Colorspace vs. Contrast Ratio - AVS Forum
Forum Jump: 
Reply
 
Thread Tools
post #1 of 246 Old 02-25-2007, 01:26 PM - Thread Starter
Advanced Member
 
mfogarty5's Avatar
 
Join Date: May 2006
Location: Charlotte, NC
Posts: 608
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 8 Post(s)
Liked: 19
I have read with great interest some of the threads in this forum. Specifically I found the debate below between the PC RGB colorspace and the studio RGB colorspace to be very interesting and confusing.

http://archive2.avsforum.com/avs-vb/...hreadid=416292

Specifically it appears that there is disagreement as to whether black than black(BTB) and whiter than white(WTW) are the intent of the creator or the result of poor encoding.

In any event I don't want to rehash the debate, but I do have two questions about this colorspace conversion as it relates to my display.

I have a 1080p LCD that accepts 1:1 pixel mapping over the VGA input, but not the HDMI input(At least not yet. There is a rumored firmware upgrade). As a result I am looking at getting an external video processor to get better de-interlacing and scaling. It appears that the VGA input is expecting PC RGB(0-255) rather than Video RGB(16-235). I understand the argument that Video RGB should be maintained from source to display, but I have to questions about this as it relates to my LCD.

First, I have a 10-bit display and it is my understanding that HDMI 1.1 is only capable of passing an 8-bit signal and that HDMI 1.3 will be capable of passing a 10-bit signal. VGA, on the other hand is capable of passing a 10-bit signal. If the above statements are true, then since I have an HDMI 1.1 display I am limited to passing 8-bit color via HDMI, but could pass 10-bit over VGA.

Second, one of the weaknesses of LCDs is their contrast ratio. Wouldn't converting a DVD's YCbCr(16-235) signal to a PC RGB(0-255) increase the contrast of my LCD at the expense of clipping BTB and WTW information? I understand that one limitation of performing the 16-235 to 0-255 conversion is banding, but shouldn't the fact that I have a 10-bit display mitigate this issue?

If what I have said is true it seems that connecting my LCD via VGA(0-255) could be superior because doing so would increase the perceived contrast ratio of my LCD and enable 10-bit processing in the external video processor rather than the internal video processor. The only drawback would be clipping of BTB and WTW information.

Is what I have said true or am I totally off base?

New: 2x Cisco 8642 w/ multi-room. Yes, that's right TWC has a multi-room DVR!!

Old: 2x SA8300HD TWC Navigator MDN 2.4.6-19
mfogarty5 is offline  
Sponsored Links
Advertisement
 
post #2 of 246 Old 02-25-2007, 01:39 PM
AVS Special Member
 
jvincent's Avatar
 
Join Date: Jan 2003
Location: Ottawa, Canada
Posts: 3,348
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 11
You're offbase.

The digital values used to represent colours have nothing to do with your perceived contrast ratio. That is entirely a function of your display.

If the display is set up correctly where "black" is displayed as the lowest possible light output and "white" is the maximum then you will have the greatest contrast ratio. This is true whether you send it a 10-bit or 8-bit or 24-bit encoded signal.

Banding usually occurs when a conversion is done without enough bit precision which results in rounding errors. These are manifested as banding.

If you feed the display via VGA with PC levels (0-255) these are still 8 bit values. If the conversion of the source signal (assuming YCbCr DVD, which is 8 bit) to PC levels was done with sufficient bit depth you may be OK. But if it isn't, banding city.

The fact that your display has 10-bit processing means that it uses 10-bits for the internal calculations when you adjust colour, brightness, etc. on the display.

Hope that helped.
jvincent is offline  
post #3 of 246 Old 02-25-2007, 02:42 PM - Thread Starter
Advanced Member
 
mfogarty5's Avatar
 
Join Date: May 2006
Location: Charlotte, NC
Posts: 608
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 8 Post(s)
Liked: 19
Quote:
Originally Posted by jvincent View Post

You're offbase.

The digital values used to represent colours have nothing to do with your perceived contrast ratio. That is entirely a function of your display.

If the display is set up correctly where "black" is displayed as the lowest possible light output and "white" is the maximum then you will have the greatest contrast ratio. This is true whether you send it a 10-bit or 8-bit or 24-bit encoded signal.

Banding usually occurs when a conversion is done without enough bit precision which results in rounding errors. These are manifested as banding.

If you feed the display via VGA with PC levels (0-255) these are still 8 bit values. If the conversion of the source signal (assuming YCbCr DVD, which is 8 bit) to PC levels was done with sufficient bit depth you may be OK. But if it isn't, banding city.

The fact that your display has 10-bit processing means that it uses 10-bits for the internal calculations when you adjust colour, brightness, etc. on the display.

Hope that helped.

Thanks.

I understand that the bit depth and contrast ratio are two different issues.

I shouldn't have referred to PC RGB as 0-255 on my display. I guess I am trying to figure out is that if an external VP processes 10-bit signals whether or not I can pass it as 0-16,581,375(10-bit) over VGA. If this is the case, and maybe even if it isn't, there shouldn't be any banding on my LCD since it has 10-bit processing.

I'm still confused about the setting of black. If reference black is set at 16, then isn't that artificially raising the black level if there are 15 levels beneath it? If I converted 16-235 to 0-255(8-bit)/0-16,581,375(10-bit) wouldn't that increase the contrast?

New: 2x Cisco 8642 w/ multi-room. Yes, that's right TWC has a multi-room DVR!!

Old: 2x SA8300HD TWC Navigator MDN 2.4.6-19
mfogarty5 is offline  
post #4 of 246 Old 02-25-2007, 03:01 PM
AVS Special Member
 
jvincent's Avatar
 
Join Date: Jan 2003
Location: Ottawa, Canada
Posts: 3,348
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 11
WRT to why reference black is at digital 16, I'd suggest a read of Chris Wiggles guide.

Short answer, it doesn't increase contrast, since that is a function of the display.
jvincent is offline  
post #5 of 246 Old 02-25-2007, 08:28 PM - Thread Starter
Advanced Member
 
mfogarty5's Avatar
 
Join Date: May 2006
Location: Charlotte, NC
Posts: 608
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 8 Post(s)
Liked: 19
Quote:
Originally Posted by jvincent View Post

WRT to why reference black is at digital 16, I'd suggest a read of Chris Wiggles guide.

Short answer, it doesn't increase contrast, since that is a function of the display.

I did read the Chris Wiggles guide and he relies heavily on the munson/spears thread that I linked in my original post.

Here is the relevant paragraph from the ChrisWiggles guide.

"Mapping Studio levels to PC levels can be done a few ways. Sometimes all the levels are just shifted down 16 steps, thus clipping off BTB data, but not introducing banding/contouring or clipping highlight details. If there are other PC level sources fed to the display, their whites will be substantially brighter. The clipping of BTB data is undesirable and the brightness mis-match is also noticeable, however note that there will be no banding problems. Usually the levels are expanded: digital 16 (black) is shifted down to 0, and 235 (white) is shifted up to 255 (or sometimes a value slightly lower than 255) thus expanding the numerical range between black and white to match PC levels. Note that this doesn't improve contrast in the final image. In this case, re-mapping Studio levels to PC levels will destroy BTB and peak white image data, and introduce banding/contouring artifacts because of the expansion. This is also undesirable."

1. I understand that mapping 16-235 to 0-255 will clip BTB and WTW information. I could probably live with that.

2. I also understand that on an 8-bit display that banding can occur, but I have a 10-bit display so this should be less of an issue if it is one at all.

3. What I don't understand is how mapping 16-235 to 0-255 doesn't increase contrast ratio.

Per Wikipedia, "The contrast ratio is a metric of a display system, defined as the ratio of the luminosity of the brightest and the darkest color the system is capable of producing."

An LCD uses the PC RGB colorspace of 0-255 which means that at 0 the backlight filter is trying to prevent as much light as it possibly can and a setting of 255 means that no filter is being used and that the backlight filter is letting in all light.

Setting reference black at 16 means that the authors are setting black 15 levels ABOVE the lowest level my display is capable of achieving. Setting reference white at 235 means that the authors are setting white 20 levels BELOW the lowest level my display is capable of achieving.

Believe me I would rather pass all signals in Video RGB over HDMI, but I also want to maximize my display's capablitlies. I would love for someone to explain to me or direct me to a post that explains how mapping 16-235 to 0-255 does not increase usable contrast ratio on a digital display.

New: 2x Cisco 8642 w/ multi-room. Yes, that's right TWC has a multi-room DVR!!

Old: 2x SA8300HD TWC Navigator MDN 2.4.6-19
mfogarty5 is offline  
post #6 of 246 Old 02-25-2007, 08:35 PM
Advanced Member
 
MASKOAA's Avatar
 
Join Date: Oct 2006
Posts: 551
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 10
I dont know if this pertains to you or not but I have a Upscaling DVD player hooked up to my TV through HDMI and it has the option of RGB FULL RANGE, RGB, or COMPONENT for video modes.....after expermenting with all 3 I found Component to look the sharpest with the best levels of black and white. The other 2 crushed my blacks and blended shadows together.
MASKOAA is offline  
post #7 of 246 Old 02-25-2007, 09:58 PM - Thread Starter
Advanced Member
 
mfogarty5's Avatar
 
Join Date: May 2006
Location: Charlotte, NC
Posts: 608
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 8 Post(s)
Liked: 19
Quote:
Originally Posted by MASKOAA View Post

I dont know if this pertains to you or not but I have a Upscaling DVD player hooked up to my TV through HDMI and it has the option of RGB FULL RANGE, RGB, or COMPONENT for video modes.....after expermenting with all 3 I found Component to look the sharpest with the best levels of black and white. The other 2 crushed my blacks and blended shadows together.

Your findings make sense because the HDMI input is probably expecting the Component(YPbPr) color space.

In order to do a true comparison, you would have to change that same setting in BOTH the DVD player and the display.

New: 2x Cisco 8642 w/ multi-room. Yes, that's right TWC has a multi-room DVR!!

Old: 2x SA8300HD TWC Navigator MDN 2.4.6-19
mfogarty5 is offline  
post #8 of 246 Old 02-25-2007, 10:06 PM - Thread Starter
Advanced Member
 
mfogarty5's Avatar
 
Join Date: May 2006
Location: Charlotte, NC
Posts: 608
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 8 Post(s)
Liked: 19
Actually both ChrisWiggles and stacey spears agree that using 0-255 instead of 16-235 on a digital display like an LCD would increase contrast ratio. In fact Stacey Spears could gain a 50% increase in contrast ratio by using 0-255 instead of 16-235.

I think part of the problem is that most of the "experts" detail their findings based on CRTs, but most of the people here at AVS have digital displays.

Quote:
Originally Posted by Stacey Spears View Post

You do lower your CR when you calibrate to have all values above 235 visible. I went from 1500:1 down to 1000:1 when I did that. Too me it is worth the loss in CR to not have blow out whites.

http://www.avsforum.com/avs-vb/showt...&&#post5034296

Quote:
Originally Posted by ChrisWiggles View Post

Digital displays have very limited on/off CR capabilities, so if you align the display's peak white to the reference white point of the video, then incomind peak whites will be clipped. If you align both the peak white points, the display's and the incoming video, then all that data will be maintained and rendered correctly on the display. The latter situation will lead to lower in-scene contrast ratios as measured between black and reference white, but data above reference white will be included. This tradeoff is a weakness of the display type, as the display has a limitation in on/off contrast ratio.

http://www.avsforum.com/avs-vb/showt...&&#post6078146

New: 2x Cisco 8642 w/ multi-room. Yes, that's right TWC has a multi-room DVR!!

Old: 2x SA8300HD TWC Navigator MDN 2.4.6-19
mfogarty5 is offline  
post #9 of 246 Old 02-26-2007, 02:52 AM
Member
 
Daniel.N's Avatar
 
Join Date: Oct 2002
Location: Umeå, Sweden
Posts: 72
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 10
There's a world of difference between being having >235 values visible, and expanding to PC 0-255 levels. They are not the same thing. Neither Chris nor Stacey meant this. They were both talking about maintained 16-235 video levels!

--Daniel
"Ignorance is bliss"
Daniel.N is offline  
post #10 of 246 Old 02-26-2007, 04:27 AM
AVS Special Member
 
Geof's Avatar
 
Join Date: Feb 1999
Location: Eden NY
Posts: 6,007
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 2 Post(s)
Liked: 33
Quote:
Originally Posted by mfogarty5 View Post

Per Wikipedia, "The contrast ratio is a metric of a display system, defined as the ratio of the luminosity of the brightest and the darkest color the system is capable of producing."

An LCD uses the PC RGB colorspace of 0-255 which means that at 0 the backlight filter is trying to prevent as much light as it possibly can and a setting of 255 means that no filter is being used and that the backlight filter is letting in all light.

Setting reference black at 16 means that the authors are setting black 15 levels ABOVE the lowest level my display is capable of achieving. Setting reference white at 235 means that the authors are setting white 20 levels BELOW the lowest level my display is capable of achieving.

I don't think that's true. I can't imagine any display manufacturer throwing away contrast range on their displays. They're not setting black level 15 steps above the lowest black of your display. It's just remaped to level 16 instead of level 0. If "X" is best black level and "Y" is the brightest white level then X=0 and Y=255 in the PC world but X=16 and Y=235 in the video world. Notice that the entire range of the display remains the same in either case. There is more granularity (smaller steps) in the PC world than in the video world but the contrast range is the same. Consider this too....some displays have more contrast range than others - they're certainly not all the same in this regard - even though they all use the same video levels.

Geof
Geof is offline  
post #11 of 246 Old 02-26-2007, 06:01 AM
Member
 
Daniel.N's Avatar
 
Join Date: Oct 2002
Location: Umeå, Sweden
Posts: 72
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 10
Agreed; while there MAY be displays on the market that really are that badly designed, it's not like that for normal displays. If you first calibrate so that 255 gives maximum light output and then crank contrast so that whites are crushed you will get maximum light output from somewhere <255 and all the way up to 255. There is really a LOT of misunderstanding about levels going on here. Hooray for forums like this, where such misconceptions are quickly replaced with knowledge :-)

--Daniel
"Ignorance is bliss"
Daniel.N is offline  
post #12 of 246 Old 02-26-2007, 05:25 PM - Thread Starter
Advanced Member
 
mfogarty5's Avatar
 
Join Date: May 2006
Location: Charlotte, NC
Posts: 608
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 8 Post(s)
Liked: 19
Quote:
Originally Posted by Daniel N. View Post

There's a world of difference between being having >235 values visible, and expanding to PC 0-255 levels. They are not the same thing. Neither Chris nor Stacey meant this. They were both talking about maintained 16-235 video levels!

Agreed; while there MAY be displays on the market that really are that badly designed, it's not like that for normal displays. If you first calibrate so that 255 gives maximum light output and then crank contrast so that whites are crushed you will get maximum light output from somewhere <255 and all the way up to 255. There is really a LOT of misunderstanding about levels going on here. Hooray for forums like this, where such misconceptions are quickly replaced with knowledge :-)


Quote:
Originally Posted by Geof View Post

I don't think that's true. I can't imagine any display manufacturer throwing away contrast range on their displays. They're not setting black level 15 steps above the lowest black of your display. It's just remaped to level 16 instead of level 0. If "X" is best black level and "Y" is the brightest white level then X=0 and Y=255 in the PC world but X=16 and Y=235 in the video world. Notice that the entire range of the display remains the same in either case. There is more granularity (smaller steps) in the PC world than in the video world but the contrast range is the same. Consider this too....some displays have more contrast range than others - they're certainly not all the same in this regard - even though they all use the same video levels.

Thanks Geof. Your comments were much better than Daniel N.'s condescending remarks.

What I am having a hard time understanding is that if what you are saying is true then no LCD would be capable of showing BTB or WTW since levels 1-16 would be the same and 235-255 would also be the same.

As far as I can tell with a digital display, you can either pass BTB and WTW while limiting contrast ratio OR you can clip BTB and WTW information and get maximum contrast ratio.

I would really appreciate if someone could explain to me why the statement above is untrue rather than simply saying that I have a misconception(daniel N.).

As much as it pained them to do so, I have provided 2 quotes from ChrisWiggles and Stacey Spears that agree with me.

New: 2x Cisco 8642 w/ multi-room. Yes, that's right TWC has a multi-room DVR!!

Old: 2x SA8300HD TWC Navigator MDN 2.4.6-19
mfogarty5 is offline  
post #13 of 246 Old 02-26-2007, 05:59 PM
 
ChrisWiggles's Avatar
 
Join Date: Nov 2002
Location: Seattle
Posts: 20,730
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 14
Quote:


3. What I don't understand is how mapping 16-235 to 0-255 doesn't increase contrast ratio.

Let me see if I can explain this clearly. There are few different considerations involved.

The first is the dynamic range of the content itself. Studio levels(nominal 16-235) has increased dynamic range over PC levels(0-255), while also ever slightly less bit depth within the reference range. This may seem counterintuitive because the numerical range is larger with PC levels. So keep in mind that the dynamic range of the content itself is slightly greater with Studio levels.

Regardless, the next consideration is how to display the ranges. We want black to be black on the display, obviously. This is a simplificationi, but I'll move on from that. But what should white be? On a CRT there isn't a hard clip point, so it's not really a problem. We set white to a bright enough level, and we enjoy our image. With digital displays our contrast range via the display is very low, and we have these relatively hard clip points including for white. We have very different aims when calibrating. When setting white on a digital display if you don't maximize the limited CR capability of the display you're not getting as good an image as you can. Now you see that we have a problem that we did not previously have to consider whith CRT which doesn't really have a maximum white point at all and doesn't have a limited CR range that you're trying to maximize (since it's essentially infinite).

Do you choose to align peak white of the display to peak white of the content, which would be 254? This would be technically correct and maintain all video content.

Do you choose instead to boost white level beyond that and say choose to align the display's peak white to the content's reference white which would be 235? Now we're clipping or colorshifting peak whites that may be present and that may or may not be significant. But we're getting better CR performance out of our display.

Do we choose come other compromise point?

These are decisions that each user encountering this compromise should make on their own. The more severely you cut into the video content, the more likely visibly negative consequences will result. This will depend partly on the specific content, and also your pickiness and preferences. But basically yes, the more whites you clip off yes you are expanding the contrast in the parts of the image that are left. Of course, a good argument can be made that this is a very worthy compromise if done conservatively and with purpose and awareness.

Now, remember that this is different than just remapping 16-235 to 0-255. Doing that won't provide any contrast ratio advantages that you couldn't achieve by simply aligning the display to 16-235 in the first place. But you do force the choice and decrease the content's dynamic range, and also may likely end up suffering banding problems (though so many systems have such terrible banding problems it may not be a particular concern).

Quote:


2. I also understand that on an 8-bit display that banding can occur, but I have a 10-bit display so this should be less of an issue if it is one at all.

Remember that uunless you have a way to feed that display higher bit-depth (i.e. not DVI or RGB via HDMI), then you are going to run into banding problems because there is an 8-bit bottleneck there.

Quote:


Actually both ChrisWiggles and stacey spears agree that using 0-255 instead of 16-235 on a digital display like an LCD would increase contrast ratio.

I agree, but to be clear this is aligning the display to 16-235 instead of 16-254. I do not advocate re-mapping the range to 0-255 and then aligning to 0-255. And the same statement can also be taken to an illogical extreme. The more you raise your white level, the greater contrast you're going to get for what's left. Does this mean that it's a good idea to turn up your white level to the absolute maximum? Of course not. The point is to get the best and most accurate image possible overall, and with a digital display with limited on/off CR range and hard clip points this involves some compromises.
ChrisWiggles is offline  
post #14 of 246 Old 02-26-2007, 06:20 PM - Thread Starter
Advanced Member
 
mfogarty5's Avatar
 
Join Date: May 2006
Location: Charlotte, NC
Posts: 608
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 8 Post(s)
Liked: 19
Thanks Chris. I really appreciate your post.

Quote:
Originally Posted by ChrisWiggles View Post

Remember that unless you have a way to feed that display higher bit-depth (i.e. not DVI or RGB via HDMI), then you are going to run into banding problems because there is an 8-bit bottleneck there.

Actually the whole reason I began researching color spaces is because my 1080p LCD can only achieve 1:1 pixel mapping over the VGA/D-SUB input. This input expects to receive PC RGB(0-255) signal and I wanted to understand the impact of converting Video RGB(16-235) to PC RGB. As I understand it, a VGA/D-SUB cable can carry a 10-bit signal.

Quote:
Originally Posted by ChrisWiggles View Post

Now, remember that this is different than just remapping 16-235 to 0-255. Doing that won't provide any contrast ratio advantages that you couldn't achieve by simply aligning the display to 16-235 in the first place. But you do force the choice and decrease the content's dynamic range, and also may likely end up suffering banding problems (though so many systems have such terrible banding problems it may not be a particular concern).

I agree, but to be clear this is aligning the display to 16-235 instead of 16-254. I do not advocate re-mapping the range to 0-255 and then aligning to 0-255. And the same statement can also be taken to an illogical extreme. The more you raise your white level, the greater contrast you're going to get for what's left. Does this mean that it's a good idea to turn up your white level to the absolute maximum? Of course not. The point is to get the best and most accurate image possible overall, and with a digital display with limited on/off CR range and hard clip points this involves some compromises.

I have a lot to learn about calibration and I still don't understand the quote above, but I will continue to re-read it until it sinks in.

Thanks again for your help.

New: 2x Cisco 8642 w/ multi-room. Yes, that's right TWC has a multi-room DVR!!

Old: 2x SA8300HD TWC Navigator MDN 2.4.6-19
mfogarty5 is offline  
post #15 of 246 Old 02-26-2007, 11:56 PM
Member
 
Daniel.N's Avatar
 
Join Date: Oct 2002
Location: Umeå, Sweden
Posts: 72
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 10
I blame the condescending..ness on the language barrier Point here, in few words, is that you have a fixed dynamic range of your display, you have content stored as 1-254 out of which somewhere between 16-235 and 16-254 are values you want to map to the range of your display.

From a pure contrast standpoint, whether you calibrate so that 16-235 uses your displays entire range, or if you expand 16-235 to 0-255 and then calibrate that to use your displays entire range, the end result is the same, in terms of contrast. This is sort of what basic calibration is about; matching the range of your display to what your feeding it.

But, while everybody calibrate so values <16 are not seen, values >235 really need to be seen. Otherwise clouds will look like marshmallows. There are details >235 well worth to "show". I tend to calibrate so that I can "see" up to 240-245 something. Still, most "white" content will be at 235 exactly. This means I will not use my display's whitest white for 235, only 245 and above will use that white. So, I'm saving a bit of my displays range for values >235, those values aren't used as much as those <=235. This will, yes, lower my contrast for the 16-235 values, but I will still use the entire range of my display for the input range I've decided to show. The reason I don't show all the way up to 254 (instead 240, or 245) is I don't want to reserve too much of my display's limited range for those >235 values.

--Daniel
"Ignorance is bliss"
Daniel.N is offline  
post #16 of 246 Old 02-27-2007, 01:31 AM
AVS Special Member
 
dlarsen's Avatar
 
Join Date: Mar 2001
Location: Beaverton, OR, USA
Posts: 1,908
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 11
Quote:
Originally Posted by Daniel.N View Post

Otherwise clouds will look like marshmallows.

Can you post up a lossless, (.png) full-frame, StudioRGB level, image capped from a mainstream (big studio), telecined, DVD that will show this alleged damage you purport if WTW (>235) is clamped? (or if it is rendered at sRGB levels)?

If you can't post such an image, can you site a more specific title/scene/timestamp and I can see if I can do it? I can post up both the otherwise unaltered clamped and unclamped images in a A/B fashion and we can all observe the clouds turn to marshmallows. If they do.

I've done this dozens of time now. Only ONE person ever reported being able to perceive ANY difference at all. My diagnosis is that person was delusional. Maybe you can show us?

Dave
dlarsen is offline  
post #17 of 246 Old 02-27-2007, 06:33 AM
AVS Special Member
 
dr1394's Avatar
 
Join Date: May 2002
Location: Mizar 5
Posts: 3,140
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 3 Post(s)
Liked: 19
It's easy to find compressed frames with BTB. In these frames, I've changed all the Y pixels with values below 16 to 235. Real detail or just MPEG-2 quantization noise?









Ron

HD MPEG-2 Test Patterns http://www.w6rz.net
dr1394 is offline  
post #18 of 246 Old 02-27-2007, 07:05 AM
 
tbrunet's Avatar
 
Join Date: Jul 2005
Location: Savannah, GA
Posts: 2,750
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 12
This is the same demo (different example) where you invert pixel values for illustration? I never said that analog filter ring or HF edge detail did not overshoot or actually modulate said excursion.

If my memory is correct, Dave will not able to provide a meaningful A/B comparison if you obliterate the original (BTB/WTW) extra information... thus making clamped vs un-clamped image comparison impossible.
tbrunet is offline  
post #19 of 246 Old 02-27-2007, 11:49 AM
AVS Special Member
 
dlarsen's Avatar
 
Join Date: Mar 2001
Location: Beaverton, OR, USA
Posts: 1,908
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 11
Ron,

Thanks for posting up the images. I have a few comments.

1) The images you posted are about BTB when the discussion was more about WTW (>235). It seems even Chris recommends setting up ANY display such that BTB values are clamped or not visible so all those highlighted pixels you show would be rendered imperceptible by that. Are you recommending otherwise? That we should reserve some of our displays range to render these BTB pixels you highlight?

2) As Thomas noted, by inverting the BTB pixels, you have altered the very pixels in question making further analysis, measurements, or A/B comparisons impossible. I have to ask, if the claim is that these pixels are perceptible image information, why do you need to invert them in an effort to show them? If it's perceptible image information' why can't it be show as such in an A/B comparison between two otherwise unaltered images?

3) Please note the large number (I'd guess almost half of them) of BTB pixels in the images you posted are well into the black bar area. How can that possibly be image information' when it's not even part of the active image area? If the black bars are to be black, what are all those highlighted pixels you show that are in the black bar area? If they aren't black as they should be, I maintain that those pixels are wrong and are errors. Filter ring and rubble. Note that clamping them would correct' them.

Even though I think the discussion was more about WTW (>235) than BTB, I think it would be interesting if you could post up the original, unaltered, lossless, StudioRGB caps and A/B it against the same image where the BTB is clamped. I predict that no one will be able to perceive any difference at all. I'm sure you can set up such a comparison but I can if you'll post the original unaltered, unclamped images. I did say otherwise unaltered'. Perhaps you have some WTW(>235) examples as well? We should be able to do the clouds turning to marshmallows test if it is infact perceptible image information encoded on all our DVDs.

Dave
dlarsen is offline  
post #20 of 246 Old 02-27-2007, 12:08 PM
AVS Special Member
 
dlarsen's Avatar
 
Join Date: Mar 2001
Location: Beaverton, OR, USA
Posts: 1,908
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 11
Quote:


Real detail or just MPEG-2 quantization noise?

As I noted above, how can all those BTB pixels you highlighted that are in the encoded black bar area (~ half of them?) possibly even be a candidate for real detail'? This should be no detail there. If those black bars should be solid black then those that aren't are ERRORS or NOISE. If we agree that all those highlighted pixels in the black bars are errors, then I don't believe it bodes well for the others either as I suspect they are in the image for the same reasons they are also in the non-active image area. Again, errors and noise.

I see the same thing with WTW(>235) too. It's usually distributed in a very EE like fashion. A halo around a bright object rather than towards the center. Just like EE and filter ring.

Dave
dlarsen is offline  
post #21 of 246 Old 02-27-2007, 12:15 PM
AVS Special Member
 
dlarsen's Avatar
 
Join Date: Mar 2001
Location: Beaverton, OR, USA
Posts: 1,908
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 11
Ron,

One other note. It appears the images you posted are with sRGB levels rather than StudioRGB levels and won't contain ANY BTB or WTW(>235) pixels so there is nothing to analyze or measure.

Dave
dlarsen is offline  
post #22 of 246 Old 02-27-2007, 01:54 PM
AVS Special Member
 
Erik Garci's Avatar
 
Join Date: Oct 1999
Location: Durham, NC, USA
Posts: 4,770
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 10
dlarsen,

In an archived post, I pointed out a scene in the opening of Toy Story 2 that seems to have significant amounts of B'<16 and R'>235, although I don't have a studio RGB capture to confirm it. Stacey Spears also posted some images in that thread.
Erik Garci is offline  
post #23 of 246 Old 02-27-2007, 02:18 PM
AVS Special Member
 
dlarsen's Avatar
 
Join Date: Mar 2001
Location: Beaverton, OR, USA
Posts: 1,908
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 11
Quote:
Originally Posted by Erik Garci View Post

In an archived post, I pointed out a scene in the opening of Toy Story 2 that seems to have significant amounts of B'<16 and R'>235, although I don't have a studio RGB capture to confirm it. Stacey Spears also posted some images in that thread.

And as I believe I noted, I considered a CGI based image as somewhat of a special case as they may not encounter the same telecine work flow as 'film' would. See Poynton. CGI does utilize 'superblack' and 'superwhite' for purposes other than the actual image information. The CGI montage elements on DVE are a good example of this. If you observe a histogram on this montage, The CGI based footage has a clear and distinctive signature relating to BTB/WTW(>235) that none of the film based segments have.

Also, if memory serves, that wasn't a full frame A/B image. If one was to post just a hyper-enlarged single pixel, then maybe we could perceive the difference. I don't recall seeing any full-frame, from telecined, mainstream studio, A/B comparisons where the only difference between the two images is the clamp. (Except for all the ones I have posted.)

Dave
dlarsen is offline  
post #24 of 246 Old 02-27-2007, 02:39 PM
AVS Special Member
 
Erik Garci's Avatar
 
Join Date: Oct 1999
Location: Durham, NC, USA
Posts: 4,770
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 10
Quote:
Originally Posted by dlarsen View Post

CGI does utilize 'superblack' and 'superwhite' for purposes other than the actual image information.

In the comparison image of Buzz's arm, the R'>235 values seem to be actual image information.
Erik Garci is offline  
post #25 of 246 Old 02-27-2007, 02:51 PM
AVS Special Member
 
dlarsen's Avatar
 
Join Date: Mar 2001
Location: Beaverton, OR, USA
Posts: 1,908
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 11
Yes, but I have to admit, even blow up to these extremes, I have a hard time perceiving any difference at all. I suspect that if this was presented full-frame in a A/B manner, there would be even less. Again, I also note, GCI based images can be a somewhat special case.
dlarsen is offline  
post #26 of 246 Old 02-27-2007, 03:44 PM
 
ChrisWiggles's Avatar
 
Join Date: Nov 2002
Location: Seattle
Posts: 20,730
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 14
We've been over this terrain before Dave. If you feel things should be different, join SMPTE and create a new video standard.
ChrisWiggles is offline  
post #27 of 246 Old 02-27-2007, 03:57 PM
AVS Special Member
 
dlarsen's Avatar
 
Join Date: Mar 2001
Location: Beaverton, OR, USA
Posts: 1,908
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 11
Quote:


join SMPTE and create a new video standard.

No need. sRGB already is a standard and doesn't seem to need any reserved range (BTB/WTW) to create very good looking 8 bit images. Better that what I can hope to get from a DVD. No Marshmellows there. As far as new standards go, I recall Don and Stacy (when they were associated with MS) commenting about how LongHorn (the successor to XP) was going handle this whole issue between StudioRGB and sRGB. The OS was going to determine if one was viewing video' or graphics' and dynamically change levels on the fly. What happened to that? I've seen no mention of this feature' making it's way into Vista. Musta gotten dropped? Why? Perhaps it turns out to be harder than it first seemed to make the distinction in the land of convergence. Perhaps the cure turned out much worse than the purported disease?

Dave
dlarsen is offline  
post #28 of 246 Old 02-27-2007, 04:23 PM
 
ChrisWiggles's Avatar
 
Join Date: Nov 2002
Location: Seattle
Posts: 20,730
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 14
I don't not believe sRGB has ever been incorporated as any SMPTE/ITU standard for video or broadcast. I could be mistaken, but as far as I am aware it is for the computer world for still images and is an IEC standard. If you wanted to apply this standard for moving images, you would need to create a new standard appropriate for that task. Using RGB of course, is inappropriate for video broadcast/distribution regardless.
ChrisWiggles is offline  
post #29 of 246 Old 02-27-2007, 04:46 PM
AVS Special Member
 
dlarsen's Avatar
 
Join Date: Mar 2001
Location: Beaverton, OR, USA
Posts: 1,908
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 11
It is a IEC standard. (IEC 61966-2-1) It's the default color space for Colour Measurement and Management in Multimedia Systems and Equipment'. It's also Windows default color space and has been for a long long time.

Multimedia. Multi. Media.

As I noted, with convergence, the distinction between the two is blurring fast. I really don't understand why you make the distinction to images that are moving versus ones that are static. sRGB can also render very good looking 8-bit images in motion too. (as most anyone viewing motion images on a PC can attest)

Are you suggesting that BTB/WTW has advantages in the temporal domain as well?

Dave
dlarsen is offline  
post #30 of 246 Old 02-27-2007, 05:09 PM
 
ChrisWiggles's Avatar
 
Join Date: Nov 2002
Location: Seattle
Posts: 20,730
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 14
Quote:
Originally Posted by dlarsen View Post

It is a IEC standard. (IEC 61966-2-1) It's the default color space for Colour Measurement and Management in Multimedia Systems and Equipment'. It's also Windows default color space and has been for a long long time.

Multimedia. Multi. Media.

As I noted, with convergence, the distinction between the two is blurring fast. I really don't understand why you make the distinction to images that are moving versus ones that are static. sRGB can also render very good looking 8-bit images in motion too. (as most anyone viewing motion images on a PC can attest)

Dave

No, what I'm suggesting is that the horse died several years ago. I'm not going to debate this with you again for the umpteenth time. It's a waste of my time and of everybody else's time.

We have well-established standards that have been in use around the world for many years. If you feel a new video standard is in order, take your case to SMPTE and propose something new.

Quote:


Are you suggesting that BTB/WTW has advantages in the temporal domain as well?

No.
ChrisWiggles is offline  
Reply Display Calibration

User Tag List

Thread Tools
Show Printable Version Show Printable Version
Email this Page Email this Page


Forum Jump: 

Posting Rules  
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are Off
Pingbacks are Off
Refbacks are Off