than an analog waveform, but the nature of clipping remains the same.

This can occur in the display, or previous to the display. And it can occur at different points in the display as well.

None of this has any technical accuracy. It's like you know the words, but have no grasp of their meaning. It would take pages and pages to unravel this.

I will try a simple example... digital audio... there is an effect a lot of people call "clipping" in digital audio. If you playback a CD, for example, and capture the analog output of the CD player and display it in real time on a 'scope, many discs will appear to have the highest peaks "clipped". This is encoded on the CD. It is not wrong, the CD has accurately captured the digital master. The problem exists in the digital master. What is the problem? Is it clipping? I say, no, it is NOT clipping. It is merely a mistake. The engineer set the level of the analog input signal too high for the dynamic range of the digital audio system he was working in (16 bits in the case of CD audio). NOTHING clipped in the digital domain, it just ran out of bits because of human error on the part of the engineer. Everything worked exactly as it should work. The only problem was that the engineer set the analog input level too high and this produced something sort of like clipping, but there was no clipping of ANYTHING in the digital domain, and the analog input signal as it existed on tape was not clipped. In effect, the engineer setup a poorly designed lookup table for the analog voltage to digital bits conversion table.

In consumer digital video, the sources we are talking about are already digital. From disc player to video display - there is no clipping in the digital domain. Anything you see on the display that you don't like is the result of controls that are allowed some degree of luminance compression from mild to severe and this may be unequal in each color depending on how the controls are designed.



And those peaks, which every digital display has, is a clip point. At that adjustment level and beyond, you begin to incur irreversable clipping of the video signal. All levels beyond a certain value are clipped off the signal. They are not reproduced by the display, and the image is clipped. Detail is removed permanently from the image.

Sorry, this is also technically incorrect. "The video signal" in digital displays is NEVER clipped (except in some plasma panels addressed elsewhere). Certainly you see luminance compression and unbalanced color - but it does not come from any SIGNAL being clipped (except the plasma case again). Period. It comes from the display doing exactly what it has been designed to do... good or bad. No values are lost, none are "missing" - and every pixel is doing exactly what has been encoded in the lookup table, right or wrong. Poor design that allows luminance compression and color shifts as you change the contrast control is simply poor design. There is no clipping of any electrical signal involved. Again, you can declare you're going to call the look of those missing steps or color shifts Clipping, but there's no SIGNAL (digital electroic) being clipped anywhere.


I've explained this a half-dozen times to you already in this thread. I fail to see how you have such difficulty understanding this.

Again, explaining it incorrectly half-a-dozen times doesn't make it right.


Here is the definition of clipping by Extron:


Here is the definition of clipping provided by Charles Poynton:



This is exactly what you yourself have described when you raise white level (or lower black level) beyond a digital display's maximum. Values above that point are eliminated, and the signal excursion reaches an extreme.

No, it isn't what I've been describing. No value between 0-255 in these digital displays is EVER eliminated at the electrical/electronic level. Severe adjustments of user controls can, as I've said a NUMBER of times (are you reading this?) cause poorly designed controls to assign lookup table values that produce luminance compression or complete loss of steps so the images have an APPEARANCE that there has been clipping, but if you look inside the display to see what really happend... there was no loss of ANY digital value from 0-255, the display simply provided poorly conceived controls that use all the numbers, but assign inappropriate levels to some of the numbers. All the numbers are there, all the pixels are doing exactly what they are being told to do by the display.

Going back to the example of the car with a top speed of 105. Let's say there's a digital accelerator pedal. It is encoded with values from 0-255. 255 is "top speed", 0 is idle. Let's say when you buy the car, there's a lookup table in the engine computer that says, for throttle position 230, set throttle to 90% open. For throttle position 231, set throttle to 90% open, and for every position from 230-255, the lookup table says "set throttle to 90% open. This limits top speed to 92 mph instead of the maximum the car is capable if the throttle is 100% open. Is this clipping? I say no. Everything is doing exactly what it is being told to do by the program in the engine computer. YES, top speed is being limited. And you might want to call this clipping, but it isn't. It is merely a design/programming issue. You could, in theory, hack the engine computer and make a new lookup table that would allow the throttle to achieve 100% open. You haven't changed how the car handles any digital (or analog) signal, but the top speed restriction is gone. Lookup tables for pixels are the same sort of thing except there are 6.2 million pixels wanting to know what their level of transparency is supposed to be for the next 1/60th of a second. If software/firmware tells them an inappropriate level - that's all it is, inappropriate programming, it's not technically clipping since the signal is always intact.


This is the normal practice to locate the peak white point on a digital display, by using the display's white clip point. Yet you maintain, in the face of universal procedure and common sense, that digital displays CAN NEVER CLIP!

Again you are missing the whole point - and this continues to go over your head. I never said digital displays never clip - I did say at the beginning that I don't lose steps as I run Contrast up and down in the well-designed displays I have worked on - Michael pointed out that the Westinghouse LCDs do that. Fine. Point taken. But if you use the model of clipping being the intentional or unintentional peak-limiting of an electronic SIGNAL... clipping never happens in digital displays. When we see problems, we are not seeing a clipped elctronic signal, we are seeing and measuring is the result of poor programming or design. To make a projection lamp optical system less expensive, it may have not quite the right proportions of red, green, and blue light and these lookup tables that control the pixels are used to compensate... up to a point. Similarly, a backlight with a less ideal light spectrum may be selected for a low-cost LCD and the lookup tables are used as somewhat of a compensator for the light spectrum not being ideal because that's cheaper than using the right filters or a more expensive light source. I have no problem with displays like the Westinghouse LCDs having luminance compression problems that you and everybody else want to call clipping. But if people are using the word clipping to indicate that electrical signals are being clipped to cause what you are seeing on the display, that's incorrect. What you are seeing on the display is a result of lower cost optics or light sources and using software to try to compensate.


And yet you claim, unbelievably, that despite the normal definition of clipping, and despite the universal procedure to locate a digital display's white point by observing the display's clip point, that digital displays can't actually clip at all.

Not at the electronic digital signal level, no.


That is a ridiculous and unsustainable assertion, particularly in the face of the numerous examples I've provided throughout this thread, not just this post.

Except that some of your "examples" make my point, not yours. The fact that you think all of those examples "prove" there is electronic clipping of signals causing visible loss of steps or color shifts indicates you don't understand the underlying causes.


But again, perhaps you, having sat through a seminar or two, know more about video than Charles Poynton. Or Joe Kane. Or Guy Kuo. Or the other thousands of engineers the world over who use the standard methodology that they describe in aligning digital displays. Apparently they're all on drugs, and you know better, those displays aren't clipping at all because it's not even possible for them to clip!

Or perhaps it is because I've worked with imaging sytems for 34 years that I actually know what I'm talking about. You've watched movies that were digitally mastered on equipment I worked on.


It's your made-up definition against the entire established video engineering community.

There's nothing made-up in my definition. If you understood digital imaging systems, that would be clear to you. The calibration community has selected a word of convenience. If you limit the scope of the word to describing luminance compression (loss of steps) or color shifts at extreme contrast settings and everybody agrees to use the word to describe THAT... fine, whatever. But that clipped scope photo you posted is clipping and that's not what causes the luminance compression and color shifts you see in video displays (except many plasmas which I covered elsewhere).


So in ever so clear terms, IT IS CLIPPING.

Only if you give up blaming the visible effects you are calling clipping on the digital signal in electronic form being electronically clipped in any way. If you want to have 2 completely different defintions of the words clip/clipping you have to recognize the difference between the 2 usages. Because what you see on video displays that is being called clipping is not coming from electronic signals being clipped - its coming from a combination of poor design of the controls or using programming/lookup tables to compensate for illumination system issues rather than physically fixing the illumination system problems.

But again, according to you, clipping a signal isn't actually clipping. Why? "Because Doug Blackburn said so." If you're not already aware, that is not a valid argument. Most people outgrow that rudimentary level of reasoning in grade school.

You clearly do not understand digital circuits and digital video displays. Clipping a signal is always clipping a signal. But the luminace effects you are calling clipping do not result from any electronic signal being clipped. If most people outgrow your reasoning in grade school, why do you keep subjecting me to it?

It all depends on what is is

I answered all your questions, but you chose to ignore the answers or they went right over your head.

No, you most certainly did not. I have repeatedly asked you to source or even define on your own what you mean by clipping.

I have repeatedly attempted to communicate to you what I mean by clipping. I did so in my own words, with pictures, and with definitions from other sources.

You have forwarded no definition whatsoever by what is meant by clipping.

You have forwarded no definition of your own words either.

In addition, I also forwarded a definition of levels "compression" as used in this context, and included a reference to Greg Roger's definition as well.


You keep referring to "clipped signal" as a key indicator of clipping. More than once, I tried to explain to you that in the digital domain, THERE IS NO SIGNAL CLIPPING. PERIOD. EVER. You never lose numbers.

This is the major problem we are having in this conversation. This is an erroneous understanding. Clipping absolutely does occur in the digital domain, and is repeatedly a concern in any digital video system.

It is similar in nature to analog clipping, but is far more absolute. There is little ambiguity about digital clipping because it is a numerical hard boundry that is absolute.

For instance, in 8-bit content, you have a hard, absolute maximum at 255. It is impossible to represent a value above this. It is an absolute maximum. If you attempt to feed values above this to an encoder for instance, they will be clipped off.

It appears that in your conception, this would not be "clipping" because it would require user error to do so. In other words, it appears that since it simply is not possible for the digital device to do anything other than stop at 255, the signals previous to this location are not actually being clipped, but subject to some kind of user error that results in some other, undefined, damage to the signal.

I strongly disagree with this bizarre conception of clipping, and is clearly contrary to the numerous definitions of clipping I provided from reference/professional sources.

You are essentially sending numbers over wires. In our case, the numbers are 0-255. None of them are EVER clipped when they exist as digital signals. But you keep saying signal signal signal and providing definitions of clipping that apply only to analog signals that have no application in anything digital.

Well, your ignorance of digital clipping does not mean that digital clipping does not occur. Anyone at all familiar with digital images is familiar with the problems of clipped content. Indeed, I have written an entire guide that delves into devices that clip the video range back to 16-235. This is objectively measurable, and objectively visible on test patterns. Not only that, if you have a DAC step, you can also measure it visibly in the analog domain, and the resulting analog waveform is identical to what you'd expect if you clipped the signal in the analog domain.

The loss of values is clipping whether it occurs in the analog domain or in the digital domain.

These are not particularly detalied, but certainly include discussions of digital clipping:

http://en.wikipedia.org/wiki/Clipping_%28signal_processing%29#Digital_processing

http://en.wikipedia.org/wiki/Clipping_%28audio%29

Again, it appears that this is the fundamental understanding-gap. If you are of the idea that clipping simply does not occur in the digital domain, then your position makes sense. Unfortunately, this is a silly position, which hopefully you will change as you learn more about the digital domain.

If you insist on applying analog signal-level teminology to a luminance effect caused by setting pixels to inappopriate values (because of poorly designed controls on the display, not because any signal is EVER being clipped), more power to you.

I'm not using analog signal-terminology at all. I'm using the term "clipping." The term clipping applies to signals, not just particular kinds of analog signals. They apply to any signal that has an excursion, whether in the domain of voltage, or numbers, or whatever the domain may be.

The definitions I provided make no mention of analog signals at all, instead they refer to signals generally. And further quotations from the same sources, particularly Poynton, Joe Kane, and Guy Kuo all include explicit mentions of clipping in digital circuitry or of digital levels.


Would you say a 4 cylinder Toyota Camry clips at 105mph just because it can't attain 120mph? No. Then why is it OK to say one video display clips at 40 fL just because another display can achieve 50 fL?

I never said that. Please point to anywhere in any post where I said anything of the sort. In fact, I never used ANY physical measurements or quantities of light at all, EVER.

It's not. What if you have 2 displays, LCD panels, they both achieve 80 fL of peak white level. On 1 display the contrast setting is 100 to achieve 80 fL. Turn down the contrast to 99 and you read 79.9 fL. So this one has a nice linear contrast control. And you can see all the steps from 200-255. The other display produces 80 fL with the contrast control set to 50 and you see all the steps from 200-255. You reduce the contrast control to 49 and you measure 79.7 fL. You increase the Contrast setting to 100. You still measure 80 fL, but all the steps from 180 to 255 measure 80 fL.

You say the latter case is clipping.

Okay, lets go step by step through the logic.

The display circuitry is set in a way that values above digital level 180 are not reproduced correctly. Instead, they are all reproduced at the same maximum as produced by the input signal of 180, the 80fL in your example.

In your example, the signal has been forced by user error, into a situation where signal excursion above level 180 is eliminated. No signal values above level 180 are reproduced correctly, since they are all reproduced at the same luminance as level 180.

In other words, in the example you provided, you have

Forc[ed] a signal to a certain maximum(or minimum) level, so as to avoid excursion above (or below) that level.

That is exactly what you have described in your example, and this is exactly Charles Poynton's definition of clipping that I provided previously.

I say the latter case is just poorly designed controls in the video display.

And if you stop and look at the definition of clipping provided by Charles Poynton, it is a literal and explicit definition of EXACTLY what you have described.

It's compression of the gray scale.

I've already explained and referenced my definition for compression. I have repeatedly stated that it is a reduction of slope but not to a slope of 0.

There is NO SIGNAL BEING CLIPPED if you examine the electrical signal, it still contains every number from 0-255, the video display merely alters the lookup table that drives the pixels to various levels of transparency. Nothing is being clipped, but the pixels are being driven inappropriately. Yes, you lose visible steps. You say "that's clipping" and I say "It may look like clipping, but there is no clipping at the signal level that is causing that appearance, it is simply a poorly designed control." Every pixel in the "clipped" display is doing EXACTLY what the display is telling it to do.

How can you claim that? The signal, i.e. the display circuitry is saying "get brighter" or "get more reflective" or "get more translucent" and it cannot. It isn't doing what it should be doing according to how the signal is aligned. In this particular example it's not the display's fault, but rather user error in misaligning the display OR feeding a signal to which it cannot fully respond. This is the exact definition of clipping:

Clip (v) Forcing a signal to a certain maximum(or minimum) level, so as to avoid excursion above (or below) that level.

When you have a clipped electronic signal, the output device (say a loudspeaker) cannot reproduce the original analog signal because that signal has been permanently damaged by the clipping of the analog amplifier output.

You can say "Whatever, I'm still calling it clipping."

No, I'm saying that the entire video engineering community agree that it is so, having arrived at a well-understood definition of clipping as:

Clip (v) Forcing a signal to a certain maximum(or minimum) level, so as to avoid excursion above (or below) that level.

That's why the various calibration explanations from Joe Kane and Guy Kuo BOTH explicitly describe the digital white-clip point on a digital display in their alignment methodology.

and I would say "Whatever, as long as you understand, that's not really what you are seeing. You are only seeing something that LOOKS sort of like clipping might look if clipping could exist in the digital domain."
\

No.

I am saying that it IS clipping, and it DOES exist in the digital domain.

There really is nothing to argue about here. It's right there in plain english.

They are observing something... usually luminance compression, not clipping. They and you call it clipping because some non-engineering person used the term and it has been repeated many times since then.

No, that is the engineering definition of clipping.

If you're calling Charles Poynton "some non-engineering person" then feel free.

Clipping is something that happens to an analog signal.

And a digital signal:

http://www.atpm.com/6.03/digitalaudio.shtml
Clipping
Both analog and digital media have an upper limit beyond which they can no longer accurately represent amplitude. Analog clipping (or overdrive or distortion) varies in quality depending on the medium. A tube amplifier, for example, has a much warmer distortion than a solid state amplifier. In each case the upper amplitudes are being altered, distorting the waveform and changing the timbre, but the alterations are slightly different. Digital clipping, in contrast, is always the same. Once an amplitude of 1111111111111111 (the maximum value in a 16 bit resolution) is reached, no higher amplitudes can be represented. The result is not the smooth, rounded flattening of analog clipping, but a harsh slicing of off the top of the waveform, and an unpleasant timbral result.


In a digital display, you can have a condition where 0-15 are never displayed (display won't show blacker than black). Some people might try to call that clipping.

Those people are called video engineers.

But it is not clipping. It is merely the display setting up a lookup table that has digital 0 displayed with the same pixel transparency as digital 16. Digital 1 is also assigned the same transparency as digital 16. Nothing is clipped, all the numbers are still there, and all the pixels are doing exactly what the display is telling them to do.

Except that they aren't. There is a disconnect between the input signal, and the result or output signal. And in this case, the particular error is a clipping error, a maximum or minimum have been exceeded, and signal excursion beyond that maximum or minimum have been lost.

Clip (v) Forcing a signal to a certain maximum(or minimum) level, so as to avoid excursion above (or below) that level.


As I said before, I have 34 years of experience with professional imaging equipment, the last 20 years were 100% digital imaging products in product development, engineering, and product support roles. Engineers often know what they are talking about and tend to buck the technically incorrect use of engineering concepts no matter how "popular" the incorrect usage may become.

And despite all that, sadly, you are not even aware that clipping happens in the digital domain.


Explaining it wrong doesn't make it right. Explain it 1000 times and it will still be TECHNICALY wrong, even if it is in popular use. I have no problem understanding what you are describing. But you aren't describing clipping. You keep saying "the signal is clipped"... the signal the signal the signal. And I keep saying there is NO SIGNAL CLIPPING in what you are seeing. You are seeing the effects of a display's controls doing something wrong that is NOT clipping.

That's why I have repeatedly cited other reputable sources to bolster my explanation.

But again, maybe Charles Poynton, who wrote *THE* definitive book on digital video, is just including non-technical, popular-use definition in the glossary of his book.

Now, if you completely give up your obsession with THE SIGNAL, and you say "I'm going to call this effect where there are missing or compressed measured LIGHT steps CLIPPING even though it has NOTHING to do with the SIGNAL." OK, you are hijacking an analog electrical signal term and applying it to luminance performance of a video display just for convenience. I can almost live with that if you just get over the fact that it has NOTHING TO DO WITH THE SIGNAL being clipped and everything to do with how the controls work (incorrectly if there's a visible problem) in the display.

You are the one who is strangely of the belief that clipping only occurs in the analog domain.

That belief is ridiculous.


This is not about belief. This is about what is or what is not clipping of signals as you keep focusing on. Setting black level and white level has nothing to do with clipping any electrical signal. It has to do with the luminance performance of the display in question. Period. Either the display is doing the right thing or it is not.

What is the right thing? If there is an adjustment with enough range, you will be able to push the display circuitry into a situation where it is clipping the signal.

That is exactly what you described in your example above, and exactly as described in Poyton's simple and effective definition of clipping:

Clip (v) Forcing a signal to a certain maximum(or minimum) level, so as to avoid excursion above (or below) that level.


[[[[ There is 1 exception to this that I have mentioned in previous posts - in plasma displays, there can be something that I can accept as being real clipping when 1 color cannot achieve the same luminance levels as the other colors - in that case, the display is trying to drive the color in question brighter (with higher digital numbers) but that color simply can't get any brighter]]]] This is not the same thing that is happening in LCDs where the 3 layers of pixels are all essentially identical (or where there are 3 LCD panels or LCoS chips that are identical).

So when one color in an LCD or LCOS display cannot acheive the same luminance level as another color, how is that not clipping? That occurs on almsot every such display on earth. And the few that don't, instead of clipping one color first, will clip all three at the same time.


If you believe there is ANY electronic signal being compromised in what you are calling clipping, you are deluded and truly out of your element - and this final attempt to get you to understand that in digital products, there is no clipping of signals fails, you have no business arguing your point from your technological vacuum.

What should I believe? There are little gnomes?

Of COURSE there is a signal being compromised in what I am calling clipping. It's the video signal! It's being clipped! In the digital domain! You can measure it! You can SEE it! And it looks like sh*t!

I think it's hilarious that you are of the idea that you are going to explain not only to me, but to Charles Poynton, Guy Kuo, Joe Kane, and others, that there is never any clipping in digital signals.

It's ridiculous.

Aside from all the above examples I have already provided, here is a published papers by the Audio Engineering Society that deal with digital clipping explicitly:

Aliasing in Digital Clippers and Compressors
http://www.aes.org/e-lib/browse.cfm?elib=12036


Sorry, gotta run.

Sorry I was interrupted during my previous post, but to continue professional references and to devices that clip digitally, and practices that strive to avoid this result in audio and video workflow, (or to use it advantageously in some circumstances):

Here is an article that goes into significant detail about digital clipping:
http://www.tcelectronic.com/media/lund_2004_distortion_tmt20.pdf

It includes the following excerpts in a discussion of digital audio clipping:

Digital clipping can happen several places during
production: Inside an audio workstation (typically on
the mix-buss or in the plug-ins), deliberately in various
types of dynamics processors (yes, also inside a TC
Finalizer or other types of digital limiters), or even
inherited from digital sample libraries. Clipped kickdrums
or snares may be chosen for creative reasons, but
if the sample is brought close to full scale during mix or
mastering, unpredictable results will occur.

Clipping in the digital domain should be practiced with
caution - and for artistic reasons only by an engineer
that likes alias distortion and knows what she is doing.
As long as clipped audio stays clear of full scale in the
final mix by a substantial margin (at least 6 dB), it
won’t “explode” as an accident waiting to happen until
the CD meets a broadcast station or a consumer player.
Lower level clipped audio will stay the way it is heard
during the mix, and how it was meant to be, but how
does the engineer avoid making productions that
reproduce with unpredictable results?

Here is an article in IEEE Transactions on Broadcasting
http://ieeexplore.ieee.org/xpl/freeabs_all.jsp?arnumber=127422

Here is an EBU article that goes into detail about digital clipping:
http://www.ebu.ch/en/technical/trev/trev_310-lund.pdf


Here is a patent for a digital video clipping circuit:
http://www.freepatentsonline.com/5233410.html
An interpolating digital signal clipping circuit provides pre-clipping of the luminance component of a video signal.

Here is another for a "digital clipping circuit with soft clip characteristics":
http://www.freepatentsonline.com/EP0335317.html

Here is an article from the journal AES that discusses clipping in the digital domain:
http://www.aes.org/e-lib/browse.cfm?elib=6004



I think it's fairly clear that you are incorrect about your assertion that clipping can't happen in the digital domain. It happens in the digital domain regularly with audio and video signals, and is a common problem. And as discussed in the above articles, it can sometimes be utilized to positive effect. But the basic point is that clipping most certainly does happen in the digital domain.

That's why video engineers use the term "clipping" to describe the nature of what happens when you exceed a display's maximum with an input signal, or adjust the display's circuitry in such a way that the input signal exceeds the display's adjusted-maximum. It is a classic and textbook example of digital clipping.

That's why it is used to locate and align a digital display's black and white points to the content properly so that the entire video range is properly represented on the display without any of the video signal's digital excursion clipped off. That the excursion in the digital domain is represented numerically rather than by analog signal voltage is immaterial to the fact that it is, indeed, still a signal, and that it is, indeed, still being clipped off in these circumstances.

Except signal excursion has nothing to do with clipping in the digital domain. If the signal starts out containing the values 0-255, it will have the values 0-255 right up to the instant a signal is applied to a pixel element. If you reduce the "signal excursion" of a digital signal at some point you lose the entire sigal, you don't lose high numbers first or low numbers first. In a digital signal, all numbers are "equal" they are merely encoded differently so they can be extracted accurately.

That's simply false. I can't tell you how many devices I've run across that clip to 16-235. Values below 16 are clipped off. Values above 235 are clipped off. There are other devices that do other things as well. And any digital display with enough adjustment range can be adjusted (improperly) such that a portion of the video range is clipped off because it exceed's the displays as-adjusted range.

Your explanation here flies in the face of all digital signal theory, and betrays your own ignrance on the topic. You seem to be obsessed with the signaling down the wire, rather than the interface levels.

Again, I refer you back to Poynton's book, and the quotes I already provided which discuss digital video signal clipping. I will provide further passages that discuss the clipping I have explained here later if you'd like. Online he does have his color and gamma FAQs, which include for instance this small discussion (which is expanded in his book) that discuss head and toeroom clipping of digital video:

http://www.poynton.com/notes/colour_and_gamma/ColorFAQ.html
The matrix of Eq 4 will decode standard Y'CBCR components to RGB components in the range [0..255], subject to roundoff error. You must take care to avoid overflow due to roundoff error. But you must protect against overflow in any case, because studio video signals use the extremes of the coding range to handle signal overshoot and undershoot, and these will require clipping when decoded to an RGB range that has no headroom or footroom.

So I do not understand how you can claim that clipping does not occur in the digital domain. It unquestionably does. It's not a semantic quibble about whether the term clipping properly applies to some particular example that is similar but maybe not quite the same as clipping. One can say factually that clipping absolutely DOES occur in the digital domain. That is a fact. Pure and simple.

If you do not understand that basic fact, then I'm wasting my time discussing the nature of clipping in digital displays, since obviously if you're of the belief that clipping can't happen in digital displays, then there is no discussion possible if one participant is making completely fallacious claims not based in reality.

The issue in digital displays is after the numbers are extracted, you will still have a range of numbers from 0-255. Something has to control the video display. Typically this is a lookup table that essentially says "for the digital value '255' send this voltage to this pixel" usually that value will make the pixel 100% transparent which is what you'd want for 255. Digital 254 should send a voltage that makes the pixel just a little less transparent (or flash the pixel less frequently). The Contrast control can change the top end of the lookup table, the brightness control can control the bottom end of the lookup table. In a properly designed display, nothing inappropriate will happen as you adjust contrast and brightness.

I do not know what you mean by "nothing inappropriate will happen." You've made this claim several times in relation to "a properly designed display." You've not explained what you mean by this.

What happens depends on the adjustment range and the input signal. Obviously if the input signal has a limited range, or the adjustment range is limited such that you cannot ever create a situation where you are exceeding the display's capabilities to properly recreate that signal range, then you won't be able to clip anything, and maximum settings on the display's adjustments may be appropriate in order to maximize the available contrast range.

But such a design is unuseful. The entire reason that adjustments exist on displays is to accomodate different video standards with different ranges of values, and this also includes digital video.

This means that in order to be at all useful to anybody, the adjustment range must be significant enough to incorporate these various ranges. It necessarily follows that in order to accomodate a smaller input range, the display's adjustment range must be adjustable such that it could accomodate that smaller range. This means that in that position, which is proper and correct for the smaller range, it would be clipping off excursion beyond that for a larger input range.

In a poorly designed display, the lookup table could end up telling the video display that for digital values 200-255 send them all the same voltage (or frequency) to make them 100% transparent.

I would say that the display is properly designed, because if it cannot be adjusted that way in some situations, that means it doesn't have an adjustment range that is useful to anybody. What is happening in this instance however is user error in misadjusting the display. The result is a clipped image.

For example, the two dominant digital image standards that we run into every day is digital video (as per Rec 601 or Rec 709 for example) or graphics (as per sRGB). The former uses a 16-235 nominal excursion range. The latter uses a 0-255 excursion range. (both for 8-bit digital content, obviously the range is different for greater bit-depths). To accomodate both legitimate ranges, a display must be adjustable for black and white level. Ignoring for a moment some other concerns about peak white for the sake of example and simplicity, let's say that we adjust a display for a digital video input. We align the display's peak white point to correspond with digital level 235. We align the display's black point to correspond with digital level 16. These levels represent the nominal white and black points for standard-adherent digital video.

After doing so, we have properly aligned the display to this input range, which is appropriate for digital video content. Do you agree?

Continuing, lets replace that source with a graphics source that follows the sRGB standard. This standard defines black at digital level 0, and white at digital level 255. We leave the display unchanged.

Now, we have a display that is no longer properly aligned. It outputs it's maximum white for level 235, yet there is input signal excursion beyond that. In other words, our whites have been clipped off by this misalignment.

In the same way, our black have also been clipped off below 16. There is signal excursion below 16 in the input, that has been clipped off by this misalignment between display and source. This is not poor display design, but rather user error. The display is properly designed because this adjustment position is CORRECT for a video source, but WRONG for a graphics source. The ability to accomodate these various ranges should be present on any display, otherwise it limits the display's usability to only certain limited circumstances. The cause of this clipping is user error since the operator failed to re-adjust the display to the new input range.

I find it strange that you seem to be implying that a display with the capability to accomodate various input standards is "poorly designed." Every professional digital display and monitor I've worked with allows for exactly the type of adjustment I just described, and usually significanly more than that. Indeed, that is why nearly every display on earth includes black level and white level adjustments (brightness/contrast, brightness/picture, or whatever the labels may be). I simply cannot understand how it is you can claim that these displays are "poorly designed."

I will try a simple example... digital audio... there is an effect a lot of people call "clipping" in digital audio. If you playback a CD, for example, and capture the analog output of the CD player and display it in real time on a 'scope, many discs will appear to have the highest peaks "clipped". This is encoded on the CD. It is not wrong, the CD has accurately captured the digital master. The problem exists in the digital master. What is the problem? Is it clipping? I say, no, it is NOT clipping. It is merely a mistake. The engineer set the level of the analog input signal too high for the dynamic range of the digital audio system he was working in (16 bits in the case of CD audio). NOTHING clipped in the digital domain, it just ran out of bits because of human error on the part of the engineer. Everything worked exactly as it should work. The only problem was that the engineer set the analog input level too high and this produced something sort of like clipping, but there was no clipping of ANYTHING in the digital domain, and the analog input signal as it existed on tape was not clipped. In effect, the engineer setup a poorly designed lookup table for the analog voltage to digital bits conversion table.


And that's called clipping. I mean, that is the definition of clipping. It's what was described in the above journal articles that I referenced. It is exactly as described by Poynton's definition of clipping:

Clip (v) Forcing a signal to a certain maximum(or minimum) level, so as to avoid excursion above (or below) that level.

That's exactly what the engineer did wrong. He tried to force the digital range beyond the hard limits. This caused portions of the input signal to be clipped off above that level. That is exactly what clipping is, and in this case it occured by wrongly driving the ADC.

I mean, you literally are writing out an identity. You give an example where you describe a signal being clipped or limited, and then you turn around and try to say that it isn't being clipped or limited!

If that is not clipping, then WHAT IS CLIPPING? What is your personal definition of clipping? Do you have any professional references to back this definition?

What in Poynton's definition which I have repeated time and again, is wrong?
What is wrong about Joe Kane's description of digital clipping?
What is wrong about Guy Kuo's description of digital clipping?
What is wrong about the various journal articles which explain digital clipping?

You've not addressed any of these examples I've provided. All you're saying is "gee no way that's not clipping." Or "gee no way, digital can't clip at all." These are completely ridiculous assertions that have no basis in reality. Not only is it plainly false by anyone's reading of the english language, but it flies plainly in the face of the most respected video authorities around. It flies in plainly in the face of standard audio engineering practice and terminology.

Sorry, this is also technically incorrect. "The video signal" in digital displays is NEVER clipped (except in some plasma panels addressed elsewhere). Certainly you see luminance compression and unbalanced color - but it does not come from any SIGNAL being clipped (except the plasma case again). Period. It comes from the display doing exactly what it has been designed to do... good or bad. No values are lost, none are "missing" - and every pixel is doing exactly what has been encoded in the lookup table, right or wrong. Poor design that allows luminance compression and color shifts as you change the contrast control is simply poor design.

You're simply wrong on each of these counts. It's not poor display design, it's user error in misadjusting the display. That causes clipping of values which are lost in the reproduction of the image due to excursion beyond a limit.

There is no clipping of any electrical signal involved.

Well duh, it's digital. It's digital clipping of digital values. We haven't been talking analog at all.

Again, you can declare you're going to call the look of those missing steps or color shifts Clipping, but there's no SIGNAL (digital electroic) being clipped anywhere.

Um, the digital signal is there. And it's being clipped. If you have no signal, then you're right, there is no clipping, but then there is no image either. I'm assuming you're intelligent enough to understand we are talking about situations that actually have sources with signals hooked up to a display. As such, there is a signal there, and in this context we're talking about a digital signal. And in this case the video range that the signal conveys is being clipped off. This does not occur during transmission obviously, but during interface or processing operations.

No value between 0-255 in these digital displays is EVER eliminated at the electrical/electronic level. Severe adjustments of user controls can, as I've said a NUMBER of times (are you reading this?) cause poorly designed controls to assign lookup table values that produce luminance compression or complete loss of steps so the images have an APPEARANCE that there has been clipping, but if you look inside the display to see what really happend... there was no loss of ANY digital value from 0-255, the display simply provided poorly conceived controls that use all the numbers, but assign inappropriate levels to some of the numbers. All the numbers are there, all the pixels are doing exactly what they are being told to do by the display.

So what about devices that clip the range numerically? What about display inputs that do this. Are you aware that there are many devices out there that do this? That's called clipping.

There are many devices out there, unfortunately that clip to the 16-235 range. Values outside this range, for instance the sRGB range, are simply clipped off. Is that not clipping? And how is that any different than a display that you have misadjusted which does essentially the same thing through adjustment error?

Going back to the example of the car

Let's not, it's a terrible analogy.

I never said digital displays never clip

Really? You repeatedly made the claim that you can never clip in the digital domain.

I found the following statement made by you to be fairly absolute in that regard:

More than once, I tried to explain to you that in the digital domain, THERE IS NO SIGNAL CLIPPING. PERIOD. EVER.

Or

In our case, the numbers are 0-255. None of them are EVER clipped when they exist as digital signals.

Or

Actually, the signal in a digital video display NEVER clips. Ever.


Or

In a digital world, the signal never clips.

If a digital display is not operating at least in part in the digital domain, then it isn't a digital display is it?

I don't know how much clearer your own statements can be.

And I don't know how many more examples I can provide which illustrate the fallacious nature of these claims.

There absolutely is clipping in the digital domain. If you are not even aware of that, then moving onto more complex discussions about what's happening in displays and whether or not it is clipping is futile since you're not in a position of understanding to have that discussion.

If you would perhaps stop being so blindly stubborn, you could backtrack a moment and learn a little bit about the digital domain. It would be unfortunate if you allowed ego to obscure possibilities at learning from the numerous sources I have cited for you. You can ignore my assertions entirely for all I care, but I would hope that you would at the very least take heed of professional sources.

But if you use the model of clipping being the intentional or unintentional peak-limiting of an electronic SIGNAL... clipping never happens in digital displays.

Except that it does. All the time. On $30K displays. On $100K displays. On $200K displays. On ever display I've ever seen with enough adjustment range.

When we see problems, we are not seeing a clipped elctronic signal, we are seeing and measuring is the result of poor programming or design.

We are seeing a display that has been driven into clipping by operator error/misadjustment.

I have no problem with displays like the Westinghouse LCDs having luminance compression problems that you and everybody else want to call clipping.

The probelm is that the "everybody else" in that statement is the entire video engineering community, and indeed the entire engineering community at large. You can do the same thing with digital audio in the digital domain with similar controls, you can drive digital circuits into clipping. And it's called clipping. Why do engineers call it clipping? Because that is the DEFINITION of clipping.

What you are seeing on the display is a result of lower cost optics or light sources and using software to try to compensate.

What about the $60K DPI projector I'm sitting next to? That clips too if you misadjust it.

What about the $200K JVC 4k projector I watched last week? That clips too if you misadjust it.

Find me a single, ANY digital display on planet earth that doesn't clip if it has sufficient adjustment range. With ANY light source.

You can't do it. Not possible.


There's nothing made-up in my definition.

There absolutely is. I have more than a half-dozen times provided you with the definition of clipping.

You have not once explained your own definition, yet it appears to fly in the face of every professional source that I could find. You have provided NO reference for your definition whatsoever. If you did so, at least I could understand how it is you arrived at your conception of the term. But you haven't. You've not cited ANYTHING in ANY of your posts. EVER. Not ONCE!

If you understood digital imaging systems, that would be clear to you. The calibration community has selected a word of convenience.

No, the calibration community has continued to use the term clipping in the same way as it is used in the professional community as illustrated by the various articles I linked you to above.

It is using the term clipping in the same way as Charles Poynton.

But again, perhaps you think Charles Poynton simply included a popular, convenient, and misleading definition of a term in his glossary.

The same Charles Poynton who includes an entire section in his book dedicated to clearing up confusion between the term 'luma' and the term 'luminance.' Yet that same author who is so concerned with precision of terminology, decided to include a completely non-technical, popular, and convenient definition of 'clipping' because I guess I was just lazy?

If you want to have 2 completely different defintions of the words clip/clipping you have to recognize the difference between the 2 usages.

What 2 different usages!? I have stuck with one, singlar definition the entire time, from the very beginning. I am using it one, coherent, and intelligible way, with professional references.

You have forwarded no professional references for your unclear and confused definition of clipping.

You have made the completley errant claim that nothing can clip in the digital domain. That has been thoroughly refuted.

And fundamentally, that is really the problem we're having in this discussion, is that you have a limited understanding of the digital domain, thus your bizarre contention that you can't clip digitally. Well, you can. Once you get over yourself and behave with a little bit of intellectual rigor, we can have a useful discussion. But proceeding to discuss digital displays when you are very weakly informed about the digital domain in digital is not very productive for anyone, particularly when you are content to be so stubborn about your unawareness.

I mean, after 34 years in video, have you ever even READ Poynton's book? Obviously not.

If most people outgrow your reasoning in grade school, why do you keep subjecting me to it?
+1

Doug,

FWIW Chris is unable to comprehend that solid state light emitting devices are ANALOG and the “voltage” driving these components are produced by a given LUT per D/A. ALL 255 (8-bit) discrete analog amplitude values are uniquely represented, so whether or not the light emitting device is able to ideally track all 100% quantized levels is another matter.

By definition LEDs have a non-linear light transfer response characteristic (lazy S-shape), so again you are correct in that COMPRESSION would be the more technically correct adjective.

Maybe Chris will be kind enough to list all the DIGITAL display panels that are not driven by ANALOG VOLTAGE:p

[QUOTE=tbrunet;14165849...the “voltage” driving these components are produced by a given LUT per D/A. [/QUOTE]


Really?

Sorry, please disregard thomas. No, it's not really.

The LUT is not the DAC. The LUT is for de-gamma.

And all DLPs are not driven analog, obviously. Nor are plasmas.

{a bunch of repetitive stuff that does nothing to advance any discussion}


No original thought or room to explore an "heretical" concept is permitted in your world. If it doesn't follow your rules and your quotes, it's wrong. And that's all there is to it.

When I refuse to give in to your brand of internet bullying, calling me igorant is your fall-back tactic. Even though I have been an AVS member for 8 years under a nickname, I have only recently chosen to use my real-world name. In my entire professional career, I have never been called igonrant by anybody and find your repeated attacks and insults to be offensive in the extreme. This started out as a discussion of a different way of looking at what happens when you see something inappropriate in a video display. You have devolved it into personal attacks and insults which are prohibited by forum rules. I've been polite and gentlemanly trying to offer an alternate view of what I believe is incorrect application of a concept. I have not complained to moderators about your offensive comments because I believe they reveal your character and have no impact on me directly even though, at the time, it was pretty clear you wished they did. I also don't believe in running to daddy for protection from the big bad forum basher.

If I was annoyed by any of your insults and managed to slip in an insult in one of my posts, I apologize. I don't think I did, and I didn't intend to.

There is no discussion going on here, so no need to continue your defense of the party line.

No, it's not really.Really?
Yep! Apparently you and Chris failed to comprehend Doug's explanation regarding fundamental design architecture.
You could, in theory, hack the engine computer and make a new lookup table that would allow the throttle to achieve 100% open.:)

No original thought or room to explore an "heretical" concept is permitted in your world. If it doesn't follow your rules and your quotes, it's wrong. And that's all there is to it.

When I refuse to give in to your brand of internet bullying, calling me igorant is your fall-back tactic. Even though I have been an AVS member for 8 years under a nickname, I have only recently chosen to use my real-world name. In my entire professional career, I have never been called igonrant by anybody and find your repeated attacks and insults to be offensive in the extreme. This started out as a discussion of a different way of looking at what happens when you see something inappropriate in a video display. You have devolved it into personal attacks and insults which are prohibited by forum rules. I've been polite and gentlemanly trying to offer an alternate view of what I believe is incorrect application of a concept. I have not complained to moderators about your offensive comments because I believe they reveal your character and have no impact on me directly even though, at the time, it was pretty clear you wished they did. I also don't believe in running to daddy for protection from the big bad forum basher.

If I was annoyed by any of your insults and managed to slip in an insult in one of my posts, I apologize. I don't think I did, and I didn't intend to.

There is no discussion going on here, so no need to continue your defense of the party line.

In other words, you don't want to want to confront the numerous professional references I provided for you which directly refute your assertions. You don't want to admit that you were incorrect about digital clipping, or about clipping in digital display devices.

You didn't answer any of the questions I asked you.

You didn't explain or support any of your claims, which have no references or supporting sources for your terminology.

I fail to see how my patient and detailed responses to your posts in their entirety over a lengthy period of time, numerous professional references and my own additional explanations to you constitutes "internet bullying."

I'm sorry, but if somebody comes on the forum and says something that is patently false, someone correcting their misinformation is not "internet bullying."

Neither did I personally insult you. Calling you incorrect, wrongly informed, or ignorant on the topic is not an insult. I didn't call you names or debase your person. But if you make statements which are factually erroneous, you should not be offended in any way when someone corrects those errors of fact.

That is all you have done here, is forward some very serious errors of fact, and I've spent an inordinate amount of time with a great deal of largely unecessary patience trying to correct those errors of fact in a way that you can understand and be rightly convinced by the technical literature.

You have completely avoided these references, you have ignored the technical literature, and otherwise have behaved in a very unprofessional fashion. Instead of considering the outside sources I have referenced you to, you simply ignore them. And conversely, you provide no technical references to back your claims.

That you believe strongly that the things you've said here are true are immaterial. All I am asking of you is intellectual honesty and correcting misinformation when it is established and clear that you are mistaken about a claim. Honesty demands this behavior, and you are grossly failing in this respect.

I don't know how else to legitimize my claims in your eyes.

I have provided various technical references from various authorities. I have asked you to consult them on your own. I have asked you to provide your own references to support your claims, because it certainly could be that I am mistaken, and that in certain areas of use your definitions do have merit. I am not aware of them, and until you provide any evidence or support for this, I must go with the most legitimate technical references I am aware of, and their usage of technical terms.

What else could convince you besides my already thorough consultation of the technical literature?

Do you have a video reference better than Charles Poynton? If so, I would very much like to know that, because I am not aware of a more thorough text on digital video.

I've asked you repeatedly to provide any such references for your understanding, and I don't know how else to get you to respond. You have utterly failed to respond to any of these requests, which I feel are not unfair requests at all. I just ask you to legitimate your claims with scientific, technical references. I'm just asking you to behave with basic integrity.

You literally will not answer any of my questions, you ignore all the technical literature, and you provide no references of your own. That is not reasonable behavior, and it is a disservice both to yourself and to this forum. This is AVScience forum. I don't think it is unfair for me to ask basic intellectual rigor of you. I do not see why you should be insulted by that, since you are an engineer; a man of science.

And that is ALL that I am asking of you. Integrity. Can you do that, please?

Where are these DACs that you are talking about? I have not seen them in the circuits that drive panels. Can someone point to a schematic or block diagram that shows one? I was not aware that this was the case.

I suspect that the notion that panels are driven by analog signals comes from the fact that the waveforms that do get applied to panels may look like messy rf signals, and digital signals are not simply bi-state square waves. There is some effective low pass filtering and noise in the circuits, but not intentional, dedicated D/A conversion that I am aware of.

To everyone, rather than continuing the vitriole and demeaning each other, why not just post some documentation of what you have to say. Some of us are here to share information and learn rather than peck at each others eyes. I see lots of assumptions and opinions about not just the subject of the thread, but each others' intent and meaning. Just the facts, please.

LCD and LCOS display pixels are individually controlled by voltage, which control how translucent or opaque (or reflective in the case of LCOS) they are. So yes it does eventually go back to being driven by per-pixel "analog" voltage.

DLPs and plasmas do not because they are PWM/PDM.

To everyone, rather than continuing the vitriole and demeaning each other, why not just post some documentation of what you have to say.


http://www.pacificdisplay.com/lcd_backlights.htm
Controlling Backlight Brightness
One can also very the LED backlight brightness by simply varying the DC current to the LED's

NOTE that the backlight is NOT "driven" by binary or HEX values!

DLPs and plasmas do not because they are PWM/PDM.Still varying the on/off ratio of an ANALOG waveform;)

http://www.pacificdisplay.com/lcd_backlights.htm
Controlling Backlight Brightness
One can also very the LED backlight brightness by simply varying the DC current to the LED's

NOTE that the backlight is NOT "driven" by binary or HEX values!

Are we talking about backlighting or panel drive? I thought the discussion was about the signal to the panel. I still want to know where the DAC is that you talk about.

LCDs are transmissive and are controlled by ANALOG voltage. The backlight is simply the light supply that is varied by the (LCD valve) magnitude of the analog signal.

This analog vs digital discussion is rather pointless, IMO. The exact nature of how devices are addressed varies both within and between technologies. Unless you are discussing a specific device and understand the way in which it deals with the signal, you really are not providing any useful information. Some LCDs for instance are addressed with continous voltages, some with discrete levels. It depends on the interface IC used and the type of LCD. My point was that there is not a specific DAC stage in the output to the panels that I have seen, though the actual signals may resemble analog more than what we think of as digital. The difference between analog and digital becomes blurred when we get to high data rates and multiplexed signals. The technologies start to merge. LVDS signalling used to communicate to most panels these days is not an inherently digital technique. The signals sent across it are initiated as discrete levels in most cases, however.

Get the datasheets, schematics, and application engineering data for the devices that you want to discuss, then we can make sense of these debates. All this generalizing and assumption really gets nowhere.

I don't think it's really relevant to the thread anyway.

The difference between analog and digital becomes blurred when we get to high data rates and multiplexed signals. The technologies start to merge.There’s zero gray area about the subject matter. Maybe Chris can provide us with an example of ONE digital display panel technology that responds directly to binary or hex data?

Good luck with that in advance!

There’s zero gray area about the subject matter. Maybe Chris can provide us with an example of ONE digital display panel technology that responds directly to binary or hex data?

Good luck with that in advance!

Thomas, I have repeatedly asked you to stop trolling me.

That being said, I already told you that DLPs and plasmas are not analog, they are PWM/PDM. You are correct (which is amazing) about LCDs and LCOS panels which are controlled by DC voltage. Whether you consider that analog or not is not really a useful discussion, and off-topic. But DLPs and Plasmas are not controlled this way, obviously.

I already told you that DLPs and plasmas are not analog, they are PWM/PDM. Varying the ON/OFF duty cycle of a "analog waveform" does NOT qualify as digital binary or hex data;)

"Virtually all AMLCDs (Active Matrix LCDs)
are designed to produce gray levels—intermediate
brightness levels—between the brightest
white and the darkest black a unit pixel
can generate. There can be either a discrete
number of levels—such as 8, 16, 64,
or 256—or a continuous gradation of levels,
depending on the LCD Driver IC (LDI) used."

http://www.wde.com/pdf/wsr123_reprint.pdf
Widescreen Review • Issue 123 • September 2007

Varying the ON/OFF duty cycle of a "analog waveform" does NOT qualify as digital binary or hex data;)


Pulse Width Modulation is Pulse Width Modulation. To argue whether it is analog or digital is pointless. It can be used to convert digital to analog, but at the level of controlling a DMD, the input is clearly digital. At the level of controlling a PDP, you could make a case for the output stage acting as the DAC, but what is the point?

In any of these cases you can potentially run into the limits of the system at some point. What you choose to call analog or digital, just like what you chose to call clipping, does not inform anyone about the nature of the problem. That will be specific to the system. Without context, these debates really don't have much relevance.

So does anyone know what was going on specifically with the LN40A450 that the OP has and did he ever get his question answered adequately?

Greetings

The original poster had his answer by post #11.

Here it is at #84 ... sidetracked.

regards

Nothing was sidetracked.. thanks to Doug's expertise!
LCDs don't clip no matter how bright the backlight or Contrast control are set. So you won't get a color shift (but the color could have errors all the time.

Since we're on the topic;

Why do some test disks encode at 0 IRE and others at 7.5 IRE?


BTW: I'm just kidding :p

You didn't explain or support any of your claims, which have no references or supporting sources for your terminology.

When you are not in agreement with "the establishment", there are no references. That doesn't make a differing point of view wrong--this is proven daily as previously undocumented concepts are proven to be correct in all branches of science. I use my experience and analysis of the processes to reach my own conclusions. I almost always agree with published technical information. In this case, I don't accept the "established" use for reasons I've explained.


I fail to see how my patient and detailed responses... and my additional explanations constitutes "internet bullying."

It's not what you did, it's how you did it. The systematic intolerance in the replies comes across as pedantic, dogmatic, and insulting - whether that was your intenet or not, they do. You could simply have answered that you understand what I was saying (say, about a passive device like a light valve being incapable of clipping), but you choose to go with the definition as you understand it from other sources (which is fine with me, it's your choice). That would have been the end of everything. But you had to launch into multiple point-by-point attacks with destruction of a differing point of view as the sole goal. And in spite of your protests to the contrary, your tone and words ARE highly insulting of my training (which includes all the references you mentioned), experience, and accomplishements in the industry. You know nothing about me except that I don't think it's appropriate to call incorrect video effects from light valve displays clipping.


I'm sorry, but if somebody comes on the forum and says something that is patently false, someone correcting their misinformation is not "internet bullying."

It is bullying when it becomes a vendetta of endless personal attacks and an effort to stomp out an opposing point of view. You get so worked into a lather you can't help calling me ignorant (and that's exactly what you did) and you couldn't stop yourself from "feel sorry for your calibration customers" (may not be an exact quote, but it's pretty close) when you have no personal knowlege of my calibration expertise. The quality of caliibraion I perform is completely unrelated to whether incorrect video response from light valve devices is called clipping or not. The fact that it is cheked for and corrected is the only thing relevant to performing a quality calibration, not what name I choose to assign to it.


Neither did I personally insult you.

Yes... you did.


Calling you incorrect, wrongly informed, or ignorant on the topic is not an insult.

Why do you feel compelled to call me incorrect, wrongly informed, or ignorant of the topic AT ALL??? It's a character thing. You can make your point without doing any of those things. That's called being balanced and diplomatic. Going your way is bullying and offensive in the extreme - and that's why you have at least 1 guy who antagonizes you - it's your approach that is so offensive it pisses people off. You clearly have a lot of knowledge, but your approach to using it in this thread is as I've already described, no need to repeat my opinions here. Your approach in this thread has been the problem, not the fact that you think I'm wrong. Yet, you did a great job with a post in another thread. Compltely different approach, completely different result. If this exchange had gone the same way as that other post, I wouldn't have any problems with our disagreement.


I didn't call you names or debase your person.

Yes... you did. Including comments that call my ability to perform a quality video calibration into question.


But if you make statements which are factually erroneous, you should not be offended in any way when someone corrects those errors of fact.

As mentioned already - if you had made a simple statement of disagreement based on common usage of the term, it would have all been over in a post or 2. But you had to enter into a obliteration mode because I have a different way of looking at the issue compared to published definitions. There was nothing factually erroneous in my assertions. It is clear from my position that, when it comes to visble effects from passive light valve devices, that I don't agree with calling that clipping. That's all I'm saying. There is nothing factually incorrect in how I describe how the undesired effects are produced - the only thing I don't agree with is calling it clipping. There is no need to attempt to destroy a person's point of view simply because the person doesn't agree with a definition in general use. I never said there weren't "problems", only that I didn't think it was right to call it clipping in the case of light valve devices in video displays.


That is all you have done here, is forward some very serious errors of fact, and I've spent an inordinate amount of time with a great deal of largely unecessary patience trying to correct those errors of fact in a way that you can understand and be rightly convinced by the technical literature.

No, that's merely your interpretation - I have a problem with terminology. Not with any underlying science, engineering, or design issues. This is a personal thing it is not worth the time you have invested in it. It is worth the time I have invested in it because it is my (differing) point of view. I pursued the explanation only because you failed to acknowledge you "got" what I was saying - as already stated, if you'd simply said something like "I understand how you could reach that conclusion, but I don't agree with it, nor do the published references I can find." It all would have been over. But you had to go into seek-and-destroy mode instead. Crush the infidel for having an opposing point of view. It's simply a disagreement over what is called clipping and what should be called something else (in my opinion).


You have completely avoided these references, you have ignored the technical literature, and otherwise have behaved in a very unprofessional fashion. Instead of considering the outside sources I have referenced you to, you simply ignore them. And conversely, you provide no technical references to back your claims.

This is exactly the sort of offensive dogma you pepper your replies with. Of course I ignored them. I don't agree with them on how clipping is defined in passive light valve devices.


That you believe strongly that the things you've said here are true are immaterial.

This is your dogma speaking again - no room for anything that might be the slighest bit different. If this different opinion I hold would change or affect anything, I could see getting uppity about it... maybe. But in reality it is a small detail that affects NOTHING. NOTHING AT ALL. Except the name I prefer to call something.


All I am asking of you is intellectual honesty and correcting misinformation when it is established and clear that you are mistaken about a claim. Honesty demands this behavior, and you are grossly failing in this respect.

My "intellectual honesty" comes from being willing to admit in a public forum that I don't agree with what is referred to as clipping as applied to light valve video devices. Your only objection is that my point of view doesn't match what is published in your treasured references. Well whoop tee dooo. Is that really worth the effort you have put into trying to destroy me for having a different point of view?


I don't know how else to legitimize my claims in your eyes.

You can't legitimize your claims in my eyes. There is NOTHING you can say or do that will change my assessment of what is or isn't clipping in a light valve video display product. I don't CARE how legitimate your "claims" are because I know that my definition is outside the "establishment". But you want to crush me because of that. It's not going to happen. Why don't you try to get me to convert to some other religion or political party. I figure that will burn up another 8 or 10 hours out of every one of your days. Make me your "project". Fix me. Bend me to your will. Yes, bully me. Go ahead. I haven't had enough yet.


I have provided various technical references from various authorities. I have asked you to consult them on your own. I have asked you to provide your own references to support your claims, because it certainly could be that I am mistaken, and that in certain areas of use your definitions do have merit. I am not aware of them, and until you provide any evidence or support for this, I must go with the most legitimate technical references I am aware of, and their usage of technical terms.

Yet more evidence of your dogmatic need to CRUSH me into insignificance. I don't need references, none of them will change my mind. I've been there and done that. In fact, I'm beyond that... I'm actually thinking for myself. That's a dangerous concept for your tidy little scientific world. It's a good thing that science encourages dissenting points of view or we'd never make progress - we'd still think that blood-letting was a good thing and that any speed over 40mph would kill a human beng. You're just going to have to learn to live with the fact that I am never going to agree with you or your precious references on this point. That will probably be the thought that launches your next tantrum.


What else could convince you besides my already thorough consultation of the technical literature?

Nothing. But keep up your thorough consultation of the technical literature.


Do you have a video reference better than Charles Poynton? If so, I would very much like to know that, because I am not aware of a more thorough text on digital video.

For this matter? Me. I'm the reference. Quote me. "Light valve devices in video displays do not clip in the digital domain, but they can produce visible effects many people describe as clipping." That's my quote. Is it world alterning? Does it change the way ANY video display works? Is it worth all the effort you have put into squashing it to dust?


I've asked you repeatedly to provide any such references for your understanding, and I don't know how else to get you to respond.

Because you are missing the entire point - completely. There's nothing wrong with having an issue with a definition. It's not like I'm saying Ohm's Law is wrong or that Red, Green, and Blue aren't really the primary colors.


You have utterly failed to respond to any of these requests, which I feel are not unfair requests at all.

They aren't unfair requests. They are ignorant requests. Just quoting you. Are you OK with that?


I just ask you to legitimate your claims with scientific, technical references. I'm just asking you to behave with basic integrity.

I am the techinical reference. I hearby legitimize (not my "claim" but my opinion) that light valve devices do not clip in the digital domain. Now drop and give me 100 more references that prove how wrong I am.


You literally will not answer any of my questions, you ignore all the technical literature, and you provide no references of your own. That is not reasonable behavior,

It may not be reasonable behaviour in your uptight little world, but it works fine for me. Your questions don't matter in the least. All that matters is why I think light valve displays don't clip - and once again, I am quite willing to admit they may, under some conditions, produce effects that people are calling clipping.


and it is a disservice both to yourself and to this forum.

Nobody is getting worked up about this except you. Forums do allow for opinions to exist, don't they? There are no thought police running around making sure everybody has the same opinion are there? Oh, maybe I asked the wrong person.


This is AV[I]Science forum. I don't think it is unfair for me to ask basic intellectual rigor of you. I do not see why you should be insulted by that, since you are an engineer; a man of science.

Part of intellectual rigor is considering alternative findings and alternative points of view. You are free to accept or reject those alternatives, but you can't and shouldn't stomp different points of view out of existence - unless it's cold fusion as promoted many years ago.


And that is ALL that I am asking of you. Integrity. Can you do that, please?

Your agenda is clearly "defense of territory at all costs" which has nothing to do with integrity. My agenda is simply thinking of something in a different way. There's nothing you've said or quoted I don't understand. The whole problem that is getting you so wound up is that I won't agree with your "science". All the posts, all the bullying, all the indignation, it's all because my assessment of what happens in light valve displays isn't something I think should be called clipping. If you had any integrity, you'd stop with the non-stop attacks and let my opinion stand as it is. If you have no integrity, there will be yet another round of attacks that will fail to do anything to get me to change my mind.

If you wanted the term changed, wouldn't a review panel with your peers be a better place to change the "established view"?

When you come on a public forum and saying what something is, when it is NOT (your own admittance) then there is a high potential for confusion among the "not as elite" (definetely myself included).

Let's say for instance I decided that R, G, and B were NOT the primary colors. Would I:

A) Go on a public forum and claim I know the true primary colors and they are what I termed, Bred, Plurple, and Gorange?

or

B) Discuss it with the scientific comunity, who could actually review my thesis, and in turn (when proven correct) change the nomenclature?

It's fine to have a differing opinion, it's when you spew it as fact to the people who don't know what your talking about that there is a problem.

If you don't like the term, then change it (via some other higher means than a AV scientific forum).

I am one of the (I'm sure) many people who don't understand what your discussing. So how am I to know the correct terminology, without the added confusion?

Again, your opinion on clipping for digital devices is fine, hell, it may even be the correct way of thinking that no one else but you discovered yet.

But until you change it with someone that can do something about it, don't color the science and established facts that exist, it's REALLY confusing.

As is always said; "just my 2 cents",

-Brian

When you are not in agreement with "the establishment", there are no references. That doesn't make a differing point of view wrong--this is proven daily as previously undocumented concepts are proven to be correct in all branches of science. I use my experience and analysis of the processes to reach my own conclusions. I almost always agree with published technical information. In this case, I don't accept the "established" use for reasons I've explained.

That's a specious argument. In fact it's totally ridiculous.

The term clipping was invented and defined to describe certain kinds of effects that occur in signals. It was established and used by engineers in a clear an well-understood way. It still is used in that same way.

Inventing your own definition of the term, and then asserting that as fact, and indeed making numerous claims that are simply false is what makes your position wrong. It also makes it extremely unuseful because you inject confusion into discussion that otherwise would be very clear.


It's not what you did, it's how you did it. The systematic intolerance in the replies comes across as pedantic, dogmatic, and insulting - whether that was your intenet or not, they do. You could simply have answered that you understand what I was saying (say, about a passive device like a light valve being incapable of clipping), but you choose to go with the definition as you understand it from other sources (which is fine with me, it's your choice).

Again, I have a forceful argumentative style, and when someone is making clearly false claims that grossly misinform readers, I tend to rebut these wrong claims so that others reading will not be likewise wrongly informed.

And if you notice in this thread, you began with the completely bizarre claim that:

Most modern digital displays can't have the white level set with ANY test pattern because they show every shade of white even beyond 100% white which is represented by a digital 235 for all 3 colors (red, green, and blue).

LCDs don't clip no matter how bright the backlight or Contrast control are set. So you won't get a color shift (but the color could have errors all the time.

To which I responded in surprise with: "huh? Sure they do."

And Michael responded with the same degree of surprise.

There is no way that any part of that assertion is correct. Thus my surprise.

When you then launched into a bizarre and unfounded explanation of these assumptions of yours which are obviously wrong in both practice and theory, I made a fairly brief post that corrected these. And I've been saying the same things in as many different ways as I can to you since then. And that did include a great deal of consternation and surprise that you held so strongly such bizarre and so obviously incorrect views.

I mean, the claim that digital displays can't have their white level set with ANY test pattern is absolutely bizarre.

That would have been the end of everything. But you had to launch into multiple point-by-point attacks with destruction of a differing point of view as the sole goal. And in spite of your protests to the contrary, your tone and words ARE highly insulting of my training (which includes all the references you mentioned), experience, and accomplishements in the industry. You know nothing about me except that I don't think it's appropriate to call incorrect video effects from light valve displays clipping.


Well, I do know that you believe, as your statements above have illustrated, that it is not possible to align the white or black points of digital displays since you do not believe they can ever clip.

Well, they do clip as has been established beyond any doubt or confusion. And that's what one observes while calibrating to locate the display's maximum white point, and also the display's black point. And indeed, they hard clip in a way that is far more unambiguous than the rather soft margins on CRTs, for example. You can see it on ANY digital display with sufficient adjustmenr range, which is darn near all of them. That includes cheap and poorly designed displays, and professional displays. And all displays in between.


It is bullying when it becomes a vendetta of endless personal attacks and an effort to stomp out an opposing point of view. You get so worked into a lather you can't help calling me ignorant (and that's exactly what you did) and you couldn't stop yourself from "feel[ing] sorry for your calibration customers" (may not be an exact quote, but it's pretty close) when you have no personal knowlege of my calibration expertise.

Ignorance is not a personal insult. Ignorance is a state of unknowing on a topic. I didn't insult your intelligence which is obviously not lacking. No matter how smart a person is, their areas of knowledge are limited. Every person is ignorant about things. In this case, you are not very well informed about digital video, and digital displays. I mean, it follows that if you do not believe that digital displays have any maximum white point since they can never clip, then clearly you cannot properly calibrate such a display since locating that white point is crucial to properly setting white level for the highest quality image.

This is why I referenced Guy Kuo and Joe Kane's explanations of the methodology to locate this point on digital displays while calibrating, in the hopes that you could eliminate this area of unknowing for you. It's why I've taken so much patience (which did fray at certain points, thus some of my frustrated statements) with you because you should know about digital video and how these displays behave.

Now if it's simply the word "ignorant" that is offensive, I do not mean it that way, and there isn't any connotation in that word that should be taken as insulting. I, like anyone else, is ignorant about far more things than I know.

The quality of caliibraion I perform is completely unrelated to whether incorrect video response from light valve devices is called clipping or not. The fact that it is cheked for and corrected is the only thing relevant to performing a quality calibration, not what name I choose to assign to it.

But what confuses me is the above claims that there simply is no clip point on digital displays, and that you can just set them to their maximum. Well, that's simply not so. Doing that grossly degrades the image. Why? Because the image is usually very heavily clipped in such a setting.

Your unfounded belief that they won't ever clip no matter how high the contrast setting is is a very bad understanding to have because it leads to a horrible image if you set a display that way. I frankly wouldn't care at all if you had a misunderstanding that lead to no damage or no effect in what other people might draw from it. If you called it, for instance, "smashing" instead of "clipping" and made your own distinction of "smashing" in the digital domain but were perfectly able to identify the "smash" point on a digital display then it really doesn't matter to me or anyone else. I think it would be silly, but you wouldn't be doing any damage to the image quality of your own displays, that of your clients, or to others who may read your posts and follow your advice.

Why do you feel compelled to call me incorrect, wrongly informed, or ignorant of the topic AT ALL??? It's a character thing. You can make your point without doing any of those things.

If someone is wrong, you generally have to say they are in some way mistaken. None of the terms you list is at all insulting, nor should they be.

If I said that and LCD display uses one LCD panel that changes colors to create a color image, you would probably say: "No, you're wrong about that. There are almost always 3 colored panels which combine to make a colored image." And you might expplain further, or I might go look for more information, and after reading and realizing "gee, I was wrong about that." I would admit that, and then I wouldn't keep explaining something that's totally erroneous.

That's called being balanced and diplomatic. Going your way is bullying and offensive in the extreme - and that's why you have at least 1 guy who antagonizes you - it's your approach that is so offensive it pisses people off.

Thomas is another story, and he has no place in this thread. His posting history on AVS is long and sordid, and his many claims speak for themselves.

You clearly have a lot of knowledge, but your approach to using it in this thread is as I've already described, no need to repeat my opinions here. Your approach in this thread has been the problem, not the fact that you think I'm wrong. Yet, you did a great job with a post in another thread. Compltely different approach, completely different result. If this exchange had gone the same way as that other post, I wouldn't have any problems with our disagreement.

I don't have any problems with our disagreement either, and I don't see why you should be insulted. I made my argument more strongly only after you continued to assert things that were clearly incorrect. If you observe the original posts in this thread you will see that.

I made a fairly short and basic normal post in answer to the original poster that described the clip point in order to locate the white point of the display in order to assist him in calibrating it. You made a claim that such a point doesn't exist and that digital display's won't ever clip. I reacted briefly in surprise, as did Michael. It was a strange assertion you made, and I didn't even spend time to refute it because it was so far out of left field.

Only later did I go into further detail in response to more posts from you asserting the same thing. And I made several detailed posts before I got so frustrated that I got more explicit in my arguments.


Yes... you did. Including comments that call my ability to perform a quality video calibration into question.

Well it follows logically. Whether I point out the obvious or not, anyone reading would go: "huh?"

In fact, that's what I did in the first response. "huh?"

And I'm still doing that now. But now it's due to surprise at your unwillingness to admit your error or the limits of your understanding here.

As mentioned already - if you had made a simple statement of disagreement based on common usage of the term, it would have all been over in a post or 2. But you had to enter into a obliteration mode because I have a different way of looking at the issue compared to published definitions.

Well, actually if you look back to the posts in this thread, there were several posts that were completely neutral in which all I did was explain briefly, and then in significanly more detail, and then again in even more detail, what you should already have known or at least should recognize that you didn't know and then gone and educated yourself about.

But you didn't do that.

There was nothing factually erroneous in my assertions. It is clear from my position that, when it comes to visble effects from passive light valve devices, that I don't agree with calling that clipping. That's all I'm saying. There is nothing factually incorrect in how I describe how the undesired effects are produced - the only thing I don't agree with is calling it clipping. There is no need to attempt to destroy a person's point of view simply because the person doesn't agree with a definition in general use. I never said there weren't "problems", only that I didn't think it was right to call it clipping in the case of light valve devices in video displays.

You said a lot more than that.

You said that anything in the digital domain cannot clip. (which is wrong)

You said that you can crank the contrast level up to the maximum without any clipping or colorshifint. (which is wrong)

You said that "Most modern digital displays can't have the white level set with ANY test pattern because they show every shade of white even beyond 100% white" (which is wrong)



My "intellectual honesty" comes from being willing to admit in a public forum that I don't agree with what is referred to as clipping as applied to light valve video devices. Your only objection is that my point of view doesn't match what is published in your treasured references. Well whoop tee dooo. Is that really worth the effort you have put into trying to destroy me for having a different point of view?

Destroy you? In what way? By stating the obvious that you are mistaken about a few thing that you've said here? The one digging the hole here is you. I mean, making up definitions on your own and using yourself as a reference is pretty hilarious.


You can't legitimize your claims in my eyes. There is NOTHING you can say or do that will change my assessment of what is or isn't clipping in a light valve video display product. I don't CARE how legitimate your "claims" are because I know that my definition is outside the "establishment".

Well, thank you for at least admitting that you are beyond reasonable behavior here. No professional or technical reference will dissuade you from your beliefs. No evidence or examples will shake your own personal thinking. No amount of science will shake your unscientific inventiveness.

I don't need references, none of them will change my mind.

Fantastic. We need more of this in the world. :rolleyes:

Me. I'm the reference. Quote me.

Great. Today founds the beginning of Doug Blackburn's personal reality. Doug Blackburn is the new authority on everything video. We will all heed Doug's expertise, and his powerful creativity to make up new terms and concepts as he sees fit. From this point on we should all ignore everything we have learned or read before this day, and celebrate a new reality as elucidated by Doug Blackburn. A new math. A new science. A new physics. An entirely new way of knowing the world.

Should I or others encounter your posts in the future, we should take care to recognize that your claims are based totally on your own creative ideas about engineering and science, and are a unique and separated reality apart from our own common reality.

Because you are missing the entire point - completely. There's nothing wrong with having an issue with a definition. It's not like I'm saying Ohm's Law is wrong or that Red, Green, and Blue aren't really the primary colors.

It isn't? Didn't you just go through how now amount of technical literature or objective fact will dissuade you from your beliefs? Why not just make up a new ohms law too while you're at it. Make resistance something that increases current flow instead. I mean, as long as you're just making things up and admitting that you don't care what any objective reference or logic or reason says.

Why not put Purple as a color in the rainbow too?

I mean, if you're going to create your own alternate reality, at the very least go all the way with it. I mean, I could respect that if you created your own complete way of knowing.


I am the techinical reference. I hearby legitimize (not my "claim" but my opinion) that light valve devices do not clip in the digital domain.

Go for it! Doug Blackburn's new science. A new objective reality.

I'm very interested now!

All that matters is why I think...

:rolleyes:

I think you had 1 or 2 good points there Chris :D its all about keeping an open mind "Verses" A closed mind .....

While I get the point of a good technical discussion, the stakes just aren't high enough warrant this much emotion. One discussion in the AVS forum is not going to change the nomenclature everyone uses to describe this condition. Call it clipping. Call it a "run out". As long as we all understand the condition we mean we can have productive conversations here.

Me. I'm the reference. Quote me.

Anybody would have to admit that this statement is absurd. I have been following this thread (if nothing else for the drama:)) and this statement can't possibly help your case.

That's called being balanced and diplomatic. Going your way is bullying and offensive in the extreme - and that's why you have at least 1 guy who antagonizes you - it's your approach that is so offensive it pisses people off.

I hate to tell you this, but having tbrunet agree with you is not a good thing. He basically trolls Chris because Chris has proven his ideas wrong in the past. As abrasive as Chris is, he normally doesn't get into discussion without knowing ahead of time that he is correct... I have to give him that much.

If you wanted the term changed, wouldn't a review panel with your peers be a better place to change the "established view"?

There has been review as in "Don't you think it is odd to say a passive device can clip?" and "Can you think of any other passive devices that are said to clip?" Yes, and no, respectively, are the general responses (paraphrased). But when the question is posed "Do you think it's possible to get people to stop using "clipping" in reference to a passive device like a light valve controlled video display?" The typical answer boils down to "too much inertia".

I hate to tell you this, but having tbrunet agree with you is not a good thing.Considering you and Chris are obviously unable to even comprehend what actually define IRE..

Chris's Source Settings Guide
http://www.avsforum.com/avs-vb/showthread.php?s=&postid=4969789#post4969789
“An IRE simply a representation of volts”

That statement by Chris is false: IRE is a relative linear scale. It doesn't refer to any particular voltage or digital level until you specify the signal standard being used.

Now apparently Chris can't even grasp how simple passive devices fundamentally operate;)

Thomas is another story, and he has no place in this thread. His posting history on AVS is long and sordid, and his many claims speak for themselves. According to Greg Rogers you don't even comprehend the "definition" IRE units. I've corrected your misunderstanding in the past and still your Source Settings Guide is wrong!

Your attempt at correcting Doug use of technical terminology is literally the preverbal:

"Pot calling the Kettle black"

Considering you and Chris are obviously unable to even comprehend what actually define IRE..

Chris's Source Settings Guide
http://www.avsforum.com/avs-vb/showthread.php?s=&postid=4969789#post4969789
“An IRE simply a representation of volts”

[I]That statement by Chris is false:

Now apparently Chris can't even grasp how simple passive devices fundamentally operate;)

How exactly are you using the second quote to disprove the first?

Even gregr, who you use so often to point out that we know nothing about IRE, disagrees with your use of terminology.

http://www.avsforum.com/avs-vb/showthread.php?p=13491136#post13491136

I swear you make half of this crap up just to try to get a rise from people.

How exactly are you using the second quote to disprove the first? IRE are not simply “Volts”, but rather a relative linear scale representing percent (%) luminance from the blanking level to reference white.

Even gregr, who you use so often to point out that we know nothing about IRE, disagrees with your use of terminology. Actually Greg is spot on about the nature of IRE's and you btw are trying to cloud the issue. You and Greg did not grasp that I was referring to "ANY" hypothetical (%) IRE value. In retro I see how it was misinterpreted, but the following quote by Greg supports my position 100%..pun intended:)
If I say you need a 75 IRE digital signal, you can unambiguously figure out exactly what that is in 8-bit, 10-bit, 12-bit digital video using video-levels (16-235 for 8 bits) or PC-levels (0-255)

If I say you need a 75 IRE digital signal, you can unambiguously figure out

IRE can simply be unnecessarily confusing when the source is always typical video material, for example http://www.avsforum.com/avs-vb/showthread.php?p=14173946#post14173946 For the purpose of this forum, many people have no real need to ever be introduced to the IRE term because a black to white percentage reference with less ambiguity regarding video levels can be used. For the purpose of communication, sometimes simple definitions like "Clip (v) Forcing a signal to a certain maximum (or minimum) level, so as to avoid excursion above (or below) that level" can be beneficial so that there is a clear reference to what is being discussed.

IRE are not simply “Volts”, but rather a relative linear scale representing percent (%) luminance from the blanking level to reference white.

In an analog video system, IRE is "simply" a measure of voltage.

You can apply IRE to just about anything you want, but in the context of analog video signals (which is the context of the Source Settings Guide), IRE is "simply volts."

I don't know why I even bother with this. All you are going to do is take a bunch of quotes out of context and twist them to your liking.

IRE can simply be unnecessarily confusing when the source is always typical video materialSince theres no digital video formats in the consumer domain that contain or encode (7.5%) setup, the digital "%" level and the IRE value will be the same.

Actually Greg is spot on about the nature of IRE's and you btw are trying to cloud the issue. You and Greg did not grasp that I was referring to "ANY" hypothetical (%) IRE value. In retro I see how it was misinterpreted, but the following quote by Greg supports my position 100%..pun intended:)

How does that quote not also support the position that "IRE is simply volts" for an analog system. All it basically says is that you can determine what 75 IRE represents if you are given all of the other contextual information that is needed.

Since theres no digital video formats in the consumer domain that contain or encode (7.5%) setup, the digital "%" level and the IRE value will be the same.

Avia labels black 7.5 IRE, not 0 IRE. Depending on how the end-user uses a calibration disk, black could be said to be output at either 7.5 IRE or 0 IRE. That does not match what you said, so how is that not generally confusing?

Strictly speaking the information is actually YCbCr, but for simple understanding no one labels things that way. Of course everything I've brought up so far can be altogether avoided if black is defined as 0% and white as 100%. The latter percentages are simply far simpler to communicate to an end-user than getting into either what IRE or YCbCr mean. This is the same point I was making with Doug's attemps at a more complex definition of clipping, for the purpose of this forum it's simply not necessarily important to have all that information and a simple definition works for the task at hand.

If you would have understood the profound intention of that quote

You're so full of it, but I think you're amusing none-the-less. Like I said the information is YCbCr and understanding IRE brings nothing to the table for a general end-user calibration. As you've said in some contexts 0 IRE matches 0%, but when it comes to outputting from a player 0% is less ambiguous because black can vary when using the IRE term. So with percentage you have a defined level if you know the source, but with IRE you have to know how things are output which introduces uncertainty. Like I've offered before, if I'm dead wrong then by all means please enlighten me as to how IRE is so "profound".

Avia labels black 7.5 IRE, not 0 IRE.
The disc (lable) is broken by design!
Depending on how the end-user uses a calibration disk, black could be said to be output at either 7.5 IRE or 0 IRE. That does not match what you said, so how is that not generally confusing?

From Tektronix Standards:
"Eventually the IRE (later to be the IEEE) established a unit of measure for video signals. This "IRE unit'' was defined as 1% of the video range from blanking to peak white, without reference to the actual signal voltage."

Note theres no reference to pedestal or setup amplitude.
Strictly speaking the information is actually YCbCr, but for simple understanding no one labels things that way. (Y)Luminance or (Y')Luma, it makes no difference, the IRE measure is relative one.

The disc (lable) is broken by design!

Not exactly, because with default settings most players don't output an analog signal at 0 IRE. While I personally don't use connections like component, some people still do so there are cases where the Avia disk is correctly labeled in the way IRE has been commonly used. If they were to instead use 0 IRE for labeling then there would be cases that the player would output a commonly defined 7.5 IRE level for black even though the disk reads 0 IRE. In your championing of IRE I still fail to see how setup doesn't come into play, but in contrast that item never arises if using percentages from black to white. I think you're again fighting for a losing cause, but I did read through everything Doug wrote.

In your championing of IRE I still fail to see how setup doesn't come into play, but in contrast that item never arises if using percentages from black to white.Do you know what the definition of an IRE is?

From Tektronix Standards:
Eventually the IRE (later to be the IEEE) established a unit of measure for video signals. This "IRE unit'' was defined as 1% of the video range from blanking to peak white, without reference to the actual signal voltage.

Guys, let it go. You will never convince thomas of anything. Greg Rogers and Guy Kuo tried to explain some very basic things to him, and he accuses them of not knowing what they're talking about either.

All engaging him will do is drag a thread into the mud and get it closed.

You can add him to your ignore list instead:
http://www.avsforum.com/avs-vb/profile.php?do=addlist&userlist=ignore&u=7507853

The school where I’m employed http://www.scad.edu/ just purchased several of these HP reference monitors.



This part I bolded scares me. Is he teaching students his ideas?

Or maybe he just meant he works in the custodial arts as a hydro-ceramic engineer!?

This "IRE unit'' was defined as 1% of the video range from blanking to peak white

See another term that has to be defined - blanking. Okay, so blanking is 0v in an analog output. So if the player outputs an analog signal with a setup of 7.5 IRE, then black is 7.5 IRE. That matches Avia, but again you said that Avia's label was broken by design and black should be 0 IRE. Well then if black was labeled zero IRE, but black was at 7.5 IRE then the label would no longer be corrent. My best guess is that by your logic this situation about how one IRE label would always be incorrect in certain situations depending on the end-user must be why IRE is so "profound."

See another term that has to be defined - blanking. Okay, so blanking is 0v in an analog output. ."No, blanking on those shinny disc you call DVDs is equal to black level.

Setup was necessary for the real world limitations of terrestrial broadcast transmitters. Whether or not your DVD player adds setup is irrelelvant with respect to the encoded signal amplitude.

0IRE = DIGITAL 16 (8-bit) = 0% Stim

100IRE = DIGITAL 235 (8-bit) = 100% Stim

EDIT/
Note to self: Maybe Guy Kuo should take some notes on the subject matter:)

There has been review as in "Don't you think it is odd to say a passive device can clip?" and "Can you think of any other passive devices that are said to clip?" Yes, and no, respectively, are the general responses (paraphrased). But when the question is posed "Do you think it's possible to get people to stop using "clipping" in reference to a passive device like a light valve controlled video display?" The typical answer boils down to "too much inertia".

Since when is a display a passive device? Since when is an LCD a passive device?

A passive device would be like a screen. That won't clip anything, and you're right that it won't because it is passive. We haven't been discussing passive devices. We've been discussing electronics, specifically displays and their behavior. And nothing we've discussed is passive.

Unless you're talking about the hughes-JVC ILA displays from way back that were CRT-driven, I'm not sure how in any way you could characterize anything we've discussed as passive. There's nothing passive about it.

blanking on those shinny disc you call DVDs is equal to black level.

IRE, as exactly defined by your quote, ultimately depends on how the information is output. If the player adds setup in an analog situation, then the black level in IRE terms changes. In percentage terms black level remains the same when the player adds setup, but when the player adds setup once again very clearly the IRE changes. So the percentage label is always correct by definition, and with IRE the label is either correct or incorrect depending on how the player used. By a very strict definition of the media as what is being labeled, sure black is at 0 IRE, but when a definition breaks down in use its counter-productive to use the strict version when a simpler one works for the ultimate task at hand. Like I said if the most strict definition was beneficial, then you might as well have all calibration disks labeled with YCbCr because that's actually what they contain.

IRE, as exactly defined by your quote, ultimately depends on how the information is output. .What a given player does will not reverse engineer itself onto the calibration disc nor "change" the encoded levels.

During production its possible to raise the encoded (image) master pedestal (1%, 4%, 7.5%, 20%, ect) above the blanking level, this release for example has that relative IRE level effectively embedded into the media.

Avia is mistaken when they produced a disc with a 7.5% lable when in reality 7.5% setup was NOT encoded.

Avia is mistaken when they produced a disc with a 7.5% lable when in reality 7.5% setup was NOT encoded.

LOL:rolleyes:

If I say you need a 75 IRE digital signal, you can unambiguously figure out exactly what that is in 8-bit, 10-bit, 12-bit digital video using video-levels (16-235 for 8 bits) or PC-levels (0-255)If one can encode 75 IRE then the same is true for 7.5 IRE:p

What a given player does will not reverse engineer itself onto the calibration disc nor "change" the encoded levels...

Avia is mistaken

Some players do use the IRE term for their analog black level settings. So it's unnecessarily confusing for most people to ever learn how the player might typically have black at 7.5 IRE and the media could be said to have black at 0 IRE. That just brings up questions like - well if the media is at 0 IRE shouldn't I use the 0 IRE setting on the player? Then you have to explain about NTSC levels, and how after using NTSC levels that if the disk was labeled by how you want then even though that's not the NTSC standard that the output doesn't match the disk's labels. Like I said, ultimately from a simple-to-understand perspective if the player adds setup then black could be said to be at 7.5 IRE from the player. The first Avia disk came before digital consumer connections, so this sort of labeling a typical analog output was reasonable from a practical general-end-user perspective even if it's not a strictly by-the-book match.

Some players do use the IRE term for their analog black level settings. So it's unnecessarily confusing for most people to ever learn how the player might typically have black at 7.5 IRE and the media could be said to have black at 0 IRE.What some players do or whats confusing for most people is irrelevant to the fundamentally correct technical "definition" of the unit in question.

The same level of detail is true with respect to Doug Blackburn's succinct explanation regarding engineering terminology.

What some players do or whats confusing for most people is irrelevant

I couldn't disagree more. Avia is a company and I seriously doubt they're not-for-profit. Their ultimate bottom-line is to produce and sell a product. If the product is too confusing for the target audience then that goes against their #1 intent of producing and moving a product. I'm not thrilled by how they use the term saturation differently than TomHuffman's guide, but they do define their term so I figure they probably give further details on their labels somewhere.


The same level of detail is true with respect to Doug Blackburn's... terminology.

I don't think Doug ever offered a succinct definition, but again likewise with as basic question as what was asked to start this thread Doug's definition offers no value to someone asking such a question.

I don't think Doug ever offered a succinct definition,Not only did Doug give a clear definition, but he walked you through step by step how the LCD fundamentally operates and one artifact in particular;). If you managed to miss this, that in itself is quite revealing!

Again terminology - succinct http://dictionary.reference.com/browse/succinct

That would mean a very to-the-point definition. If he had a clear definition as you say it could convey the point in a few sentences or at most a few paragraphs. "Clip (v) Forcing a signal to a certain maximum (or minimum) level, so as to avoid excursion above (or below) that level" from Chris is very concise as to what he means when he uses that term. Drawing the line between clip and crush as Doug uses the term is not even relatively half as well defined in a simple to communicate and discuss way.

That's a specious argument. In fact it's totally ridiculous.

Again insults. Wouldn't it be more appropriate to just disagree rather than categorize?

Look, this will be my last point by point reply in this thread - the pulled muscle I've been working on has improved and I have work to do. At the end of the message, I have asked for a different approach if you choose to reply... 4 discussion points that are intended to focus your comments, stop your editorializing (i.e. insults), and to take less time to reply to.


The term clipping was invented and defined to describe certain kinds of effects that occur in signals. It was established and used by engineers in a clear an well-understood way. It still is used in that same way.

It was invented to describe a very specific effect in the analog domain - it was invented before digital existed. The term was adopted (I say inappropriately in many cases) to describe all digital domain problems even if they are unrelated to the original use and intent of the term.


Inventing your own definition of the term, and then asserting that as fact, and indeed making numerous claims that are simply false is what makes your position wrong. It also makes it extremely unuseful because you inject confusion into discussion that otherwise would be very clear.

I owned up to the initial statement when reminded there are LCDs that compress and lose steps when viewing white PLUGE or other white/above white - and I also said I haven't seen any of those myself except the Westinghouse Michael mentioned which I had forgotten about. You don't have to keep picking at that scab since I already admitted the exceptions. Is this understood by you - or do you want to keep going after it some more?

I also pointed out that there's no reason for that condition to exist since the programmer can create (and many do create) a LUT matrix that avoids making multiple digital levels have the same degree of transparency or even to have insufficient differences in transparency between steps. The light valve device does only what it is told to do by the programmer. Do you have ANY problem with these statements? The fact that there are displays with problems doesn't mean the problems are universal or required by the technology - and I have many examples... everything I have worked on in a year... where the display does not lose steps.


Again, I have a forceful argumentative style, and when someone is making clearly false claims that grossly misinform readers, I tend to rebut these wrong claims so that others reading will not be likewise wrongly informed.

On the other hand, maybe your forceful argumentaive style is just a character issue.



And if you notice in this thread, you began with the completely bizarre claim that:

A claim that I owned up to and to which you no longer need to refer at all. But then your agenda to crush the infidel just won't allow you to let it go, will it?


When you then launched into a bizarre and unfounded explanation of these assumptions of yours which are obviously wrong in both practice and theory, I made a fairly brief post that corrected these. And I've been saying the same things in as many different ways as I can to you since then. And that did include a great deal of consternation and surprise that you held so strongly such bizarre and so obviously incorrect views.

What is bizarre here is that you can't let go. You're obsessed with me. You can't handle a different point of view - even if it has NO BEARING on anything except what word describes a visible problem.


I mean, the claim that digital displays can't have their white level set with ANY test pattern is absolutely bizarre.

And once again, you are boring in on something I admitted I'd missed and there are SOME displays in this category with problems due to programming issues.



Well, I do know that you believe, as your statements above have illustrated, that it is not possible to align the white or black points of digital displays since you do not believe they can ever clip.

Look, I KNOW you are not stupid enough to belive what you just wrote here. And you know damn well that anybody can use a black PLUGE and set peak white level to 30 fL (or any other desired level) and look for missing steps in appropriate white level test patterns even if they don't think they are causing clipping when the below black stripe is adjusted to have the same brightness level as the 0% stripe. Statements like these in your replies are simply more evidence of the vindictive, nasty nature of your online persona.


Well, they do clip as has been established beyond any doubt or confusion.

What they do has come to be CALLED clipping - and all I'm saying is that I don't agree with the term.

Please explain how a passive device that has transparency levels set by a LUT matrix clips any electrical signal in the same manner an analog amplifier can be driven into clipping in the analog domain. I say there is a difference between losing or removing numbers in the electronic domain intentionally (removing the numbers 0-15 and/or 200-255) and not being able to see steps for all those numbers when viewing the display... for example... if the input signal used the full range of 0-255, but somewhere in the display, before the LUT, 0-15 and/or 200-255 disappeared... I could see calling that clipping. You have permanently removed information from the digital input signal before the LUT. But if 0-255 are still present as inputs to the LUT but you can't see below black and/or steps are missing at the top end, I don't think it is accurate to call that clipping since no digital values are ever "gone". If 0-255 are present at the LUT and you can't see all the steps from the display - what you see isn't clipping - no digital values were lost. You want to call the latter clipping anyway. I see clipping as a bad thing. Yed by your definition, intentionally setting 0-15 to the same level as 16 is clipping. By my definition, setting 0-15 to produce the same luminance level as 16 is simply correctly setting up a display for consumer video sources - 0-15 do not cease to exist in the digital (electronic) domain, they are simply all handled as if they were 16.


And that's what one observes while calibrating to locate the display's maximum white point, and also the display's black point. And indeed, they hard clip in a way that is far more unambiguous than the rather soft margins on CRTs, for example. You can see it on ANY digital display with sufficient adjustmenr range, which is darn near all of them. That includes cheap and poorly designed displays, and professional displays. And all displays in between.

I agree, when steps are lost in "modern" displays, it is more unambiguous that it has been with CRT. However, CRT and plasma can REALLY clip because of their nature as active devices... the pixels produce the light and you can have a color that gives up before the other colors... the display can be sending 255, but the pixel can only get as bright as the level it can produce with a 200 input. In that case, the LUT is programmed properly, but the pixel is too weak to do the LUT's bidding. That's really clipping - unambiguously.

You have used "clipping" in reference to the DIGITAL SIGNAL being clipped, when the digital signal was NOT clipped - all the numbers 0-255 are still present in the digital domain and the display is still processing 0-255 per the programming and control settings. The issue is that the LUT tells the light valve to do something inappropriate that causes the visible defect of lost steps or color shifts.



Ignorance is not a personal insult. Ignorance is a state of unknowing on a topic. I didn't insult your intelligence which is obviously not lacking. No matter how smart a person is, their areas of knowledge are limited. Every person is ignorant about things. In this case, you are not very well informed about digital video, and digital displays.

This is yet another insult. You are taking a disagreement about whether you can call what happens in a light valve display clipping or not and from this ONE issue, making a statement that I am not very well informed about digital video displays. That is insulting in the extreme. If I was not very well informed about digital video displays, I couldn't even make the distinction between elimination of digital values in the display and not having any digital values elminated but still having an undesirable visible appearance.

As I said previously - I am well aware of the published definitions of clipping - I simply look at what happens in a different way I think there is a distinction between removing digital values like 0-15 and/or 200-255 (or some other combiation of values) from the digtal domain, and a programming or control operation error (or intentional use of the control) causing a visible effect without removing values in the digital domain.


I mean, it follows that if you do not believe that digital displays have any maximum white point since they can never clip, then clearly you cannot properly calibrate such a display since locating that white point is crucial to properly setting white level for the highest quality image.

You are really letting me down with this statement. You are much better than this. You know damn well that "you do not believe that digital displays have any maximum white point since they can never clip" is completely wrong. Geez, really. You have to be able to do better than this. I set the maximum white level to an appropriate level for the viewing conditions and confirm that everything looks and measures well using test patterns. There are LCDs out now with white levels so high you can't even get them down to 30 fL - so those with darkened theater rooms are doomed to having eyestrain problems unless they use a fairly significant level of bias lighting. Since I calibrate almost exclusively for darkened theater rooms, the only time I'm interested in PEAK white level is when I work with a projector - and even then, not using the peak level may be desirable in some cases. Once the display is setup for the theater room viewing conditions, I always see how far the owner can move controls before negative effects begin to appear - or whether a different mode can produce a different result in both the darkened room or brighter room. But you don't know that because you know almost nothing about me, yet you seem to be able to make this silly leap that somehow because I don't think the use of the term clipping is right. Use of the term and appearance/calibration of displays are completely unrelated in this case.


This is why I referenced Guy Kuo and Joe Kane's explanations of the methodology to locate this point on digital displays while calibrating, in the hopes that you could eliminate this area of unknowing for you. It's why I've taken so much patience (which did fray at certain points, thus some of my frustrated statements) with you because you should know about digital video and how these displays behave.

I've had dinner with Joe Kane, though he may or may not remember me. I'm sure he's met so many people, as have I, that you don't remember everybody. I can't recall meeting Gary Kuo, but it's possible I have. I've heard Joe's presentations many times and had the chance to discuss some interesting issues with him at dinner. And as I've pointed out before, your attempts to educate me are quaint but a waste of time as I've seen it all before in personal reading or formal training.


Now if it's simply the word "ignorant" that is offensive, I do not mean it that way, and there isn't any connotation in that word that should be taken as insulting. I, like anyone else, is ignorant about far more things than I know.

Thanks for that, but again, you aren't aware of what I know or don't know about digital video and you feel that because I don't agree with you and what's published witin a very narrow definition of whether the use of 1 word seems right/appropriate or not... that this disagreement is a matter of ignorance. It's not ignorance. In fact, to have the opinion I have in the first playce requires knowlege of the issue. It's a matter of having difficulty with the use of a word to describe things that I don't think should be called clipping in very specific applications and situaions. I believe using the same term for conditions that can legitimately be called clipping as well as conditions I believe should be called something else fails to recognize "real" clipping and other undesirable conditions that are not caused, literally, by clipping of any electronic signal. as clipping was defined in the anlog domain decades ago. (see points at end of this reply for some specific statements and questions)


But what confuses me is the above claims that there simply is no clip point on digital displays, and that you can just set them to their maximum. Well, that's simply not so. Doing that grossly degrades the image. Why? Because the image is usually very heavily clipped in such a setting.

You are making another leap of silliness... I never said you could just set digital displays to their maximum. You are making things up in your head that are fueling your belief that I have a pathetic lack of knowledge about what I'm talking about. I sensed that you were doing something like this, but I couldn't pin down exactly where you were getting so far off track. As I posted earlier, before reading this, I set self-contained video displays to an appropriate max white level depending on the viewing conditions. Most of the time the room is dark or nearly dark and 30 fL is about where the displays end up unless there are other problems/issues that can be used by using a slightly different peak white level. The worse thing that has been happening lately is that some LCDs won't get any darker than 40 fL or so. I hate having to leave a display that's too bright... but I do recommend adding color temp balanced bias lighting to avoid eyestrain.


Your unfounded belief that they won't ever clip no matter how high the contrast setting is is a very bad understanding to have because it leads to a horrible image if you set a display that way.

Would you GIVE THIS UP ALREADY? I have already conceeded that there are displays like the Westinghouse Michael mentioned that will produce undesirable effects when set too bright. I was wrong to apply that to "every" light valve display. And why would you think I would EVER leave ANY display set to the peak white level? That's all coming from you... assumptions to fuel your indignation - I certainly never said "crank it up and leave it there". Ever.


I frankly wouldn't care at all if you had a misunderstanding that lead to no damage or no effect in what other people might draw from it. If you called it, for instance, "smashing" instead of "clipping" and made your own distinction of "smashing" in the digital domain but were perfectly able to identify the "smash" point on a digital display then it really doesn't matter to me or anyone else. I think it would be silly, but you wouldn't be doing any damage to the image quality of your own displays, that of your clients, or to others who may read your posts and follow your advice.

Well, I hate to tell you, but this pretty much sums up my concern about the use of clipping completely. I've alreay conceeded the point that there are displays like the now-proverbial Westinghouse that will do undesirable things when cranked up. But I have a whole slew of others that don't do anything untoward when Contrast is set to the maximum value - they get brighter, but there is no clipping and little or no color shift. That doesn't mean I would ever leave ANY display set to the peak white level and I'm really disappointed in you thinking I would and making the leap that somehow I cannot calibrate a video display.


If someone is wrong, you generally have to say they are in some way mistaken. None of the terms you list is at all insulting, nor should they be.

It's all in how you choose to do it - and it is your choice. You can disagree and provide your evidence without the use of anything derogatory in any way. Whether you indend to be offensive or not isn't important - the end result is important. And if the other party is offended merely saying you didn't intend to offend or the other party shouldn't be offended is a cop-out. The offense was committed, intended or not. And in every case, it's avoidable. OK, maybe some psycho will go overboard if he's politely told that you're not supposed to be viewing images with 10,000K color temperatures or higher - not much you can do about that.


If I said that and LCD display uses one LCD panel that changes colors to create a color image, you would probably say: "No, you're wrong about that. There are almost always 3 colored panels which combine to make a colored image." And you might expplain further, or I might go look for more information, and after reading and realizing "gee, I was wrong about that." I would admit that, and then I wouldn't keep explaining something that's totally erroneous.

If the issue was that simple, the above scenario would be reasonable. And in fact, when Michael said (paraphrasing) "What about the Westinghouse LCD in the training class that loses more steps the higher you set the contrast control?" I immediately said (paraphrasing) "Oops, I completely forgot about that one, and if there's one there are probably many others also... I just haven't seen any I could remember, but I do remember the Westinghouse now that you mention it." That's pretty much your above scenario, is it not? Now let that part of the discussion go! Please!


Thomas is another story, and he has no place in this thread. His posting history on AVS is long and sordid, and his many claims speak for themselves.

Nevertheless, something sicked him on you. A different approach may have produced a different end result. Or not - you never know. As above, sometimes there's nothing you could do to change the outcome. But based on this thread, I wouldn't be too surprised if a little more diplomacy wouldn't have created a different outcome.


I don't have any problems with our disagreement either, and I don't see why you should be insulted. I made my argument more strongly only after you continued to assert things that were clearly incorrect. If you observe the original posts in this thread you will see that.

I'm insulted because you have somehow come to the conclusion that I have no business questioning anything that's published. And are completely dismissive of my view of what should or shouldn't be called clipping. And I'm also insulted by your... insults (just can't think of any other word for it) about how I should be unable to calibrate a video display because I don't like the word clipping as applied to light valve displays and as applied to digital displays when the undesirable effect has nothing to do with digital values being eliminated (i.e. when you can't see blacker than black, but 0-15 continue to exist inside the display and continue to be processed by the display per the programming and control settings). I even have a problem saying there's any clipping going on in a display that never displays blacker than black. If the display shows 16 as the minimum luminance level (0% white), an 0-15 are nowhere in evidence, even if 0-15 are removed somewhere in the display, it's tough for me to call that clipping because the display is FINE. Clipping has always been a BAD thing. It is a GOOD thing to have the display showing 0-16 as all the same 0% white luminance level.


I made a fairly short and basic normal post in answer to the original poster that described the clip point in order to locate the white point of the display in order to assist him in calibrating it. You made a claim that such a point doesn't exist and that digital display's won't ever clip. I reacted briefly in surprise, as did Michael. It was a strange assertion you made, and I didn't even spend time to refute it because it was so far out of left field.

Let's not hash this over AGAIN. I've already conceeded the point and moved on to explain what I was getting at in the first point. You have to let this go now. When you're having a discussion and someone says - OK, you're right, there are some displays that do that (and some that don't, which is all I have been seeing within recent memory, save that Westinghouse which I'd forgotten about.)


And I'm still doing that now. But now it's due to surprise at your unwillingness to admit your error or the limits of your understanding here.

Geez o Pete. I did admit the initial error. Get off this horse PLEASE.


You said that anything in the digital domain cannot clip. (which is wrong)

If the digital input range is 0-255 and the digital output range to the LUT is 0-255... was there any clipping in the digital domain?

When the digital signal path contains all digital values from 0-255 right up to the input of the controller/LUT for the light valve device, can a light valve device that is operating properly fail to achieve the level of transparency or reflectance sent by the LUT?


You said that you can crank the contrast level up to the maximum without any clipping or colorshifint. (which is wrong)

Not for the displays I've been working on - and I have admitted that there are displays that do have the problem (how many times will I have to say that before you stop running the old tapes?). You have never admitted that there are digital displays that don't go crazy at the highest Contrast settings. So where's the agreeent on the point. I have displays here and displays I have calibrated that don't fall apart as you increase Contrast, even to the max setting. That doesn't mean ANY of them are left at the max setting. It only means you need a light meter or calibrated eyeball to set the display to some reasonable setting for your viewing conditions and a 100% white window works fairly well for that.


You said that "Most modern digital displays can't have the white level set with ANY test pattern because they show every shade of white even beyond 100% white" (which is wrong)

Depends on your perspective. From my perspective, I haven't seen any displays that did not perform like this since that Westinghouse LCD in the training class that I forgot about. I did say "Most" not "all" and not "every" - so crucify me, if more than 50% do drift off atthe top end, for not using the word "Many" instead of "Most". I mean, that's really all we're talking about here isn't it? You have your panties in a bunch over this statement because the way I worded it 51% or more "modern" displays would have to meet this wording for it to be accurate, while if I had used "many" instead, we'd be thinking 49% or fewer. From my perspective, even counting the Westinghouse, MOST of the light valve displays I've worked on in the last couple of years perform as described in that sentence.


Destroy you? In what way? By stating the obvious that you are mistaken about a few thing that you've said here? The one digging the hole here is you. I mean, making up definitions on your own and using yourself as a reference is pretty hilarious.

It was supposed to be hilarious because your vindictive intolerance and lack of consideration of anything other than the party line was getting way out of control. It was a reaction to your 'crush the infidel' attitude and inability to focus on the current issue without going back and re-entering points I have conceeded already.


Well, thank you for at least admitting that you are beyond reasonable behavior here. No professional or technical reference will dissuade you from your beliefs. No evidence or examples will shake your own personal thinking. No amount of science will shake your unscientific inventiveness.

Yes, because as we all know, science is built on the foundation that nothing that original thought is NOT allowed. Ever. Crush it.


What's hilarious is that you think everything is so black and white.
Great. Today founds the beginning of Doug Blackburn's personal reality. Doug Blackburn is the new authority on everything video. We will all heed Doug's expertise, and his powerful creativity to make up new terms and concepts as he sees fit. From this point on we should all ignore everything we have learned or read before this day, and celebrate a new reality as elucidated by Doug Blackburn. A new math. A new science. A new physics. An entirely new way of knowing the world.

Gag. We're talking about something that is almost insignificant - the point at which I'm comfortable calling a visible image defect in a digital video display clipping or whether I prefer to use a different description. Your hissy rant really doesn't quite capture that context.


Should I or others encounter your posts in the future, we should take care to recognize that your claims are based totally on your own creative ideas about engineering and science, and are a unique and separated reality apart from our own common reality.

This is yet another offensive and unnecessary "commentary" that has nothing to do with the discussion. Let's just keep the issue focused on the issue and not the world, the universe or the unfinished Unified Theory.


It isn't? Didn't you just go through how now amount of technical literature or objective fact will dissuade you from your beliefs? Why not just make up a new ohms law too while you're at it. Make resistance something that increases current flow instead. I mean, as long as you're just making things up and admitting that you don't care what any objective reference or logic or reason says.

You did say you weren't trying to be intentionally insulting, right? What do you call this? It's another example of your character and inability to focus. In fact, I've already said that a resistor is an example of a passive device that can't clip and is analogous to the pixels in a light valve display... which also can't clip... each pixel is a light resistor.


Why not put Purple as a color in the rainbow too?

Because we're talking about 1 thing you can't focus on. And because purple being a color of the rainbow hasn't come up in an AVS thread yet.


I mean, if you're going to create your own alternate reality, at the very least go all the way with it. I mean, I could respect that if you created your own complete way of knowing.

This is another unfortunate digression. I'm talking about being in disagreement with the use of the term "clipping" and you are in "crush the infidel" mode again. Even though you keep saying you don't do that and don't intend to be insulting, you keep doing it. Over and over and over again.


Go for it! Doug Blackburn's new science. A new objective reality.
I'm very interested now!

Be as interested as you like but we're talking about whether I think it's appropriate to call something clipping or not. You focus on that - you keep looking for all manner of insults (that you can then claim aren't insults, really) and escallation. I'm refusing to participate. Let's keep the discussion to the points that I made in this post. You are free to agree or disagree with them. I'd also like you to indicate whether you "get" the statement whether you agree with it or not. In fact, I'm going to ask for a miracle... how about a reply that addresses nothing but the 4 points listed below which are a new try at putting my thoughts on this in a compact form. Leave out all the other crap, and just focus on these ideas...

1) If the digital input signal contains the numbers 0,1,2,3...254,255 and the LUT/controller is processing all of those numbers following instructions written by the designer/programmer, I don't think it is appropriate to call undesirable visible problems at or near peak luminances clipping since all values from 0-255 are present and being processed for display. I understand that people call the visible effect clipping, but I think it's possible for a better, more descriptive term to come into popular use, but it's not likely to happen. This is a statement you can agree with or disagree with - I don't need a bunch of references, just a reply that clarifies exactly what clips and where the clipping happens if you don't think I have it right.

2a) Condition: The digital input signal contains the numbers 0,1,2,3...254,255 but the digital values are processed intentionally or unintentionally inside the display so the range becomes 18,19,20....199,200 prior to the LUT/controller - I woud agree with calling that clipping in the digital domain - there's no question in my mind that this would follow the general use and itent of "clipping" as it exists in analog signals. Something (information) has been removed.

2b) Would you agree that this condition is difficult or impossible to find a real-world product?

Sidebar for 2: I want to leave out signal loss or long cable runs from this discussion to keep focus on the bits - the numbers. Digital signals don't have "excusrion" the way analog signals have "excursion" - in otherwords, digital 255 while still existing as a digital signal, does not have a higher "excursion" or voltage than digital 0 - after the LUT in the controller when you are OUT of the digital domain, yes, THEN you can deal with the concept of "excursion" or voltage levels if those are appropriate/used for any given display technology. If we look at digital video signal waveforms, an "all zeros" waveform looks the same as an "all 255s" waveform. Do we agree on this?

3) If thedigital input signal contains the numbers 0,1,2,3....254,255. And the display outputs 16-255 because it is SUPPOSED TO... I don't like saying 0-15 are clipped. They still exist as digital numbers, they are still processed. The processing simply assigns them same luminance level as 16. I don't think it is accurate to say (for example) 'When I use a low luminance PLUGE pattern to set black level, I am clipping the values from 0-15.' 0-15 still exist as digital values in the display. They are not lost. They no longer appear in the image because 0-15 have all been assigned, INTENTIONALLY, the same luminance value (transparency or reflectance value in the case of light valves) as 16 (0% white). Here I'd like some acknowledgement about whether you "get" what I mean by this statement - feel free to agree or disagree - I don't need a bunch of references or value judgements. If you don't agree, I would be most interested in hearing where you think the clipping actually happens, just your thinking based on what you know is fine. Is the clipping present in the electronic signal? If you agree that the digital electronic signal itself does not have 0-15 removed, where does the clipping happen exactly? Is it in the controller/LUT when 0-15 are all assigned the same luminance value as 16 even though this is all "per plan" with no errors? Do you think it's OK/right/appropriate to call something good/desirable/proper clipping?

4) If the input is 0,1,2,3...254,255, and the output all the way up to the controller/LUT is the same and the controller?LUT outputs 0-255 levels also, but when you measure the display and read at least 1 color doing something like this (in % of the color in question) 0,1,2,3...91,92,93,93,93,93,93, etc... the pixels can't get any brighter than 93% of the desired luminance level for that color even though the digital values continue to rise all the way to 255... that's something I can get behind calling clipping. Now here's the big question... do you see how this condition can never happen in a light valve display? In otherwords, being passive, the light valve merely transmits or reflects a calibrated (by programming) amount of light for each digital level. If the light valve device itself is working properly, it will never stop (using LCD as an example) at 93% transparent when the digital numbers are telling it to go to 100% transparent. A program that tells the LUT to output transparency or reflectance levels equal to 93% white for a 255 digital level to the controller/LUT is wrong, but there's no clipping of anything electronic.

Sidebar to 4: I want to avoid discussions about operating range as it will just muddy the water, I'm well aware of the fact that 0-255 digital values and 0%-100% reflectance/transparency for any given display device are not in lock-step, you have to have some wiggle room to avoid running out of range, though manufacturers push the envelope mercilessly trying to achive higher luminance and contrast ratios. And you do need to be able to set the blackest black at either 16 or 0 without fail.

Do you think it's OK/right/appropriate to call something good/desirable/proper clipping?

In a general forum like this not everyone is going to have such a word association where clipping is necessarily an undesirable effect like the common electronics usage. To someone asking a question like the original, clip might be associated with say clipping coupons. The coupons come as a full page (0-255), but if you clip out a coupon (set a control so 0-15=16 on-screen) then the main page is effectively reduced (only levels 16 up produce different on-screen outputs). In the coupon comparison, "Something (information) has been removed" from input to output. I did find the LUT discussion interesting, but without some clear reference definition I have a hard time not just writing this off as your own word association similar to how some women have a major aversion to the word easy.

There's one thing wrong with much of this discussion. The "clipping" we are talking about has nothing to do with the actual LCD screen. Talks of "light valves" are totally irrelevant. The processing has "clipped" the signal before the panel can even try to display it. If there are 1024 levels per color(10 bit), the processing won't try to address any levels beyond 1024 (256 on an 8 bit panel). In this sense the LCD panel isn't actually part of the what we are observing. Much of this discussion seems to be assuming that the processing can address beyond the 8 or 10 bits that the panels are capable of (the processing would have to be higher bit then the panel for this to even be a possibility). It's the processing that is mapping everything to a digital level, so when there is no head room left the processing just maps everything that would be above 1024 to 1024 (or 256 for 8 bit).

But then your agenda to crush the infidel just won't allow you to let it go, will it?Did he attempt to crush the infidel or just clip the infidel? I believe clipping is a 15 yard penalty, but crushing an opponent could get you thrown out of the league.

clamp.

Nevertheless, something sicked him on you. A different approach may have produced a different end result. .Spot on observation there Doug. With my very first post on this forum I was greeted by Chris Wiggles with his obligatory long winded dissertation (sound familiar;)) regarding IRE units and how they are simply “volts”.

According to Tektronix Corp who btw just so happen design the most respected benchmark video measurement products in the industry, says:

From Tektronix Standards overview:Eventually the IRE (later to be the IEEE) established a unit of measure for video signals. This "IRE unit'' was defined as 1% of the video range from blanking to peak white, without reference to the actual signal voltage.:)

According to Tektronix Corp who btw just so happen design the most respected benchmark video measurement products in the industry, says:

From Tektronix Standards overview:

How do you interpret "without reference to the actual signal voltage?"

IRE units are relaitve..0 IRE does not always equal 0mV nor does 100 IRE always equal 700 mV

IRE units are relaitve..0 IRE does not always equal 0mV

Under what conditions in an analog output does 0 IRE not equal 0mV? All references I've seen to US NTSC levels refer to 0mv and 0 IRE interchangably.

Under what conditions in an analog output does 0 IRE not equal 0mV? All references I've seen to US NTSC levels refer to 0mv and 0 IRE interchangably.
http://www.avsforum.com/avs-vb/showthread.php?t=1009493&page=2

Here are several examples of Greg Rogers attempting to prove that I can’t measure (analog voltage per a unity calibrated chain) the original (digital) encoded IRE level that’s embedded on a given DVD test disc. While the following quotes by him are correct in that blanking level and black level are not ALWAYS at 0 mV. Greg was reaching here because of course I was talking about “standard” SMPTE RP and industry standard Rec. 601 and 709 DVD production DVD release formats.
Ever hear of MII or Betacam? They are also "traditional SDTV component systems". Betacam sync levels are not 300 mV. Betacam reference white is not 700 mV. MII black level is not at the blanking level in North America.
So if you measure the YPbPr signals from a device (including a professional DVD player) that complies with the Betacam standard and the black level isn't at blanking and the reference white isn't 700 mV above blanking, it isn't calibrated correctly?
There are also consumer DVD players (CEA standards) and DVD players made for professional standards. There are DVD players for the US, Europe, and Japan. So black is not always at the blanking level. Reference white is not always 700 mV above blanking.

In a typical US NTSC analog output I would have to agree that "black level is not at the blanking level", which by my understanding is why the standard black level is said to be at 7.5 IRE. Regardless of if setup is included or not in a US NTSC system blanking could be said to be at 0mV as far as I know. Depending on setup black could typically be at 7.5 IRE with setup, but many players also allow for a setting to output black at 0mv.

Like stated above, as far as I know in a typical US NTSC DVD system analog output 0mV = 0 IRE. You said "0 IRE does not always equal 0mV". I asked under what conditions is that true? You replied with quotes that do not reference 0 IRE and 0mV.

Like stated above, as far as I know in a typical US NTSC DVD system analog output 0mV = 0 IRE. You said "0 IRE does not always equal 0mV". I asked under what conditions is that true? You replied with quotes that do not reference 0 IRE and 0mV.I actually agree with the nominal 0mV = O IRE convention.

My point was, 0 IRE actually represents the "Blanking level" which can be the "black level". For example the THX DVD test patterns are contrived and follow the production work flow process. The encoded black level offset can deviate from the blanking level from release to release so that the blanking level and the encoded video black level are offset by x amount of mV.

IRE units are relaitve..0 IRE does not always equal 0mV nor does 100 IRE always equal 700 mV

Again... how does that disprove the fact that for an analog system, IRE is "simply" a representation of voltage.

Again... how does that disprove the fact that for an analog system, IRE is "simply" a representation of voltage.Your right:eek:

"IRE references the actual signal voltage"

I'll notify Tektronix Corp on your behalf!

The O IRE or black level

Like already stated, it's confusing to use these terms interchangably. I get you're probably talking about the disk based on prior definitions, but when you bring up mV later in the sentence it's not clear because the disk has no such reference to voltage. The only context where a voltage reference arises is in an analog output. In a typical US NTSC analog output with setup, 0 IRE does not necessarily equal black level due to typical setup which you already agreed with.


black level can deviate from release to release so that the blanking level and the O IRE Black level are offset by x amount of mV.

The disk is YCbCr information. As far as I know there is a standard defined black video level for the digital information. As far as I can figure out what you're talking about would mean that black level has deviated from the defined digital level of black. All in all I fail to see how this proves "0 IRE does not always equal 0mV" because an analog output is the only area where mV is typically discussed and you've already agreed that 0mv = 0 IRE = blanking and that setup moves the defined standard black level to 7.5 IRE (again in a typical US NTSC analog output).

As far as I know there is a standard defined black video level for the digital information.Of course and its 0mV.

But there is nothing preventing THX i.e. to encode a test pattern and the entire motion picture as well where the Black level -- is not parked at the nominal level (16) for 8-bit video, which is contrived from the 0-700mV analog model. It's possible for THX to "encode" the standard SMPTE color bar pattern with 800mV of p-p amplitude. In a calibrated environment there is A/D/A unity, the digital part needs to be transparent.

Of course and its 0-700mV

I'll avoid how black level wouldn't be defined as a range, and that this is a mistake in regards to video range rather than a specific black level as referred to in what you quoted.

Like previously stated, digital information such as DVD is not stored as mV. The disks store video information digitally as YCbCr. As far as I know there is a defined standard video black level - which would be a specific digital YCbCr value (which could be converted through a matrix to the more common RGB 16 reference). In whatever way an analog system uses that information is completely separate from the digital system. There is no reason to bring up analog in discussion of a digital format such as DVD until there's a point of conversion to analog. The reason to avoid discussing analog representation of a digital format until conversion is due to items like setup which will alter how the digital information would ultimately be represented in an analog system.

Like previously stated, digital information such as DVD is not stored as mV.

..The disks store video information digitally as YCbCr.Whether you like it or not the image information is contrived from an analog model.

http://www.filmlight.ltd.uk/pdf/technicalnotes/FL-TL-TN-0139-StdColourSpaces.pdf
Filmlight Standard Colour Spaces
9.4 Luminance-Chrominance Coding

From page 36 (8-bit column)

“Y’ Black point = 0 mV = 16”
“Y’ White point = 700 mV = 235”

Cb’ and Cr’ are 16-240 respectively

I'll avoid how black level wouldn't be defined as a range, and that this is a mistake in regards to video range rather than a specific black level as referred to in what you quoted.

Like previously stated, digital information such as DVD is not stored as mV. The disks store video information digitally as YCbCr. As far as I know there is a defined standard video black level - which would be a specific digital YCbCr value (which could be converted through a matrix to the more common RGB 16 reference). In whatever way an analog system uses that information is completely separate from the digital system. There is no reason to bring up analog in discussion of a digital format such as DVD until there's a point of conversion to analog. The reason to avoid discussing analog representation of a digital format until conversion is due to items like setup which will alter how the digital information would ultimately be represented in an analog system.

We might as well just leave this alone. He thinks you can somehow "encode 800 mV" and that setup is encoded on the disc. Heck, most of the time he proves himself wrong if you let him type long enough.

Whether you like it or not the image information is contrived from an analog model.

http://www.filmlight.ltd.uk/pdf/technicalnotes/FL-TL-TN-0139-StdColourSpaces.pdf
Filmlight Standard Colour Spaces
9.4 Luminance-Chrominance Coding

From page 36 (8-bit column)

“Y’ Black point = 0 mV = 16”
“Y’ White point = 700 mV = 235”

Cb’ and Cr’ are 16-240 respectively

Of course, it has nothing to do with the fact that 8-bit provides 256 steps and that headroom and footroom is needed for processing. No, they just contrived 16 and 235 from 0 mV and 700 mV using a magical formula. I'm sure that formula has IRE in it somewhere.

We might as well just leave this alone. He thinks you can somehow "encode 800 mV"I have DVE calibration disc and I measured exactly 800mV on one of their "special" test patterns.

In the production (encoding) process ANY LEVEL of pedestal (setup), flare, gamma, gain, ect. can be adjusted first then encoded.

..and that setup is encoded on the disc
http://www.mastersofcinema.org/reviews/onibaba.html

“In the case of the Asmik Ace Onibaba (and several other R2 Japan discs we have come across), the image — as digitally encoded on the disc — has for some odd reason had 7.5 IRE setup already added.”

“For the purposes of this comparison, VOB files containing the first few minutes of the film were lifted off of both DVDs. The VOB files were then demultiplexed and losslessly exploded into individual 8-bit grayscale (256 grey levels) TIFF files using the Linux tools transcode and DeCSS which were adapted and compiled for Mac OS X by the author. We chose to use the first 2750 frames of the film for our comparison, shown in a large montage here”

One reason that the NTSC pedestal level was incorrectly encoded on a few DVDs was because they were transferred from D2 or D3 digital composite video tape recorders (often used for LaserDisc mastering - although the LaserDiscs were not themselves digital). NTSC digital composite video includes NTSC setup (pedestal) and it should have been removed for DVD mastering, but sometimes it wasn't by mistake.

Since when is a display a passive device? Since when is an LCD a passive device?

A passive device would be like a screen. That won't clip anything, and you're right that it won't because it is passive. We haven't been discussing passive devices. We've been discussing electronics, specifically displays and their behavior. And nothing we've discussed is passive.

Unless you're talking about the hughes-JVC ILA displays from way back that were CRT-driven, I'm not sure how in any way you could characterize anything we've discussed as passive. There's nothing passive about it.

A resistor is passive. A motor-driven variable resistor is still passive. A filter is a passive device. A variable filter is still a passive device. LCDs are variable filters, DLPs are variable reflectors. So they are still passive.

For any given frame interval... say 1/60th of a second, the LCD pixel achieves a programmed transparency i.e. 50% transparent, for that pixel for that frame interval. The transparency changes for the next frame but remains constant for that frame interval. The DLP device is similar, though it reflects or deflects light at variable intervals during the 1/60th of a second to achieve 50% reflectivity or whatever other level is needed for that image/frame.

In a general forum like this not everyone is going to have such a word association where clipping is necessarily an undesirable effect like the common electronics usage. To someone asking a question like the original, clip might be associated with say clipping coupons. The coupons come as a full page (0-255), but if you clip out a coupon (set a control so 0-15=16 on-screen) then the main page is effectively reduced (only levels 16 up produce different on-screen outputs). In the coupon comparison, "Something (information) has been removed" from input to output. I did find the LUT discussion interesting, but without some clear reference definition I have a hard time not just writing this off as your own word association similar to how some women have a major aversion to the word easy.

What it really boils down to is that, to me, is the fundamental difference between clipping in a plasma panel where one color simply cannot achieve the same level as the other colors - you crank the contrast up and up and up, and at some point, one color starts to run out of gas then it flatlines while the other colors continue to get brighter. That's a physical limitation of an active light producing pixel. This sort of clipping doesn't even product lost high-luminance steps, it always produces a color shift unless by some miracle, all 3 colors give up at the same time (it might be possible, but I haven't seen that happen yet). When 1 color gives up before the others, you get color shift if you continue to advance the Contrast control.

In an LCD, you don't have the panels/colors getting out of sync (color shifts as Contrast is adjusted up and down) in well-designed displays, nor to you see loss of high-luminance steps in well-designed devices. The well-designed devices have controller/LUT programming and adjustment ranges that prevents assigning the same level of transparency to multiple digital values. A poorly designed (perhaps inexpensively designed? or less thoughtfully designed?) LCD device will allow multiple digital levels to be assigned the same transparency level which can (and does in some LCD products) produce "lost" steps.

Did he attempt to crush the infidel or just clip the infidel? I believe clipping is a 15 yard penalty, but crushing an opponent could get you thrown out of the league.

Man, you never know who the camera is going to find in the crowd at the game! LOL!

A resistor is passive. A motor-driven variable resistor is still passive. A filter is a passive device. A variable filter is still a passive device. LCDs are variable filters, DLPs are variable reflectors. So they are still passive.

For any given frame interval... say 1/60th of a second, the LCD pixel achieves a programmed transparency i.e. 50% transparent, for that pixel for that frame interval. The transparency changes for the next frame but remains constant for that frame interval. The DLP device is similar, though it reflects or deflects light at variable intervals during the 1/60th of a second to achieve 50% reflectivity or whatever other level is needed for that image/frame.

Why not refer to LCDs as transmissive and avoid the possible debate over active and passive?

But I have a whole slew of others that don't do anything untoward when Contrast is set to the maximum value - they get brighter, but there is no clipping and little or no color shift.

Can you give me an example? You keep saying this, but I've never seen such a display, and I have seen a lot of displays.

What specific digital display have you experienced, with what source, that did not have the adjustment range to be forced into clipping?

In my experience, I have never found a digital display that could not be forced into clipping with a fairly standard video or graphics source. That doesn't mean it doesn't exist, but 100% of the scores of displays I've dealt with all have sufficient adjustment range for this to occur. Professional and consumer displays, and pricepoints from cheapie to >$100K.

And because you did ask some questions, and despite the fact that you are great at avoiding nearly all the questions I posed to you, I will still answer them

If the digital input range is 0-255 and the digital output range to the LUT is 0-255... was there any clipping in the digital domain?

It can be clipped in the LUT, or it can be clipped beyond that. You seem to be of the opinion that as long as everything makes it to the LUT, then there can be no clipping ever, since somehow everything past that doesn't count.

Yet you also claim that CRTs can clip no problem, but they don't even have a LUT because they're not linear. I fail to see the difference between a phosphor saturating and reaching a maximum (which you call clipping) and an LCD element reaching a maximum (which you refuse to call clipping). There is a logical discontinuity here.

When the digital signal path contains all digital values from 0-255 right up to the input of the controller/LUT for the light valve device, can a light valve device that is operating properly fail to achieve the level of transparency or reflectance sent by the LUT?

Sure it can reach a maximum and go no higher. Every video engineer I've encountered would call that pushing a display into clipping. It's a straightfoward application of the term in a way that is completely sensible, well understood, and well established. It leads to minimal confusion, and very accurately describes the visual effects of what is occurring. But hey, what if you're in a system without a LUT, and no gamma, driven linearly? Where is your imaginary boundary where clipping can no longer occur?

You have never admitted that there are digital displays that don't go crazy at the highest Contrast settings.

There is nothing crazy about it. It's actually a positive design feature to have a significant adjustment range to allow alignment to a variety of input ranges, which I have already explained.

2b) Would you agree that this condition is difficult or impossible to find a real-world product?

Nope. I most certainly wouldn't. Happens all the time.

I want to leave out signal loss or long cable runs from this discussion to keep focus on the bits - the numbers. Digital signals don't have "excusrion" the way analog signals have "excursion" - in otherwords, digital 255 while still existing as a digital signal, does not have a higher "excursion" or voltage than digital 0

Obviously.

after the LUT in the controller when you are OUT of the digital domain, yes, THEN you can deal with the concept of "excursion" or voltage levels if those are appropriate/used for any given display technology.

Excursion is not equivalent with analog excursion only. Digital signal "excursion" is numerical, but clipping still most certainly can occur either in the digital domain or at an ADC stage by overdriving the ADC.

If we look at digital video signal waveforms, an "all zeros" waveform looks the same as an "all 255s" waveform. Do we agree on this?

Yes, I'm not talking about waveforms of a digital transmission signal. That's not relevant to anybody since it tells you nothing of the content values being conveyed. It's completely irrelevant. It's not like analog where you can scope the analog waveform and it's the same thing as the content signal. We've been way WAY past this distinction from the beginning.

3) If thedigital input signal contains the numbers 0,1,2,3....254,255. And the display outputs 16-255 because it is SUPPOSED TO... I don't like saying 0-15 are clipped.

What if it's a graphics source? Would you still not call that clipping?

I don't think it is accurate to say (for example) 'When I use a low luminance PLUGE pattern to set black level, I am clipping the values from 0-15.' 0-15 still exist as digital values in the display. They are not lost.

Well I do think it's pretty accurate to say that. You are the one assuming that clipping is necessarily bad. In this case, if it's a video source, it isn't bad, but part of proper calibration.

And yes they are lost. They aren't seen on the display are they? That means they're not present, they've been lost. Is that a bad thing? Maybe, maybe not. It depends. If it's a graphics source, that's a bad thing. If it's a video source, it may not be a bad thing. It's still clipping.

If people set their white level for video content to 235, that may not be a bad thing overall, depending on their preference. It's still clipping, levels above 235 are clipped off. Whether you judge that good or bad may depend on the situation.

Clip (v) Forcing a signal to a certain maximum(or minimum) level, so as to avoid excursion above (or below) that level.

I don't see any attribution or assumption in that definition that clipping is always bad. That's why there are numerous reasons why digital signals are clipped intentionally for good reasons. I'll certainly admit that usually clipping is used to describe negative situations, but it is not always so.

And regardless, what we've been discussing in this thread is a negative example of clipping if a display is adjusted so. However, it's not negative for the display design, since significant adjustment range is advantageous.

Is the clipping present in the electronic signal? If you agree that the digital electronic signal itself does not have 0-15 removed, where does the clipping happen exactly? Is it in the controller/LUT when 0-15 are all assigned the same luminance value as 16 even though this is all "per plan" with no errors?

It may be, it may not be. It depends on the display design, and where the adjustment is being implemented. But I fail to see why this distinction matters. You have already explained that you feel CRTs and plasmas reaching phosphor saturation are "clipping" but nothing in them is at all being lost. The entire signal is there, all the way to the phosphor. Yet you still call that clipping.

How is that any different than having the signal all the way to the LCD, but that doing the equivalent of saturating, by reaching maximum translucence?

Do you think it's OK/right/appropriate to call something good/desirable/proper clipping?

Of course. That's assuming that what is being described is clipping, and not lime jello which I find delicious and desireable, but not clipping. ;)

4) If the input is 0,1,2,3...254,255, and the output all the way up to the controller/LUT is the same and the controller?LUT outputs 0-255 levels also, but when you measure the display and read at least 1 color doing something like this (in % of the color in question) 0,1,2,3...91,92,93,93,93,93,93, etc... the pixels can't get any brighter than 93% of the desired luminance level for that color even though the digital values continue to rise all the way to 255... that's something I can get behind calling clipping.

How is that different than maxing out at 100%? It looks the same at the display either way if you adjust it to the same clip point.

In otherwords, being passive, the light valve merely transmits or reflects a calibrated (by programming) amount of light for each digital level.

And phosphor isn't passive? I mean phosphor is way more passive than an LCD which is being directly electronically controlled and manipulated. And there's not even a LUT telling it what to do in a CRT at all. So how come you say that a CRT can clip, but an LCD can't? This logic is absurd.

One reason that the NTSC pedestal level was incorrectly encoded on a few DVDs was because they were transferred from D2 or D3 digital composite video tape recorders (often used for LaserDisc mastering - although the LaserDiscs were not themselves digital). NTSC digital composite video includes NTSC setup (pedestal) and it should have been removed for DVD mastering, but sometimes it wasn't by mistake.Thank you Greg!

For providing validation regarding whether its even possible to encode setup (aka pedestal) into the digital domain, be it SDI or DVD, ect. Many forum members in this thread have suggested thats its not;)

THX is able to encode calibration test patterns at "non stnadard" levels per varying digital production workflow.

Thank you Greg!

For providing validation

Do you really believe what you post or is it just little grasp of context? Do you seriously think "Many forum members... have suggested" that it's impossible to have an unintended error during encoding? That's all the quote is about, a mistake during encoding that caused the final YCbCr levels to deviate from standard intended levels. This error means that the disk's YCbCr levels don't match up with the video encode on other disks. If levels differ from the standard, then any "calibration test patterns" would only apply to that non-standard video encode. To use the label "calibration test patterns" alone without qualification with such an encode might make it unclear to users that such patterns shouldn't be used for calibration in general because they don't match with the standard.

The post was validation to the FACT that IRE (setup) can be encoded and that (Y) luminance data is referenced from a nominal 0-700mV analog model. THX video calibration test patterns are not a mistake or error.

Chris Wiggles has posted long winded dissertations on how IRE are just volts and can't be represented on "DVDs" or even in the digital doamin at all:p

THX video calibration test patterns are not a mistake or error.

I'm not familiar with the THX patterns, and you seem well-informed, so I have a few questions.

- I've read that some of the THX patterns have different black levels, is this true? (I'll assume the answer is yes)

- If that is true, how do the different black levels on the different patterns come about? (I would have to assume the patterns are encoded to different YCbCr video levels)

- Being how there seems to be a certain defined YCbCr level for black if a video is encoded to standard levels (RGB 16-235 model), could it be considered a mistake or error for someone to calibrate to a THX pattern that uses non-standard YCbCr level for black and then use those settings when watching video that has been encoded to standard video levels?


The post was validation to the FACT that IRE (setup) can be encoded

Certainly you can encode video to whatever YCbCr levels you want, but whether or not those levels correspond to the standard 16-235 range is what matters for the video to display as intended if someone has adjusted their display to expected video levels. If you aviod standardized levels, then every video has to have its own calibration pattern.


a nominal 0-700mV analog model

For clarity of what exactly is going on, analog doesn't necessarily have to be brought up at all in such a simple discussion of a digital format like DVD.

Certainly you can encode video to whatever YCbCr levels you want, but whether or not those levels correspond to the standard 16-235 range is what matters for the video to display as intended if someone has adjusted their display to expected video levels.Thats all I'm saying on the subject matter..the range is referenced from (legacy) analog components. Image black level, gamma, gain, ect. can be adjusted in the analog or DIGITAL domain.

For example with most real world full frame images the nominal black level is not usually resting on 16 for 8-bit video.. there will be a pedestal level there and this tweak varies from (color correction) artist to artist and theres no one "standard", aslo the obligatory artifacts of gamma would crush detail in that region if one simply made all image shading the same. The same is true with respect to white, its not a good idea to blow out whites at or above 100 IRE either.

Have a nice day!

Thomas is wrong.

Thomas is wrong.

Actually, we all owe Thomas an apology. It looks like he was right all along.

http://www.avshd709.com/DiscVolts.jpg

Actually, we all owe Thomas an apology. It looks like he was right all along.I'm 100% right. If you don't know what defines a given UNITY MODEL, then maybe you should try http://www.wikipedia.org/



http://www.avsforum.com/avs-vb/archive/index.php/t-904279.html
Digital excursion beyond "YCbCr reference levels" i.e. when D/A will produce RGB voltages outside the 0-700mV unity model.
John instead of miss using that cheap Mutil meter you have, I suggest you purchase the following:

http://www.jvbdigital.com/jvb.asp?cur=2&level=sdi&page=title&title=103
Pioneer Elite DVD-F07 SDI upgrade

http://www.testequipmentdepot.com/tektronix/video/waveformmonitors/digital/wfm5000.htm
Tektronix WFM5000 HD/SD-SDI Multi-STANDARD Waveform Monitor
Parade and overlay presentation of SDI signals in RGB, YPbPr, YRGB or pseudo-composite formats

http://www.jkpi.net/DVD_NTSC.php
Digital Video Essential

John theres at least one "special test pattern" a (1-254) ramp that measures exactly 800mV and all the nominal patterns (black –white) measure exactly 0-700mV:p

I'm 100% right. If you don't know what defines a given UNITY MODEL, then maybe you should try http://www.wikipedia.org/



John instead of miss using that cheap Mutil meter you have, I suggest you purchase the following:

http://www.jvbdigital.com/jvb.asp?cur=2&level=sdi&page=title&title=103
Pioneer Elite DVD-F07 SDI upgrade

http://www.testequipmentdepot.com/tektronix/video/waveformmonitors/digital/wfm5000.htm
Tektronix WFM5000 HD/SD-SDI Multi-STANDARD Waveform Monitor
Parade and overlay presentation of SDI signals in RGB, YPbPr, YRGB or pseudo-composite formats

http://www.jkpi.net/DVD_NTSC.php
Digital Video Essential

John theres at least one "special test pattern" a (1-255) ramp that measures exactly 800mV and all the nominal patterns (black –white) measure exactly 0-700mV:p

I don't understand why I need all that expensive equipment when my cheap multimeter clearly proves that you have been correct. It is showing 800 mV, encoded right there on the disc.

It is showing 800 mV, encoded right there on the disc.LOL, that weak attempt at humor actually lacks any understanding of the difference between RMS and Peak - Peak amplitude. It seems you don't know what your 800 mV RMS reading represents:)

800 mV RMS ~ 2262.4 mV Peak - Peak

EDIT/
My post above #157 should have read 1-254 per 800mV P-P amplitude

That's a valid point, the readings should be lower. Any chance hwjohn's measurements were due to pedestal level?

That's a valid point, the readings should be lower. Any chance hwjohn's measurements were due to pedestal level?Do you use a multimeter to measure the P-P amplitude of waveforms to?

John's meter probes seem to have that fancy built in laser reader with the MPEG-2 decoder already embedded, followed by a calibrated D/A with a reconstruction filter as well? OK, I wonder how he managed to navigate the GOP sequence? Find a given I-Fame with an 8-bit digital value of "1" and the other probe touching the correlated (Y) digital 254 value?..all that considered, yes it seems his 800mV reading is spot on.

I can't believe you replied to that, deleted it and replied again. One thing I hated in grade school was the endless repetition of how too is a different word, but I suppose I get their point. Somewhere around there was also taught how sarcasm could be exampled by using a word, such as valid, completely out of context. Anyway, the new information is very helpful and I'm sure everyone will take it into account when measuring analog on DVD like hwjohn showed how.

If you're going to reply twice, then by all means make it a meaningful discussion and please fill me in on one thing I still can't figure out with how you seem to use 0 IRE to refer to the black level on each individual disk. How do you know where the level defined as black is on the DVE disk? You've seemed to state that black level isn't standardized and can be set wherever desired at the mere whim of the person doing the encoding, or at least that's the most sense I can make out of how you said gregr's post supported your position. So when DVE has a 1-254 pattern, how do you know where black is located? If you can talk about everything you've brought up regarding mere satire then this should be simple 101, but it's the same sort line of questioning I asked in regards to the THX pattern above and you passed on addressing.

I asked in regards to the THX pattern above and you passed on addressing. I’ve already explained i.e. that Y' luma along with Cb' or Cr' amplitude can be adjusted in either the analog or digital domain.

The original concept for the THX alignment patterns was for them to undergo the same production workflow as the content, so that in effect one would be calibrating to the same relative levels used by the studio and would in effect correlate for that specific encode.

You also failed to note the Tek SDI waveform monitor listed above. Since we humans live in an analog world the vertical graticule is calibrated in volts and measures P-P waveform amplitude. I personally use the following in my engineering shop:

http://www.tequipment.net/LeaderLV5750.html
Leader LV5750 Portable Multi SDI Monitor

Whether you like it or not, SD-SDI SMPTE 272M:

Blankink Level = 0mV = Black Level = 0 IRE = 0 %
White Level = 700mV = Reference White = 100 IRE = 100%

If THX or anyone else for that matter encodes a non standard test pattern, then those adjustments are embedded in the media. To suggest I can’t decode them and measure a given tweak on my scope is laughable!

http://www.jvbdigital.com/jvb.asp?cur=2&level=sdi&page=title&title=103
Pioneer Elite DVD-F07 SDI upgrade

http://www.testequipmentdepot.com/tektronix/video/waveformmonitors/digital/wfm5000.htm
Tektronix WFM5000 HD/SD-SDI Multi-STANDARD Waveform Monitor

..fill me in on one thing I still can't figure out with how you seem to use 0 IRE to refer to the black level on each individual disk.
http://www.tequipment.net/pdf/Hamlet/DigiGenIndex_manual.pdf
SERIAL DIGITAL BASICS

Page 23

The diagram Figure 5 is for 10-bit (64-940), but the Voltage (0-700mV) P-P amplitude is the same.

for 10-bit (64-940), but the Voltage (0-700mV) P-P amplitude is the same.

I'll take it that when you use 0 IRE you mean 16 from the 16-235 common 8-bit representation. With a typical US NTSC DVD player, digital 16 would be commonly refered to as 7.5 IRE as an analog output. To this point I'm still not sure you've ever explained why you consider it clear communication to refer to a digital level as 0 IRE when that same level could also be refered to as 7.5 IRE when output from a typical US NTSC analog output.


You also failed to note the Tek SDI waveform monitor listed above.

I really had no idea there was such a thing. The scopes I've used were all cheap and only accepted analog inputs. With the SDI input I'll figure it accepts digital signals and then represents them as an analog form. Of course analog can represent a digital input in a number of different ways, so I'll guess it must have controls to determine the representation. I suppose there are uses for such a feature, but for general discussion I consider it simpler to merely discuss the digital signal up until it's actually converted to analog by the playback chain. Many disk players allow for analog output levels to be altered, so discussing analog levels requires a clear definition of how the digital levels are represented with analog. Limiting discussion expressly to digital levels, until acutal conversion by the playback chain, just cuts down how many things have to be defined or implicitly understood.


Blankink Level = 0mV = Black Level = 0 IRE = 0 %
White Level = 700mV = Reference White = 100 IRE = 100%

This is not in line with a typical US NTSC analog output. The difference between this representation and US NTSC levels is how discussing analog levels without clear definitions of what you're talking about is confusing. Unless you expressly define how you're using an idealized conversion, from a digital signal, with these terms here, I really don't get how you expect someone to assume all that. Of course by the general questions that cross this forum, all those sorts of assumptions regarding analog levels isn't even relatively close to being typical.


To suggest I can’t decode them and measure a given tweak on my scope is laughable!

My comments, including the sarcasm, was generally around terms and definitions. The questions were around trying to figure out if you had any viable point. I really don't think a multi-thousand dollar scope is a necessary preresequite for where all this started with how IRE was said to supposedly be so "profound". It seems to me that most likely all the scope is doing is accepting a digital signal and representing it as an idealized analog level based on a selectable conversion. I certainly would try to avoid giving any false impression of that being actually measuring analog levels doing such a thing myself, but it just is what it is so I don't think I have a need for an exclimation point.

You know you remind me of Chris Wiggles. I mentioned to him my school has (2) Sony HDW-F500 HDCAM digital video tape recorders. I supplied him with a pdf of the operation manual. I noted the page that refers to the menu item that lets me adjust the SDI output setup level. Chris read the manual and commented ..and I quote:

"That is a misprint; you can't have setup in the digital domain"

Well that "misprint" reverse engineered itself into the machine.;)

how discussing analog levels without clear definitions of what you're talking about is confusing. .
http://www.visionaryforces.com/downloads/Sony_HDW-F500_HD_VTR.pdf
Item 712 (SETUP LEVEL)

“Adjust the level of the HD video signal output from the HD SDI connector. Adjust the setup level of the video signal.”

Why do you consider it clear communication to refer to a digital level as 0 IRE when that same level could also be refered to as 7.5 IRE when output from a typical US NTSC analog output?The definition for IRE has nothing to do with a NTSC DVD player adding 7.5% to the encoded YCbCr value.

IRE per definition is only correlated with the BLANKING LEVEL and reference white level. For the last time IRE does not correlate to a fixed NTSC analog voltage. Heres the standard SMPTE RP luminance reference signal voltage range:

Blanking level = Black level = 0mV for the weighted digital code 16 (8-bit video).

Reference White = 700mV = weighted digital code 235 (8-bit video)

Should SMPTE not use absolute voltages? Should Tektronix not attempt to measure digital video because your DVD player is adding 7.5%?

Actually, we all owe Thomas an apology. It looks like he was right all along.

http://www.avshd709.com/DiscVolts.jpg

Post of the year. :D:D:p:D

Even more hilarious is that Thomas ACTUALLY thinks this is how DVD's work. You can't even make stuff like this up!

IRE per definition is only correlated with the BLANKING LEVEL and reference white level...
IRE does not correlate to a fixed NTSC analog voltage

In an analog output isn't 0 mV defined as blanking regardless of pedestal? I thought you agreed with that here http://www.avsforum.com/avs-vb/showthread.php?p=14197630#post14197630 So if blanking is defined as 0 mV in an analog output and white is defined, then isn't IRE defined by your definition? If not, then why doesn't your IRE definition apply to a typical analog output with pedestal? I don't think anyone is refuting the first quoted part because it seems to be common in most any reference, but I can't find anything that supports the second part by how correlation is defined.

I tried to go back and see if I could find any prior coherent answer to such a basic conflict, and the only thing that stuck out was that you defined blanking as black level on DVD the same as above. With a pedestal in the analog output though, black level does not equal 0 mV as far as I can tell so the definition wouldn't apply. If instead just blanking is defined as equal to black level, then that too conflicts with the prior agreement linked in an analog output with pedestal. As far as I can tell your definition of IRE lines up perfectly with the typical mention of a 7.5 IRE black-level pedestal, but as far as I can tell for whatever undefined reason you seem to keep saying IRE does not apply in such a situation. I miss how the terms in the first quoted portion wouldn't apply to a DVD over an analog output with pedestal and how the second quoted portion would be true.

Even more hilarious is that Thomas ACTUALLY thinks this is how DVD's work. You can't even make stuff like this up!Hey Chris, did you contact Sony yet and tell them about an apparent "misprint" in their operation manual?.. so that you could explain to the Sony Corp that its NOT possible to "encode" setup in the digital domain. That setup only exist on the analog output of my DVD player. Maybe you could give them a link to your "Source Settings Guide" to clear up their confusion on the subject matter regarding the encoding digital video.

Thanks in advance!

I really don't think a multi-thousand dollar scope is a necessary..

It seems to me that most likely all the scope is doing is accepting a digital signal and representing it as an idealized analog level based on a selectable conversion.

I certainly would try to avoid giving any false impression of that being actually measuring analog levels doing such a thing myselfOK sure, the industry pros can't use waveform or vector scopes because that can't accurately represent the digital video signal. You do realize that every post production suite in the world references very similar devices in real time every day?

Oh and btw we are measuring Digital Video Levels and not "analog levels"

In an analog output isn't 0 mV defined as blanking regardless of pedestal?Yes, the blanking level is in fact the relative 0 IRE level. I will admit in my response to your question earlier regarding THX non standard test patterns, I did manage to transpose the encoded black level offset with the "0 IRE" blanking level. Still the +/- offset can be encoded and its not an ERROR.

Wether or not your DVD player adds 7.5% to the encoded signal does not change the accepted definition of IRE.

Oh and btw we are measuring Digital Video Levels and not "analog levels"

mV is a unit of analog measure. Without further context, I consider it entirely reasonable to assume that mV relates to an analog measurement. The only context where that doesn't apply is when talking ideals for an analog system or unless you were to specifically define your exact process as you've done here. To state that you're measuring mV from DVD without further context makes absolutly zero sense, and even with context it's still a bit non-sensical to say you're actually measuring mV because the scope is probably just doing an idealized relation. At least one thing I guess is commonly understood is that analog output voltage levels differ depending on connection type, and some players also allow for deviation from the defined levels, although unless a specific context is given then again all this is undefined and what is being stated is somewhat unclear.


the ENCODED pedestal level

I get what you're probably talking about here, but I think this differs from how for example Digital Video and HDTV Algorithms and Interfaces defines the term pedestal. If you're going to agree that it's improper use to take an analog term clipping and apply it in a general sense, then I'm unclear on why you consider it fine with the term pedestal.

mV is a unit of analog measure. Without further context, I consider it entirely reasonable to assume that mV relates to an analog measurement.Good luck with that! If you’re trying to imply when measuring a given SDI (digital stream) that the signal is analog. The scope display is of course analog though ;) Does not do me much good to watch decimal or hex numbers flash by in real time when I'm tweaking images.
..then I'm unclear on why you consider it fine with the term pedestal.I thought I would use the other accepted terminology black level offset, btw setup can in fact be (+/-) , I guess the later should be called setdown? The black level i.e. could be a - %.

For standard NTSC broadcast signals the pedestal is +7.5%.

If you’re trying to imply

What I said was very literal. Voltage is an analog unit of measure. A digital signal could be measured as voltage, but in any general discussion of digital signals voltage is somewhat irrelevant because variable voltage is not where the relevant information is contained in a digital signal. In a defined system there will be an ideal conversion of the digital information to an analog variable voltage, but even with a typical DVD player the different analog outputs have different ideal ways to represent the digital information. Strictly speaking if your scope receives a digital input, you are not actually sampling variable mV regardless of what the display shows. All the scope is likely doing is representing an ideal D/A conversion based on certain parameters. This is all so absolutely basic I have no idea how anyone with much introduction would be unaware that it could be somewhat confusing to discuss a digital medium such as DVD in terms of mV, unless the system in discussion is very well defined. All I was doing is pointing out that unless you clearly define what sort of conversion you're talking about when speaking of mV from a digital signal then it will only contribute to others not understanding what you mean for example like http://www.avsforum.com/avs-vb/showthread.php?p=13491136#post13491136 that you linked earlier.

A digital signal could be measured as voltage, but in any general discussion of digital signals voltage is somewhat irrelevant because variable voltage is not where the relevant information is contained in a digital signal.Well that is literally wrong!

Every single day engineers and artist are looking a digital video signals, we make adjustments like Black Level - Move Gamma Cross Point – Color Balance, these are “variable” tweaks. Are you suggesting these are irrelevant? ..which you defined as 0mv/700mV several times. You didn't say SMPTE YPbPr signals, you said "all calibrated YPbPr" signals.

..Betacam reference white is not 700 mV. MII black level is not at the blanking level in North AmericaHere Greg is reaching just like you, but his response has nothing to do with the subject of measuring the encoded (DVD) video signal levels that was being discussed in that thread.

Nominal DVD encoded "reference" Blk-Wht levels have NOTHING to do with Panasonic MII "standard" or Sony Betacam "standard" but for some odd reason they seem to align perfectly with SMPTE RP protocols?
This is all so absolutely basic I have no idea how anyone with much introduction would be unaware that it could be somewhat confusing to discuss a digital medium such as DVD in terms of mV.This organization- http://www.smpte.org/standards/ has no problem with such confusing terms like mV.

Have a nice day:)

Well that is literally wrong!

A digital signal is based around a high and a low voltage, that's what I said in the portion you quoted.


http://www.smpte.org/standards/ has no problem with such confusing terms like mV.

I figure they clearly define terms of any D/A conversion and understand that a range of mV values relates to an analog representation of information.

A digital signal is based around a high and a low voltage, that's what I said in the portion you quoted.Which part of your post address the subject matter of encoding and decoding YCbCr information?
I figure they clearly define terms of any D/A conversion and understand that a range of mV values relates to an analog representation of information.
I've clearly defined the absolute value those ones and zeros actually represent.

SMPTE seems to agree with me. NO?

The definition for IRE has nothing to do with a NTSC DVD player adding 7.5% to the encoded YCbCr value. IRE per definition is only correlated with the BLANKING LEVEL and reference white level.That's true.

For the last time IRE does not correlate to a fixed NTSC analog voltage.That would be true if you had simply said a "fixed analog voltage". However, NTSC video corresponds to a specific standard, where 140 IRE = 1 volt. Therefore, for NTSC video 100 IRE = 714.29 mV.

Heres the standard SMPTE RP luminance reference signal voltage range:

Blanking level = Black level = 0mV for the weighted digital code 16 (8-bit video).

Reference White = 700mV = weighted digital code 235 (8-bit video)

1. There are many different SMPTE standards, and your statements are NOT true for all SMPTE standards.
2. The Black Level is NOT the same as the Blanking Level in all SMPTE standards (and some professional formats).
3. The Reference White Level is NOT 700 mV in all SMPTE standards (and not 700 mV in some professional formats).
4. Most North American CONSUMER DVD players today output analog signals nominally in accordance with CEA standards, which are based on different SMPTE standards for different signal outputs. Many DVD players provide user controls that produce analog signals that deviate from those standards.
5. Prior to the CEA standards for consumer DVD players, consumer DVD players output analog signals nominally based on a variety of different standards or professional formats.
6. Some professional DVD players were designed to output analog signals based on a variety of SMPTE standards, CEA standards, and other professional formats depending on intended application.

Therefore, to say that DVD players output analog signals based on one SMPTE standard is wrong.

But why continue to discuss DVD player analog outputs when we now have consumer DVD players with HDMI outputs (and some players with SDI outputs)?

Some instruments (waveform monitors, video analyzers, oscilloscopes, etc.) represent digital signal levels as analog values, and/or analog signals/values as IRE values, based on only one SMPTE standard or professional format. That makes their representation correct only for signals intended for that standard or format. Some instruments include more than one standard or format for analyzing or displaying signals. Users have to be extremely careful to select the correct standard/format and/or only apply the results of the instrumentation to signals of the applicable standard/format or else they will obtain erroneous results, sometimes without realizing or understanding these errors.

That would be true if you had simply said a "fixed analog voltage".Fair enough, it’s what I've been trying to say and had originally posted four or five times already... “IRE are not simply volts” Most North American CONSUMER DVD players today output analog signals nominally in accordance with CEA standards, which are based on different SMPTE standards for different signal outputs.In this thread I’ve really tried to avoid the DVD player variable altogether. A given DVD player does not change the DVDs encoded YCbCr nominal reference range. I guess by accident the YCbCr digital payload being discussed appears for some odd reason to align with my waveform monitor mV graticule. Therefore, to say that DVD players output analog signals based on one SMPTE standard is wrong.I've only been trying to explain that a DVDs encoded Y’ for example 16-235 correlates exactly to a (0-700mV) A/D and D/A unity model and that also aligns with my SMPTE SDI (10-bit) infrastructure. NO?
But why continue to discuss DVD player analog outputs when we now have consumer DVD players with HDMI outputs (and some players with SDI outputs)?AGAIN I’m explaining what’s encoded on the DVD disc. You are reaching back to the hypothetical I gave you in the other thread, which btw I specifically said for the “ideal calibrated” D/A. But all this does not matter anyway I’ve already listed above a mod for your standard def DVD player which will bypass the D/A process altogether, the mod would read the digital data directly and provide me SMPTE compliant SDI stream. My Leader HD-SD SDI waveform monitor will indicate 0V for digital code level 16 and .7V for the digital code value of 235.

Also an amplitude of (1-254) will measure ~800mV. With both negative and positive relative level, no matter what Chris thinks of my test equipment.

BTW it’s just a coincidence that the SMPTE SDI P-P interface level spec is for 800mV. That is not a correlation with the encoded video signal levels.


Finally, Greg I do much appreciate the time!

it’s what I've been trying to say and had originally posted four or five times already... “IRE are not simply volts”

All that everone else has really said is that in the analog realm, with a defined standard or known situation, IRE corresponds to volts.


an amplitude of (1-254) will measure ~800mV

Like above, with a defined standard, a digital signal corresponds to analog levels. You're not actually measuring mV with your scope from a digital input. Like gregr said, "Some instruments... represent digital signal levels as analog values". The scope shows a corresponding analog level, to the digital input, as defined by the selected standard.

In both of the above situations, they're each just defined relationships based on a standard.


YCbCr digital payload being discussed appears for some odd reason to align with my waveform monitor mV graticule

YCrCb levels would generally correspond to analog levels through a chosen standard. The final analog representation depends on the standard selected. In a video playback chain the only place where analog levels actually arise from DVD is after a point of D/A conversion, and there are different standards for the conversion.


AGAIN I’m explaining what’s encoded on the DVD disc.

DVDs are encoded to digital levels. Volts are an analog unit. With a defined standard, analog units can relate to digital levels. If the standard is undefined, then discussing DVD in terms of analog levels makes no sense becuase there is not one single defined D/A conversion for DVD. Personally I consider discussing a digital medium in terms of what's actually encoded on disk to be less abstract, but if terms of conversion are defined then analog units can also relate to the stored digital information.

DVDs are encoded to digital levels. Volts are an analog unit. With a defined standard, analog units can relate to digital levels.I have no idea what standard you are talking about, but FWIW I'm defining an MPEG-2 YCbCr standard color space digital encode.

601 mapping specifies that video black (0mV) is assigned a value of 16, and video white (700mV) a value of 235. This allows for reasonable footroom and headroom in the signal. The Y'CbCr color space is a logical choice for modern broadcast, since MPEG-2 compression operates in this color space. 700mV is also the maximum excursion required for a digital-to-analog converter (DAC) to generate SMPTE-compliant component video.

Voltage excursion for 8-bit DVD/Blu-ray DIGITAL ENCODING

Y Luminance is defined as being between peak white, 700mV (235). and black, 0mV (16).
CbCr Chroma is defined as being between max positive, 350mV (240), and max negative, -350mV (16).

The values above and below these are termed out of gamut.

You're not actually measuring mV with your scope from a digital input. Like gregr said...All SD or HD (DVD/Blu-ray) productions employ engineers and color correction artist that reference similar if not the exact same wavefrom/vector monitors I use every day as well. These WF monitors have mV graticule markings that align with the video signal being encoded. They are calibrated to the same DIGITAL standards that I utilize.

My infrastructure is "calibrated” to the following Digital encoding standard:

HD-SD SDI SMPTE 259M, 272M, 299M.

EDIT/
Two standards, SMPTE 272M (for SD) and SMPTE 299M (for HD) define how audio is embedded into the ancillary space.

Y Luminance is defined as being between peak white, 700mV (235). and black, 0mV (16).

By the definitions in Digital Video and HDTV: Algorithms and Interfaces, 8-bit RGB of 235 is reference rather than peak white.

By the definitions in Digital Video and HDTV: Algorithms and Interfaces, 8-bit RGB of 235 is reference rather than peak white.I'll let dlarsen (Dave) explain to you the concept of gamut error and legal/valid excursion. I see you still have issues with the accepted definition of IRE though:)

http://www.tek.com/Measurement/App_Notes/Component_Puzzle/
“Eventually the IRE (later to be the IEEE) established a unit of measure for video signals. This “IRE unit’’ was defined as 1% of the video range from blanking to peak white, without reference to the actual signal voltage."

http://www.tek.com/Measurement/App_Notes/25_15609/25W_15609_0.pdf
Preventing Illegal Colors
"For example, R'G'B' systems have an upper gamut limit of 700 mV and a lower gamut limit of 0 mV. If any channel of an R'G'B' signal exceeds either the upper or lower limit, a gamut violation exists."

"Legal signals, however, can be invalid in colordifference formats like Y'P'bP'r. The allowed range for the luma signal is 0 mV to 700 mV, and for the color difference signals P'b and P'r, the range is +/-350 mV."

Why on the prior quote do you say "IRE per definition is only correlated with the BLANKING LEVEL and reference white level" and now you say "blanking to peak white"? There is a disconnect there as to how Digital Video and HDTV defines the terms. Going by the terms in that book, 8-bit RGB of 235 is reference white and 254 would be peak white. Commercial DVDs appear to contain specular highlights above reference white as Poynton has commented.

If you wish to expain the discrepancy in how you're using reference and peak, that's fine. Otherwise I'm all for letting this thread die. Based on your prior posting history I figure that even after gregr pointed out how your comment regarding "IRE does not correlate to a fixed NTSC analog voltage" was incorrect, you would probably still contend that was an entirely true statement.

A valid signal is one that is within color gamut and remains legal when translated into any other format. IEEE makes proper use of the adjective (peak) in their definiton of IRE. For this reason HDTV and SDTV systems should NOT exceed the 0-700mV nominal gamut.

Defining Legal and Valid Gamut (From Tek link above)
“The term gamut has been used to refer to the range or gamut of colors reproducible by a television system when a scene is illuminated by a reference white (6500-degree illuminant D for NTSC/PAL and High Definition Systems).”

“Because the values of the R'G'B' signals directly represent these colors, the term gamut is often used to refer to the voltage range of R'G'B' signals. R'G'B' signals extending outside the specified voltage range, or gamut, may produce desirable color values (albeit outside the system color gamut) on a given picture monitor, but may be clipped or compressed in subsequent signal processing, distorting the color when displayed on another picture monitor.”

601 mapping specifies that video black (0mV) is assigned a value of 16, and video white (700mV) a value of 235.Please quote the words in ITU Rec. 601 that specify video black as 0 mV and video white as 700 mV? They aren't there because Rec. 601 does not discuss analog voltage levels at all. In fact it says (section 2.2), "In the circumstances when the component signals are not normalized to a range of 1 to 0, for example, when converting from analogue component signals with unequal luminance and colour-difference amplitudes, an additional gain factor will be necessary ...". It clearly anticipates the use of multiple conversions, some of which don't conform to the one case you continue to cite (which is when Y, Pb, and Pr all have the same 700 mV nominal excursions). In any case, there is nothing about a 700 mV analog amplitude in Rec. 601.

Y Luminance is defined as being between peak white, 700mV (235). and black, 0mV (16).

By the definitions in Digital Video and HDTV: Algorithms and Interfaces, 8-bit RGB of 235 is reference rather than peak white.The reason for this confusion is because Rec. 601, an older standard, called 235 the peak white level. It also says (Table 2), "The signal level may occasionally excurse beyond level 235". Since it didn't make a lot of sense to say that the signal level could "excurse" beyond the peak level (and it caused confusion, often seen here on the forum), the terminology was changed/fixed in later standards, such as SMPTE 274 and ITU Rec. 709 (which are both HD standards). In SMPTE 274 and in Rec. 709, the corresponding 235 level is called reference white. (Perhaps tbrunet is stuck in the past? It sure seems a lot like "Groundhog Day" around here sometimes.)

Interestingly, SMPTE 170, which is an even older standard for NTSC composite analog video, also calls the corresponding white level the Reference White (defined as 714.29 mV for composite broadcast video), which really makes the "peak white" wording in Rec. 601 an unfortunate and poor choice of words (some say it was simply a terminology error that was never fixed).

Please quote the words in ITU Rec. 601 that specify video black as 0 mV and video white as 700 mV? .
http://broadcastengineering.com/mag/broadcasting_digital_video_basics/
Figure 2
"The relationship between the E’Y, E’CB and E’CR analog component signal levels corresponding to a 100/0/100/0 color bars signal as specified in ITU-R BT.601, SMPTE 274M and SMPTE 296M. Click here to see an enlarged diagram."

E'Y = 0-700mV (Blk-Wht)
E'CB = +/- 350mV
E'CR = +/- 350mV

In SMPTE 274 and in Rec. 709, the corresponding 235 level is called reference white.(Perhaps tbrunet is stuck in the past? ).http://broadcastengineering.com/mag/broadcasting_digital_video_basics/
Figure 2.
The relationship between the E’Y, E’CB and E’CR analog component signal levels corresponding to a 100/0/100/0 color bars signal and the 10-bit and 8-bit Y, CB and CR digital sample values, as specified in ITU-R BT.601, SMPTE 274M and SMPTE 296M. Click here to see an enlarged diagram.

235 = PEAK LEVEL
16 = BLANKING

Michael Robin, a fellow of the SMPTE and former engineer with the Canadian Broadcasting Corp.'s. He is co-author of Digital Television Fundamentals, published by McGraw-Hill, and recently translated into Chinese and Japanese.

Please quote the words in ITU Rec. 601 that specify video black as 0 mV and video white as 700 mV?http://broadcastengineering.com/mag/broadcasting_digital_video_basics/
Figure 2
"The relationship between the E’Y, E’CB and E’CR analog component signal levels corresponding to a 100/0/100/0 color bars signal as specified in ITU-R BT.601, SMPTE 274M and SMPTE 296M. Click here to see an enlarged diagram."

E'Y = 0-700mV (Blk-Wht)
E'CB = +/- 350mV
E'CR = +/- 350mV That's a magazine article, it's not Rec. 601. Quote the section of Rec. 601 that you claimed specifies peak (reference) white as 700 mV. It isn't there. It gives normalized signal levels (Table 3), i.e. 1.0 for E'Y, not 700 mV.

That's a magazine article,..Good luck with that observation.

“Signal Characteristics: All HDTV scanning formats originate as component analog gamma corrected E’ subscript R,G,B signals. These signals may be encoded into component analog E’Y, E’Cb and E’Cr. All analog signals have nominal p-p analog signal amplitude of 700mV corresponding to a 100/0/100/0 color bar signal.”

Hey Greg, I need to calibrate my analog waveform monitor so that I can adjust the black level for a given video frame that is to be encoded onto a standard def (ITU-R BT601) DVD release. Are you telling me I can’t do that because it “DIGITAL” and there are no volts envolved?

Can you then describe to me how I’ve calibrated my plant incorrectly? According to Michael Robin that books author, you do NOT know what your talking about.

Thanks in advance!

.. It gives normalized signal levels (Table 3), i.e. 1.0 for E'Y, not 700 mV.According to Figure 2 in that Broadcast Engineering Mag link above (note: click to enlarge icon), Michael Robin's math shows:

0mV = 16 for 8-bit (decimal)
700mv = 235 for 8-bit (decimal)

In SMPTE 274 and in Rec. 709, the corresponding 235 level is called reference white.(Perhaps tbrunet is stuck in the past? ).

http://broadcastengineering.com/mag/broadcasting_digital_video_basics/
Figure 2.
The relationship between the E’Y, E’CB and E’CR analog component signal levels corresponding to a 100/0/100/0 color bars signal and the 10-bit and 8-bit Y, CB and CR digital sample values, as specified in ITU-R BT.601, SMPTE 274M and SMPTE 296M. Click here to see an enlarged diagram.

235 = PEAK LEVEL
16 = BLANKING

Michael Robin, a fellow of the SMPTE and former engineer with the Canadian Broadcasting Corp.'s. He is co-author of Digital Television Fundamentals, published by McGraw-Hill, and recently translated into Chinese and Japanese.Again you quote magazine articles. The Y' component corresponding to 235 is not called the peak level in Rec. 709 or SMPTE 274, it is called reference white. See section 5.4 and 10.5 of SMPTE 274. See Table 5, item 5.1 of ITU-R BT.709 (i.e. Rec. 709).

You said you use SMPTE 274 in your work, so quit quoting magazine articles and give us sections of SMPTE 274 that use the term "peak white" rather than "reference white". (As I already said, the term "peak white" is used in Rec. 601.) And tell us what section of Rec. 601 defines peak white as 700 mV as you claimed (it doesn't - it defines no specific standard for analog signal levels - only normalized signal values with peak white = 1.0, but SMPTE 274 does because it defines a specific standard for digital and analog component video signals.)

When your claims are challenged, you say the standards support your claims. Then when I give you references in the standards you give us quotes from magazine articles.

“Signal Characteristics: All HDTV scanning formats originate as component analog gamma corrected E’ subscript R,G,B signals. These signals may be encoded into component analog E’Y, E’Cb and E’Cr. All analog signals have nominal p-p analog signal amplitude of 700mV corresponding to a 100/0/100/0 color bar signal.” More magazine quotes (talking about SMPTE 274). Quote the section of ITU Rec. 601 that you claimed said that peak white was 700 mV.

Hey Greg, I need to calibrate my analog waveform monitor so that I can adjust the black level for a given video frame that is to be encoded onto a standard def (ITU-R BT601) DVD release. Are you telling me I can’t do that because it “DIGITAL” and there are no volts envolved? Digital levels are not in volts. That's pretty damn elementary. A waveform monitor that accepts digital signals (i.e. SDI) can have a graticule that shows equivalent analog volts in any standard that relates digital levels to analog volts, and sometimes they include more than one standard. A waveform monitor may have multiple digital to analog electronic graticules, which show different analog levels for the same digital level.

According to Michael Robin that books author, you do NOT know what your talking about. Or you don't comprehend what he is talking about.

I've wasted enough time with this. No more replies from me.

More magazine quotes (talking about SMPTE 274). .That is a word for word quote from Michael Robins's published McGraw-Hill book.

http://books.google.com/books?id=BkCOd_d8_u0C&dq=Michael+Robin+Digital+Television+Fundamentals&pg=PP1&ots=JhT75QxWKI&sig=rNzA3ca09wU0Tcg9CLEeHXac0q4&hl=en&sa=X&oi=book_result&resnum=4&ct=result
Digital Television Fundamentals

Explains relevant SMPTE and CCIR-ITU standards

Digital levels are not in volts. That's pretty damn elementary. A waveform monitor that accepts digital signals (i.e. SDI) can have a graticule that shows equivalent analog volts in any standard that relates digital levels to analog volts, and sometimes they include more than one standard.I specifically asked you in the process of color correction or encoding a DVD master. For example:

Can professionals reference a calibrated analog waveform? Below I’ve defined in detail exactly an 8-bit (decimal) digital YCbCr component signal and how it correlates to a calibrated component waveform monitor per ITU-R BT.601 standard.

Y’ = 16-235 = 0V–700mV
CB, CR = 16-240 = +/-350mV

Greg the question IS rhetorical btw, not only is it possible but it’s done on a daily basis except the medium now is Blu-ray.

“Signal Characteristics: All HDTV scanning formats originate as component analog gamma corrected E’ subscript R,G,B signals. These signals may be encoded into component analog E’Y, E’Cb and E’Cr. All analog signals have nominal p-p analog signal amplitude of 700mV corresponding to a 100/0/100/0 color bar signal.”

That is a word for word quote from Michael Robins's published McGraw-Hill book.

http://books.google.com/books?id=BkCOd_d8_u0C&dq=Michael+Robin+Digital+Television+Fundamentals&pg=PP1&ots=JhT75QxWKI&sig=rNzA3ca09wU0Tcg9CLEeHXac0q4&hl=en&sa=X&oi=book_result&resnum=4&ct=result
Digital Television Fundamentals

Explains relevant SMPTE and CCIR-ITU standardsOK, that quote was from the book Digital Television Fundamentals, which I have. But that quote is about HDTV formats, it says "All HDTV scanning formats ...". You were writing about Rec. 601, an SD format, not an HD format. So that quote has nothing to do with what you originally wrongly claimed Rec. 601 said about 700 mV (white) and 0 mV (black). So once again, quit taking quotes from books and magazines out of context, and tell me where in Rec. 601 you claim it says the analog voltage level is 700 mV. If you can't do that, then admit it doesn't, and quit posting erroneous information.

Also look at the source you quoted, Digital Television Fundamentals, figure 2.83 on page 121 (my 1998 edition). It shows 4 different analog component video signal standards. One is NTSC related with 714 mV white level and 53.55 mV black level, another is EBU N10 with 700 mV white level and 0 mV black level. Another is Betacam with 714 mV white level and 53.55 mV black level, and another is M II with 700 mV white level and 53.55 mV black level. (The last black level is an error in the book, BTW, it is actually 52.5 mV, and the other black levels are actually 53.57 mV rather than 53.55 mV.)

So 3 of the 4 standards used as examples in the book you selected to support your claim, disprove your claim that reference white is ALWAYS 700 mV and black is always 0 mV. The book simply lists a few of the analog component video standards that prove you are wrong, as I have been telling you.

Now I have to move on and do other things. So unless you are willing to post information directly from the Rec. 601 standard, please stop posting on this topic. Quit posting information out of context from secondary sources that does not apply to your original claims about Rec. 601.

..the other black levels are actually 53.57 mV rather than 53.55 mV.)..I would NOT be soo quick to point out Michael's "errors" without first noting his obvious expertise. If you want to correct the more significant error btw, how about the table in Figure 2:)

•The ITU-R BT.601 component digital standard: The pervasive 4:2:2 format uses a wideband (limited to Fb=5.75MHz) luminance signal (E'Y) and two narrowband (limited to Fb=2.75MHz) amplitude-scaled blue color-difference (E'CB) and red color-difference (E'CR) signals.

http://broadcastengineering.com/mag/broadcasting_digital_video_basics/
Figure 2. The relationship between the E’Y, E’CB and E’CR analog component signal levels corresponding to a 100/0/100/0 color bars signal and the 10-bit and 8-bit Y, CB and CR digital sample values, as specified in ITU-R BT.601.
Greg if you ever manage to click on the icon to enlarge the diagram (Fig. 2) provided by Michael, you will note the 8-bit (decimal) code value 16 and 235 respectively appear to correlate exactly to a 700 mV p-p waveform..NO? The same seems true of the 16 and 240 codes for CB & CR?


Send questions and comments to: michael_robin@primediabusiness.com

Quit posting information out of context from secondary sources that does not apply to your original claims about Rec. 601.Paaalease!

The Broadcast Engineering Mag article is NOT “out of context” and it’s now obvious to everyone Michael Robin knows his AC from his DC:p

ITU-R BT.601 (Y’) component has the following p-p analog values for 8-bit Video

1 = -47.9 mV
16 = 0 V
235 = 700 mV
254 = 763.13 mV

ITU-R BT.601 (Y’) component has the following p-p analog values for 8-bit Video

1 = -47.9 mV
16 = 0 V
235 = 700 mV
254 = 763.13 mV

gregr's comments appear rather clear to me. He appears to be saying this is acutally defined by SMPTE 274 rather than Rec 601. Looking at "Definition of the digital signals Y, CR, CB, from the primary (analogue) signals" from http://inst.eecs.berkeley.edu/~cs150/Documents/ITU601.PDF it appears that Rec 601 uses analog levels of 0v for black and 1v for white like he said with terms to adjust between different systems. I really have no interest in reading SMPTE 274, but it seems clear that Rec 601 doesn't use the quoted terms like you've said.


it didn't make a lot of sense to say that the signal level could "excurse" beyond the peak level

Thanks for the clarification. If nothing else I found your comments interesting.

. I really have no interest in reading SMPTE 274, but it seems clear that Rec 601 doesn't use the quoted terms like you've said.I have to ask the question if you don't mind..

Did you even look at the diagram for Fig. 2 that I posted ten times already..at all?

Can you please report what value Michael Robin suggest (author Broadcast Engineering article) an 8-bit Y' (decimal) code value 235 represents i.e.

Is it 700 mV?

Did you even look at the diagram for Fig. 2 that I posted ten times already..at all?

Have you read the references for that item? It clearly says "as specified in ITU-R BT.601, SMPTE 274M and SMPTE 296M." Rec 601 is not where those voltages you keep refering to are defined. Rec 601 references 0 to 1v for black to white with the formulas it uses. So have you even read Rec 601 that I linked?

http://www.testequipmentdepot.com/tektronix/video/waveformmonitors/digital/wfm5000.htm
WFM5000
Supports HD-SDI (SMPTE 292M) and SD-SDI (ITU-R BT.601) monitoring applications.

If the machine above does not indicate 0V and .7V respectively for a nominal Y' Blk-Wht signal i.e. (16-235) 8-bit Rec. 601 video.


"it's broke"!

So have you even read Rec 601 that I linked?Yep and I'll play your game. Please provide me a link to one Y'Cb'Cr' component waveform monitor used by industry pros like me, that will indicate Blk-Wht (Y') signal that in Volts i.e. a 1V p-p waveform representing a standard reference color bar signal.

Yep and I'll play your game.

Well, I wasn't playing any game. I wanted to see what exactly is defined by Rec 601, so I looked at the document I linked. It appears to define the digital levels and it defines "primary analogue signals" that correspond with the equations it gives. Rec 601 does not appear to define typical component video, which is the key of your comments. If you've read Rec 601 then it seems to me you should either know that component video is not used in the definition of the standard, or else if that's not the case then maybe you should reference where in the standard it's defined. The component voltages you keep bringing up have to be defined somewhere, but my take is that it's not in Rec 601 like your general comments would suggest. I have no issue with how your meter can have a relationship between analog voltages and digital signals, but I'm not so sure by your communication that you understand from where such a correlation arises.

I have no issue with how your meter can have a relationship between analog voltages and digital signals, but I'm not so sure by your communication that you understand from where such a correlation arises.

http://broadcastengineering.com/mag/broadcasting_digital_video_basics/
The mathematical expressions defining the Rec. 601 & Rec. 709 signal is given in Table 1. In both standards, the color-difference scaling factors were chosen to ensure that the signal amplitudes for a 100/0/100/0 color bars signal equal 0.7 V p-p.

Whether you or Greg understand the process of (DVD, SDTV, or HDTV) content mastering or not, believe me they are ALL looking at an analog waveform, be it RGB or YCbCr. For 8-bit or 10-bit video the nominal reference range is 0-.7V

The production artist/engineer doing color correction needs real time confidence as he/she adjusts the image. If the WF indicates the Black level i.e. is offset +/- from (0V) the blanking level, it's really an accurate measurement.

ITU-R BT.601 (Y’) component has the following p-p analog values for 8-bit Video

1 = -47.9 mV
16 = 0 V
235 = 700 mV
254 = 763.13 mV

gregr's comments appear rather clear to me. He appears to be saying this is acutally defined by SMPTE 274 rather than Rec 601. Looking at "Definition of the digital signals Y, CR, CB, from the primary (analogue) signals" from http://inst.eecs.berkeley.edu/~cs150/Documents/ITU601.PDF it appears that Rec 601 uses analog levels of 0v for black and 1v for white like he said with terms to adjust between different systems. I really have no interest in reading SMPTE 274, but it seems clear that Rec 601 doesn't use the quoted terms like you've said. alluringreality, that is exactly correct. Now that you have provided an on-line link to Rec. 601, there is no excuse for anyone to continue to post erroneous information about what is contained in Rec. 601.

Greg thanks for quoting (Table in Figure 2) Michael Robin like I requested. Seems you now use the adjective "erroneous" to describe his mathematical prowess. Greg whats apparent IS you don't actually comprehend how a Multi-Standard digital waveform monitor aligns with a 0-700mV graticule.

According to Michael I’m spot on about the 700mV p-p amplitude correlation between nominal (16-235) Y' for ITU-R BT.601 8-bit video. NO? Please quote the section where I'm misrepresenting the BE Mag article!

Can you suggest just one manufacturer of Digital waveform monitors that will NOT show a RGB or Y,CB,CR industry standard color bar signal that’s not 700mV p-p.

Thanks in advance.

Do you really not understand the major conflict here? You said:

601 mapping specifies that video black (0mV) is assigned a value of 16, and video white (700mV) a value of 235

That statement is flat-out false because of the word specifies and the mV terms. Rec 601 does not use those mV terms. If you take out the mV items the statement is true, with the understanding that you're talking about digital decimal values. Not everyone in a forum titled "display calibration" only deals with the one specific context you seemingly believe applies to every context.

If you take out the mV items the statement is true, with the understanding that you're talking about digital decimal values. I’ve already thrown down my gauntlet above (Post 211). I’ve specified “decimal" probably ten times already.. so give that up. I won’t even waste my time and quote myself.. the un-edited evidence is documented in the thread.

Alluringreality,

The now infamous (Figure 2) in the Broadcast Engineering article, Michael Robin clearly notes DECMIAL & HEX. There’s no getting around both mathematical expressions which validate the fact (SD) ITU-R BT.601 or (HD) ITU-R BT.709

"defines the 8-bit Y' decimal value 235 = 700mV"

the fact (SD) ITU-R BT.601 or (HD) ITU-R BT.709

"defines the 8-bit Y' decimal value 235 = 700mV"

As far as I know the ' refers to gamma adjustment, which doesn't apply to digital values in those standards. I really have no idea how it's so difficult to get that the quoted portion is false due to the inclusion of the mV term and lack of definition for that term within the quoted standards. Yes it's possible to define how 700mV can be equal to digital 235, but Rec 601 doesn't do that.


I won’t even waste my time and quote myself..

Outstanding, quoting tbrunet is something we can definitely agree on.

Yes it's possible to define how 700mV can be equal to digital 235, but Rec 601 doesn't do that.This does and the ITU-R BT.601 standard does as well:http://broadcastengineering.com/mag/broadcasting_digital_video_basics/

"The mathematical expressions defining the ITU-R 601 & ITU-R 709 signal is given in Table 1. In both standards, the color-difference scaling factors were chosen to ensure that the signal amplitudes for a 100/0/100/0 color bars signal equal 0.7 V p-p."

Can either you are Greg go on the record and state one way or the other that Michael Robin the author of Digital Video Fundamentals is WRONG?

i.e. rather standard color bars (mV amplitude) should actually measure 1V p-p on my Digital Multi standard component Y,CB,CR or RGB WAVEFORM MONITOR?

Send questions and comments to: michael_robin@primediabusiness.com

Can either you are Greg go on the record and state one way or the other that Michael Robin the author of Digital Video Fundamentals is WRONG?

I'll state that I have not and don't plan on commenting on Michael Robin's article. My comments were around your representation of Rec 601. The major item is that Rec 601 is generally based around gamma corrected signals from black to white of 0 to 1, which is relevant for example in its definition for quantization of Y from E'y. Certainly there are ways to get around this item and use different analog levels, normalizing component levels is one example, but characterizing Rec 601 as using 0 to 0.7V is just misleading. Anway, gregr said basically the same thing here http://www.avsforum.com/avs-vb/showthread.php?p=14287905#post14287905 For what it's worth, the article does say "normalized (700mV p-p)" a few times.

It’s also apparent you don’t comprehend that the (') symbol represents gamma corrected (E) voltage. Your Berkeley ITU Rec 601 pdf uses the same (coefficients) mathematical expression to define ITU-R BT.601 thats in the BE Mag article..
http://inst.eecs.berkeley.edu/~cs150/Documents/ITU601.PDF
:rolleyes:
2.4 Construction of Y,CR ,CB via quantization of , E’R, E’G, E’B

http://broadcastengineering.com/mag/broadcasting_digital_video_basics/
:rolleyes:
And Michael’s Table 1
Component digital signal characteristics of the ITU-R BT.601 and ITU-R BT.709 standards. Click here to see an enlarged diagram.
Is the same expression and coefficients look it up for yourself!

More specifically, your Berkeley ITU link Page 6
Definition of the digital signals Y,CR,CB, from the primary (analogue) signals E'R , E'G and E'B
As far as I know the ' refers to gamma adjustment, which doesn't apply to digital values in those standards..

..but characterizing Rec 601 as using 0 to 0.7V is just misleading
Please explain these statements?

It’s also apparent you don’t comprehend the that a prime symbol correlates with gamma corrected (E) voltage.

I have no idea what this means or how it helps to advance productive discussion. It seems to me that in "8-bit Y' decimal value 235" the prime shouldn't be there, that's the point of what I said there.


Your Berkeley ITU Rec 601 pdf uses the same (coefficients) mathematical expression to define the ITU-R BT.601 thats in the BE Mag article..
Michael’s Table 1

the same expression and coefficients look it up for yourself!

This is correct. That's why it's important to understand the range that Rec 601 works with. The main key of what I commented previously is normilization, which is one way for how the 0-0.7V signal can fit with the 0 to 1 range Rec 601 uses. If you were to directly use the 0-0.7V range then by section 2.3 Quantization the digital values wouldn't fit the intended range. For example if you plug 0.7 as white without normilization it results in a digital value of 169. Instead if you normailze a 0-0.7V signal you get 0-1 which then puts white at 235 by the quantization equation. Of course there are other ways to get the output to fit with the input, but this is probably the acutal method used considering his use of the term normalized.

I have no idea what this means or how it helps to advance productive discussion. .

..but characterizing Rec 601 as using 0 to 0.7V is just misleading.Oh please! It should have read..

"its apparent you don't comprehend the (') prime symbol represents gamma corrected (E) voltage"

1)Which SDTV or DVD or HDTV standard (8-bit) 100/0/100/0 color bar signal is not coded as 16-235?

2)Which 100/0/100/0 color bar signal will NOT measure as .7V p-p on 100% of the industry standard component RGB or YCBCR waveform monitors?

Have you ever heard of a component waveform monitor ever measuring a standard reference signal.. and have it indicate 1V p-p?

Have you ever heard of a component waveform monitor ever measuring a standard reference signal.. and have it indicate 1V p-p?

Rec 601 doesn't necessarily refer to a 1V p-p like some of my earlier comments, but gregr had it right. While Rec 601 does say "e.g. 1.0 V maximum levels" at one point, generally it avoids any scale and is speaking of normalized values. In the case of a 0-0.7V black to white signal, I would think a normalized white would be 0.7V/0.7V=1. Based on reading Rec 601 and how Michael Robin's Table 1 matches Rec 601, I would have to guess that it's intended to be implicitly understood the values for Table 1 are normalized rather than actual 0-700mV considering how Rec 601 defines equations for quantization using those same terms.

..but gregr had it right..I beg to differ on many levels i.e. I asked Greg explicitly from the very beginning, probably ten times in three different threads.. the same rhetorical question:

Greg I need to color correct a given standard DVD production to be released. I would like to measure the YCBCR video black level. Are you saying I can’t because it’s digital and volts are not involved?
Digital levels are not in volts. That's pretty damn elementary.

..The book (co-authored by Michael Robin) simply lists a few of the analog component video standards that prove you are wrong, as I have been telling you.

Quote: Michael Robin
“Digital video is best defined as a means of describing the continuous analog video waveform as a stream of digital numbers.”

Now I have to move on and do other things. Unless you are willing to post information that can directly disprove my position that all NTSC/PAL DVD, SDTV, HDTV (ITU-R 601 or R 709) 8-bit or 10-bit digital video, with nominal digital excursion Blanking - Peak Level actually represents a YCBCR/RGB .7V p-p analog waveform.

disprove my position that all NTSC/PAL DVD, SDTV, HDTV (ITU-R 601 or R 709) 8-bit or 10-bit digital video, with nominal digital excursion Blanking - Peak Level actually represents a YCBCR/RGB .7V p-p analog waveform.

All you have to do is to look at Rec 601 for example to disprove such a statement; the 0-0.7V p-p you use is not the one and only available relationship. I could say the digital information represents 0-1V p-p with just as much validity. The reason for such a claim is that Rec 601 defines a relationship between normalized signals and the digital information. It's possible to take the normalized signal and equate it to any number of different analog voltage levels. It's possible to also relate the normalized levels to the 0-0.7V p-p you use, but that is not the only available relationship. What I'm saying is explicitly defined in Rec 601, and nowhere does it refer to 700mV.

“Digital video is best defined as a means of describing the continuous analog video waveform as a stream of digital numbers.”

This is fine, but Rec 601 describes a normalized signal as a stream of digital numbers, rather than explicitly 0-0.7V p-p in relation to digital values. Because the standard uses normalized levels, it can relate to any number of different actual p-p levels. All of this has already been discussed in different ways, and yet somehow you still seem to contend that the one relationship your scope uses applies to all situations. Yes it is possible to define a step by step relationship for how 0-0.7V could relate to the digital information, but Rec 601 by itself does not define any such an explicit relationship of how only 0-0.7V would relate to 16-235. Please do not again refer me to Michael Robin's Figure 2, because it seems clear enough that you don't understand how it is that relationship comes about and that it doesn't apply to all situations.

All you have to do is to look at Rec 601 for example to disprove such a statement; the 0-0.7V p-p you use is not the one and only available relationship..It IS whats being encoded and NOT A 1V p-p video waveform.. please show me a link that depicts said 1V p-p defined event. Should be easy if it actually exist.

I've shown you the standard ITU-R BT.601 or ITU-R BT.709. With the accepted nominal .7V p-p waveform which correlates to the original master :) No sync allowed!

This is so silly, but here's an example of normilization of two different values.

350mV from a 0-700mV signal would be 350mV/700mV=0.5
500mV from a 0-1000mV signal would be 500mV/1000mV=0.5

In the case of the 350mV or the 500mV, the relevant information for Rec 601 (0.5) is exactly the same.

If you don't mind I would like a valid reference, if it exist there is a pic of said waveform. Also an explanation of how you truncated the 188.97mV extra excursion beyond 1-254, yet appear to recover it magically?

According to “Digital Video Fundamentals":)

Nominal luma (Y') excursion is:

254 (decimal) +763.13mV
1 (decimal) -47.9mv

Total = 811.03mV

Please tell me how you can “encode” 1000mV in the ITU-R 8-bit payload?

If you don't mind I would like a a valid reference

I've already given my reference for discussion http://inst.eecs.berkeley.edu/~cs150/Documents/ITU601.PDF


254 (decimal) +763.13mV
1 (decimal) -47.9mv?

There is no such universal definition like this quote that applies to every situation, which is the whole point of contention. The quoted statement is true in certain situations, but to say that it is the one and only relationship is false. In fact Rec 601 expects so many relationships that in the "Definition of the digital signals Y, CR, CB, from the primary (analogue) signals" section it even says "The method is given as an example, and in practice other methods of construction from these primary signals or other analogue or digital signals may produce identical results." So not only might Rec 601 be used for analog signals, but right there it says that it could be used also with a digital signal. Look, the long and short of all of this misunderstanding is that there is no universally fixed relationship such as the above quote based wholly on Rec 601 in the case of DVD.


Please tell me how you can “encode” 1000mV in the ITU-R 8-bit payload?

Do you understand the 5th or 6th grade math example of normilization? The result of the division, the normalized level, is what the quantization equations in Rec 601 are based around. The quantization equation examples are not based on a fixed mV signal.

There is no such universal definition like this quote that applies to every situation, which is the whole point of contention. .alluringreality,

I'm going to break this down for you, ala 5th grade math:)

Let say we agree on a standard like the Rec 601 standard. This time we will call it SMPTE RP155. It's an audio standard:

-20dBFS = +4dBu = 0VU = 1.23V (RMS)

Please explain how one can encode i.e. 14 Volts (RMS)?..when

0dBFS = 12.28 Volts (RMS);)

Let say we agree on a standard like the Rec 601 standard. This time we will call it SMPTE RP155.

The suggestion of using a comparison is only another level of abstraction. The reason I am wholly focused on Rec 601 is to limit abstraction and try to be as specific as possible. I want you to focus on what exactly is stated, rather than what you believe is stated. That way maybe you can see that Rec 601 does not explicitly define for digital 235 being one certain peak white voltage.

-20dBFS = +4dBu = 0VU = 1.23V (RMS)

This is a good example of one of my points, Rec 601 generally does not use terms like V. Insead Rec 601 defines quantization of a normalized input to digital levels. Because it works with normalized levels, there is no reference to specific V levels.

...there is no reference to specific V levels.Bull

Using an agreed upon “standard” to digitally sample a given ANALOG WAVEFORM, the amplitude is defined by said mathematical expression.

If one can measure it in the analog domain, then inturn when it’s decoded using the SAME Y'CBCR math. the AMPLITUDE is the same. The digital part is transparent.

Capish

Can I do something to make that reading higher???

No, you've already stated the representation your scope uses. Unless your scope allows for different settings for how to represent the digital information, then you'll continue to look at the information the way your scope represents it. To say that what your scope shows is somehow universal and a particular digital value always equals a specific intended mV reading regardless of all other conditions, that's what I am diagreeing with. Just because your scope represents digital information a certain way, in general it's false to take that representation and to make a universal claim that somehow digital 235 is always supposed to be equal to 700mV and digital 16 is 0mV regardless of any other factors.

Here's a link:

http://www.broadcast.harris.com/product_portfolio/product_details.asp?sku=VSM_3901

The product can be configured for the different analog component standards. Look at the specifications at the bottom of the page.

Y Signal Level 700 mV or 714 mV (switchable, +/- 1 dB adjustable)
Pb, Pr Signal Level 934 mV (Betacam) or 700 mV (SMPTE) or 648 mV (MII)

Ron

Here's a link:

RonIs this an example of ITU-R BT.601.


This is my 25+ year career that we are talking about:)

I don't care if your DVD player can't match exactly what been encoded. I'm not only telling alluringreality and Greg, but yourself included for the last time. My SCOPE (aka WF monitor) is CALIBRATED! I am using the defined ITU-R standard RGB/YCBCR mode of operation. Every time I call up a standard 100/0/100/0 color bar signal. It will measure .7V p-p or else the encoded signal is NOT nominal 16-235.

What do you think Ron, is my WF monitor calibrated?;)

Unless you are willing to post information that can directly disprove my position that all NTSC/PAL DVD, SDTV, HDTV (ITU-R 601 or R 709) 8-bit or 10-bit digital video, with nominal digital excursion Blanking - Peak Level actually represents a YCBCR/RGB .7V p-p analog waveform.I'm sure you know Harris Broadcast Communications is a world leader in SD/HD Broadcast equipment. Here is a link (http://www.broadcast.harris.com/product_portfolio/product_details.asp?sku=VSM_3901) to the Harris VSM-3901 which converts 10-bit SDI digital video to analog YPbPr video. See the specifications on that page. It produces Betacam, SMPTE, or MII analog YPbPr output levels, just as I have told you (and shown you on page 121 of the very book that you cited as your reference). It converts SDI digital video to Y signals with 700 mV OR 714 mV levels. Note also that the Pb, Pr signal levels are selectable (934 mV, 700 mV, or 648 mV) for Betacam, SMPTE, or MII standards. There are many other similar professional/broadcast products that disprove your position that Rec. 601 digital video only represents 700 mV YCbCr levels. As I've also told you in the past, some DVD players were built to output Betacam levels (714 mV).

You have repeatedly said that Rec. 601 specifies 700mV analog YPbPr levels. You were given a link to a pdf of Rec. 601. You can not cite anywhere in Rec. 601 that 700mV is even mentioned. Yet you continue to post this assertion even when you now have Rec. 601 to read yourself.

There are people that come to the AVS Forum to learn about video. There are particularly people that come to this section, the Display Calibration section, that want/need to learn accurate information about video. The AVS Forum loses the great value it has as an information resource and a learning tool if someone repeatedly posts incorrect information. If I were a moderator I would have deleted this thread long ago.

Here's a link:

http://www.broadcast.harris.com/product_portfolio/product_details.asp?sku=VSM_3901 ...

RonRon, I was typing and posted my reference to the VSM_3901 before I saw your post about the VSM_3901. Amazing. I didn't mean to repeat your example.

Greg Rogers is going on the record to say I can't measure the black level that I'm encoding onto a DVD because it's DIGITAL. He must first tell me what format it is going to be output in to determine what I'm doing to the signal?:)

MY SD-HD SDI Multi Standard waveform monitor is not calibrated to measure the ITU-R BT.601 video signal.


BTW Harris Broadcast Communications, I've had 120U transmitter training at their Quincy headquarters.

Ron, I was typing and posted my reference to the VSM_3901 before I saw your post about the VSM_3901. Amazing. I didn't mean to repeat your example.
It's actually pretty cool. Great minds think alike!

Ron

Greg Rogers is going on the record to say I can't measure the black level that I'm encoding onto a DVD because it's DIGITAL. He must first tell me what format it is going to be output in to determine what I'm doing to the signal?:)That is not exactly what I said, so if you use my name provide exact quotes of what I said (in context). Nevertheless, it is true that you can't take the analog output of DVD player and determine what the digital signal levels on the DVD are unless you specify the standard used to convert from digital to analog output. That is obvious. You will obviously get different analog output levels from a DVD player that outputs Betacam levels or MII levels, rather than CEA 770.2 levels. Rec. 601 does not specify a specific digital to analog conversion.

That is not exactly what I said, so if you use my name provide exact quotes of what I said (in context). Nevertheless, it is true that you can't take the analog output of DVD player and determine what the digital signal levels on the DVD are...Greg I need to color correct a given standard DVD production to be released. I would like to measure the YCBCR encoded (video) black level. Are you saying I can’t because it’s digital and volts are not involved?
Digital levels are not in volts.That's pretty damn elementary.

..The book simply lists a few of the analog component video standards that prove you are wrong, as I have been telling you.

Then you went in detail to explain the Panasoinc MII “standard” or Sony Betacam “standard’. What do “they” have to do with my job of mastering DVD content i.e. the blanking standard or how my video image relative black level and white level are measured?

That is not exactly what I said, so if you use my name provide exact quotes of what I said (in context). Nevertheless, it is true that you can't take the analog output of DVD player and determine what the digital signal levels on the DVD are...
That is a perfect example of completely distorting what I said by quoting part of a sentence without the complete context of the sentence. You left off the rest of the sentence ... "unless you specify the standard used to convert from digital to analog output." Do you comprehend the crucial meaning of the word "unless"? That as they say, was the last straw. I will no longer respond to your posts. If you distort what I have said I will refer those posts to the moderator.

The Leader LV-5760 is doing it's own D/A conversion of the incoming SDI data. The designers have chosen to represent this pseudo waveform (there is no "real" blanking in SDI) on a .7V scale which corresponds to a particular analog interface (EIA-770.2).

The point that Greg and alluringreality are trying to make is that .7V is entirely arbitrary. It could have been .714V if the designers of the Leader LV-5760 had considered Betacam levels to be more useful than EIA-770.2 levels.

Ron

Lot's of data in the blanking.

http://www.w6rz.net/blanking.png

Ron

The Leader LV-5760 is doing it's own D/A conversion of the incoming SDI data.

It could have been .714V if the designers of the Leader LV-5760 had considered Betacam levels to be more useful than EIA-770.2 levels.
LV- 5750 Multi SDI Monitor

BTW Ron I've been referencing the industry standard "unity" model for looking at the data. The only thing analog about my monitor is the calibrated ITU D/A. The component matrix (RGB/YCBCR) and graticule is normalized for ITU-R BT.601 & Rec 709 standards. It displays E’Y, E’CB and E’CR analog component signal levels corresponding to a 100/0/100/0 industry standard color bar signal.

Thers no "real" blanking in SDI

..Lot's of data in the blanking.According to my reference the Blanking Level is clearly defined and it's NOT any 8-bit decimal code word between (1-254) like you seem to have implied;)

http://en.wikipedia.org/wiki/Serial_Digital_Interface
For SDI portions of the vertical and horizontal blanking regions which are NOT used for ancillary data, it is recommended that the luma samples be assigned the code word 16 (decimal), and the chroma samples be assigned 128 per 8-bit payload; both of which correspond to 0 mV.In other words Michael Robin is correct again in that the Blanking level for ITU-R 601 or 709 is defined as the YCbCr code words for 0 mV.

http://broadcastengineering.com/mag/broadcasting_color_bars_puzzle/
“Color bars signals are identified with four numbers, representing percentages of the maximum value of 700 mV (100 percent), with an oblique stroke between them as follows:”

“These signals may be used in their original form (component analog GBR signals), matrixed into E´Y, E´B-E´Y and E´R-E´Y (scaled or non-scaled), or encoded into an analog (PAL, NTSC or SECAM) or digital (component or composite) signal.”

http://books.google.com/books?id=BkCOd_d8_u0C&pg=PA507&lpg=PA507&dq=michael+robin+ITU-R+BT.601+100/0/100/0&source=web&ots=JhT98SxVHI&sig=rUVg1n3Qjf51u_poyYhDxdEPOfI&hl=en&sa=X&oi=book_result&resnum=3&ct=result
(Page 508)

“Table 11.9 shows the calculated signal amplitudes for individual steps corresponding to the three primary colors (green, blue and red) and their complements (yellow, cyan and magenta) of the 100/0/100/0 color bars signal with reference to (Table 11.8) ITU-R BT.601 Coded Signal Standard and the component signal “levels in mV” for ITU-R BT.601 E’Y, E’CB, E’CR components”

Note Table 11.9 on page 508,
(0-700mV) Black-White 100% Color Bar Component Signal Level in “mV” for ITU-R 601 and 709.:)

To say that what your scope shows is somehow universal and a particular digital value always equals a specific intended mV reading regardless of all other conditions, that's what I am diagreeing with.alluringreality do you really think Michael Robin just made up the "as specified in the ITU-R recommendation" part?

http://broadcastengineering.com/mag/broadcasting_color_bars_puzzle/

"Their (color-difference signals) amplitudes are expressed in percentages of the full amplitude primary signals (700 mV). These signal amplitudes are reduced by scaling factors to meet specific signal amplitude range requirements. When the allowed amplitude range of these signals is 700 mV (±350 mV), as specified in the EBU N10 Standard and the ITU-R BT 601 recommendation, the color-difference signals are given by the following expressions:

E´CB = 0.564 (E´B - E´Y), also known as PB in North America
E´CR = 0.713 (E´R - E´Y), also known as PR in North America"

alluringreality do you really think Michael Robin just made up the "as specified in the ITU-R recommendation" part?

What I think has little to do with Michael Robin's article and mainly centers on how it seems to me that you're willing to extend his statements beyond the intended context. I've already linked Rec 601, so what Rec 601 exactly says can be looked at by anyone interested. I'll say that it does not reference specific voltage levels in regards to certain digital values as you seem to claim, and US NTSC levels would be one example of where digital level 16 is not intended to equal 0mV. I question if you are primarily interested in factual accuracy, and I will merely continue to use discussions I find interesting in order to try to further my understanding. In this case, that motivation has mostly run its course.

I'll say that it does not reference specific voltage levels in regards to certain digital values as you seem to claim.
And your limited understanding of YCbCr normalization and or calibrating A/D - D/A unity is wrong. Below I've provided an ITU-R Recommendation that supports everything I've quoted Michael Robin saying (myself included) as "fact".

Testing for conformity with ITU-R Recommendations BT.601
http://www.ebu.ch/CMSimages/en/tec_text_i15-1998_tcm6-4775.pdf

ADC and DAC parameters

3.1.2 Coding ranges
Low-frequency effects can be observed by applying a 700 mV bar-on/bar-off test signal to the ADC and checking that the error is less than one quantum level in both the bar-on and bar-off conditions.
:rolleyes:
3.1.8. Response of the ADC to out-of-range signals
In the luminance channel, signal levels going more positive than 760 mV must result in the coder output remaining at level FE (decimal 254). Similarly, for input signals going more negative than -48 mV the output must remain at level 01. Levels FF and 00 are reserved for synchronization purposes.:)

--------
Michael Robin's provocative Figure 2 has perfect alignment or “conformity” with these ITU-R Recommendations. Sorry but Ron, Greg, hwjohn are wrong regarding YCbCr A/D – D/A voltage unity.

Your reference introduces another standard so I really fail to see how it speaks specifically about Rec 601. In the second paragraph it also says "The Recommendation BT.601 is not sufficient to specify fully the performance of digital video equipment." If Rec 601 really says what you continue to claim, then it seems to me only Rec 601 really needs to be discussed here. I really fail to see how US NTSC levels support your claim that specific voltages are universally related to specific digital values as defined by Rec 601. If you want to continue your argument it seems to me you really ought to address something simple like the last sentence in reference only to what is stated in Rec 601. My prior question, that was lost in the server crash, asking the value of E'Y for white for the analog levels you use in relation to the quantization section from Rec 601 was along these same simple lines of specifically discussing what is actually stated in Rec 601.

Your reference introduces another standard so I really fail to see how it speaks specifically about Rec 601..Sure OK.. "Testing for Conformity with ITU-R Recommendations BT.601" is another standard?

I think I've listed a dozen examples now illustrating how a standard .7V p-p color bar signal IS encoded per Rec 601. FWIW (16-235) does NOT represent a 1V (Y') signal, not for ITU-R 601 or 709 period:p

According to my reference(s) one can't even represent/encode a 1V p-p (Blk-Wht) Y' signal, considering Black is 0V. So please stop suggesting White or code word 235 is 1000 mV.

According to your limited understanding of Rec 601, my Multi Standard SDI (SD-HD) waveform monitor should in reality read Y' code word 235 as 1000 mV and code word 16 should indicate 0V.....please stop the madness!

Thanks in advance:)

I think I've listed a dozen examples now illustrating how a standard .7V p-p color bar signal IS encoded per Rec 601.

Absolutely no one has issue with this, within the context you're using.

So please stop suggesting White or code word 235 is 1000 mV.

I have never intended to make any such claim. I have simply stated in different ways how Rec 601 does not define an absolute relationship between voltage and digital values independent of context. With the single context you're speaking of in the prior quote the relationship you're stating is fine, but to take that voltage to digital value relationship and say that it's universal to any situation as defined by Rec 601, that is my disagreement with what you've posted. Myself and others have stated this point (how there is not one absolute relationship between voltage and digital values defined by Rec 601) repeatedly and it seems you still fail to get what exactly is stated in Rec 601. As far as I'm concerned any discussion was done long ago, being how it seems you will not reference Rec 601 in regards to what you write and how you apparently will not consider questions dealing with simple general items like US NTSC levels or E'Y in relation to Rec 601 quantization.