or Connect
AVS › AVS Forum › Display Devices › CRT Projectors › Black bars? Well, not quite, more dark gray.
New Posts  All Forums:Forum Nav:

Black bars? Well, not quite, more dark gray. - Page 3  

post #61 of 84
Originally posted by WanMan
..... According to the lecture, the left of the two black bars should become invisible, but the right one should still be visible.....My discerning eyes can see both bars. This means the black level is set too high. The left of the two black bars is not as well defined (and not as well discerned), but its visible nonetheless. This means I need to lower the brightness so that only the right black bar in this pattern is visibel, correct?

Look out. Those directions were written with display that have relatively poor black level stability in mind (aka most consumer displays). I wrote that the leftmost bar disappears because on sets with poor black level stability that slightly brighter than black bar does go away when the display is set to a good compromise black level. However, the grade of display we work with here on the forum usually has good enough black level stability to expect that bar to be visible when a good black level is set.

The left moving black bar is 1 IRE above true black. The right moving black bar is 2 IRE above true black.
post #62 of 84
Curious, Guy. If the left bar is 1-IRE about black and the right 2-IRE above black, neither assume that black is correctly set, right? I suppose what I am trying to determine is that with a black pattern (0-IRE) should any tube or raster be lit?

With the black pattern up and its 'Black' label present, I can see the raster on the two tubes being used to display this label. Should this be?

As a side note, Guy, I have been meaning to find an answer to a question I've had for a few months now regarding brightness and contrast. Can I PM it to you?
post #63 of 84
Thread Starter 
Originally posted by John Alison
We seem to have hijacked this thread. I'll get back to you privately.

John et al,
This is really good stuff and I'm sure many other folks are reading this with interest. Very much on topic as far as I am concerned and I'm glad of your (and others) valuable expertise.
post #64 of 84
Originally posted by WanMan
Curious, Guy. If the left bar is 1-IRE about black and the right 2-IRE above black, neither assume that black is correctly set, right?
Rant Mode ON!!!!

What I read on the forums about the concepts of black, encoded digital level, 0 IRE, and 7.5 IRE is a "freaking mess" to put it mildly. Many incorrect postings or partially correct postings abound and multiply that it is a wonder anyone understands what anyone else is talking about. This is worse than trying to teach newbies what reference level means.

I'll try, but my patience for this isn't very good.

1. Digital video encodes it signals with black at digital level 16 and white at digital level 235. Levels 0 and 255 are illegal for encoding any image content. Levels 1 to 15 are below black and used to handle undershoot. On a display, material from levels 1 to 15 supposed to be the same absence of light as digital 16 (black). Material from levels 236 to 254 handle above white overshoot. Picture information on pefectly mastered material has no intentionally visible image data below 16 or above 235. Black is black. Things below black are black. Nothing is brighter than max white 235. Processing and/or poor video sourcing may end up with some picture information outside the recommended range of digital 16 to 235 so preservation of those values may be of use.

2. Computer video has black at digital 0 and white at digital 255. Notice how this differs from digital video. Oh great. True black for computer graphics is at digital 0, but true black for digital video is at digital 16. Even worse, HTPC playback may or may not expand the digital video range to fill the computer video range.

3. IRE is voltage unit which was used to simplify expressions of the 0.7 volt excursion of video signal content. It was more convenient to express things in IRE than in odd fractions of of a volt. 0 IRE is at the baseline of the signal. 100 IRE is at the top of the signal range and indicates max white. If North American NTSC setup is present in an analog signal, black is represented as 7.5 IRE. If no setup is present, black is represented as 0 IRE.

4. Black means BLACK. If I say black, that unambiguously means black, true black. However, if you say 0 IRE or 7.IRE, the meaning is somewhat ambiguous because you haven't also stated whether or not setup is present. For example: 0 IRE is black if the system has no setup, but it is 7.5 IRE darker than true black on a system with setup. This is why I always refer to the bars relative to BLACK rather than an absolute IRE level. The absolute IRE level changes.

On the Avia discs, we LABEL the patterns expecting black to be output by the playback device at 7.5 IRE so the labels are correct only if the playback device represents black as 7.5 IRE and white at 100 IRE.

If there is no setup, the labeling would be slightly off. You can translate the on screen IRE labels to their zero setup IRE level using Zero Setup IRE = (Labeled IRE - 7.5) *100/92.5

So we both 0 IRE and 7.5 IRE can mean the EXACT SAME signal level as true black (video digital 16) depending on the equipment. HTPC folks probably did the most in confusing the situation by suddenly taking their "convention" of calling black = 0 IRE without also stipulating that digital 16 must also be at 0 IRE for this to be true. Any mismatch and you end up with elevated blacks or submerged blacks.

5. DVI came along and further complicated things because DVI interfaces also work with computer graphics video. Recall that computer RGB has black at 0 and white at 255, but digital video has black at digital 16 and white at digital 235. The display and source device must match in their conventions or once again you end up with submerged or elevated blacks.

6. HTPC's confuse things in the same way as DVI does because the on screen graphics and played back video may not have true black in both systems at equal digital levels. Calibrate to make black appear correct for DVD playback and you may mess up your desktop. HD played back through HTPC can have the same type of problem. The goal for each source is to make true black in THAT type of signal be black (neither submerged nor elevated) on the display.

7. No matter what the playback system, processor, or display deviced, nothing changes the on disc encoded level of black from being digital 16. In other words, you equipment doesn't magically get the disc to rewrite itself and change the values. What is encoded as black on disc is always black. It's the subsequent playback and display system that can mess things up.

8. If you have correctly understood digital levels through the above the the incredible misunderstanding it takes to come up the following statment and question should be obvious:

Calibration disc A has black at 7.5 IRE but calibration disc B has its black at 0 IRE.

No. If accurately encoded, both DVD's must have black encoded at the same digital 16. It is the playblack equipment settings that determines what IRE the true black lies. If the player has setup, then black is at 7.5 IRE. If not, then black is at 0 IRE.

Can't you put a signal on disc that is always at 0 IRE so we can compare against it?

No. No matter what digital level we encode the output IRE is varied by the player's setting. A DVD cannot overried a disc player in this regard.

To get things right, each signal stage must match the next piece of equipment's expectation for where black is on the digital range. Get it wrong anywhere along the way and you have a problem.

Labeling on the AVIA, Avia PRO, and S&VHTT discs are done to follow the convention that 7.5 IRE represents true black (digital 16) and 100 IRE represents max white (digital 235). So when I say black, I mean true black (digital 16), the black background of patterns in AVIA, or something labeled as 7.5 IRE in AVIA.

I guarantee that within a month there will be enough incorrect postings to render what I just wrote a useless reed in the wind. This is a topic that people have a knack for clearly, convincingly, and thoroughly misunderstanding.
post #65 of 84
Guy, is PAL encoded the same? I thought the black level was different, but I, like many. are thoroughly confused by all this stuff.
post #66 of 84
I'm only further confused. While I get what was just posted, I wonder if by making the playback system (Panasonic RP-56) use 'darker blacks' might be wrong, or right.

Then again, I have a significant amount of Superbowl beer in me to make common-sense seem confusing.
post #67 of 84
Making the playback system put black "darker" makes things look wrong if the display expects to see black at a lighter level. It makes things look right if the display expects black to be at a "darker" level. The key is that the two systems must agree or the display must both accept the entire signal AND have sufficient adjustment range to compensate for an unexpectedly low or high level. Just to help things out, manufacturers either don't specify or use non-sensical marketing terms to describe the levels.

Let's make it blood simple to decide the relative source/display mismatches.

If you can't adjust brightness on the display low enough to make black (digital 16) in AVIA be the darkest black possible on the display, the signal level for black is too high for the display.

If you can't adjust brightness on the display high enough to make the darkest (left) black level bar in AVIA appear, then the signal is representing black below the point at which the display clips dark end signals. The source's black level is too low for the display.
post #68 of 84
Guy- Thanks for Rant mode information and 'Blood simple'

WanMan- state of play is that with 0V (he says avoiding any mention of IRE) across your RGB inputs, then screen looks dark, but you can still play shadow puppets. However, you can distinguish black bars as discussed in 'blood simple' paragraph.

Rant on: It amazes me how video and computer geeks can make y = mx + c so complicated. PC world and video world define brightness and contrast as different functions of m and c. Anyone want an argument for fun?

Pedantry on: IRE can't be a voltage unit. It's dimensionless.
post #69 of 84
I don't what it's like in NTSC land, but generally having worked through a set up test DVD (such as the superb, but NTSC based AVIA) making sure that DVD player, processor and PJ are all correctly calibrated, I then turn to a television feed and have to immediately change the brightness and contrast on the PJ depending on the programme. 'Undershoot' and 'overshoot' is putting it mildly.
post #70 of 84
same in my setup. I would need a channel dependent brightness and contrast switch...
post #71 of 84
I have my Marquee set up to accomodate these differences in black levels by using the four color temperature presets and the 80 channel memories. I'm not using all of them, of course, but I have enough set up that I can usually find a preset channel that looks really good on any given signal.

post #72 of 84
The Marquee looks as though it's more flexible than my Barco. I have a stack of memory blocks, but only two are assigned per input resolution, and given that I want to stick to one input resolution (RGBHV 960p into port 3), my options are limited.

Sort of getting back on thread...CSI Miami or Vegas is good for setting black level. Moodily lit night time scenes. I believe you get it in high definition. Drool.
post #73 of 84
I do recognize the need for using a properly configured playback system with the software being used, but like so many other newbies the playback system is an added mystery. Thus, I do not know if my DVD-player (Panasonic RP-56), with its option for Lighter/Darker Blacks, is presenting to the CRT's input Digital 16 properly, as the DVD itself is encoded with.

And since most consumer electronics is a marketing nightmare for consumers, I'd be less than confident in any information I dig up. So, if I cannot be confident, as a consumer, in trusting the playback system and the PC-world (and all of its HTPC's and ISF-equipped laptops) the call is open to figuring out how the consumer calibration software can be confidence employed.

This consumer lacks any confidence at this point, and I'll just leave it be.
post #74 of 84
WanMan- I agree.

I'm contemplating mounting an oscilloscope in my AV rack. I could shove a luminance signal from a spare s-video socket on my DVD player into channel 1, shunt R G B from the video processor with channels 2,3 and 4, then run those grey scales.
post #75 of 84
Hey John,

This isn't really something that I'd want to consider. While my history does come with scope-skills (took electronics back in the early 1980's), this would be a path a bit much for the consumer in me.

Funny thing is that is 'IRE' is considered a bad way to describe anything then it gets used a heck of a lot. I actually asked a few times during the past couple of years if it was a unitless, linear scale that defined a voltage range and, well, still waiting on that answer.

I don't think any convention is bad if it is understood and used for the manner it is composed for. All I want to do is determine a) if my raster is being prematurely lit or not. This isn't terribly important at the moment considering the design limitations of the projector I am currently using, but learning first for the 'big installation' (XGLC) will most certainly require me to have a much more critical eye and understanding.

I suppose that I could buy a junk projector and use it, with a scope, to learn from. Kind of like making the best dang street-legal Yugo that Yugo never meant to be born. :)
post #76 of 84
I think that Guy in rant mode defines IRE quite nicely! This is my understanding:

IRE = (Video signal amp in volts/0.7 volts) x 100

which is linear in Video signal and dimensionless. (It's a normalised value expressed as a percentage)

The standard is for IRE to be in the range 7.5 to 100, so standard voltage amplitude range (across the video cable feeding e.g. your display) is 0.0525 V to 0.7 V.

If the display receives 0.0525 V, then the brightness should be set so that there is no output. This is black.

However, owing to undershoot, or non-standard use, the video signal may occupy the range 0.0V to 0.7V. In which case, the brightness setting used above will clip anything from 0.0V to 0.0525V to black.

The AVIA disk follows the standard. The voltage ouput at 7.5 IRE is 0.0525 V (or digital 16). The voltage ouput at 100 IRE is 0.7V (or digital 235)...or should be if the playback hardware is calibrated correctly.
post #77 of 84
So, I just need to measure the output voltage at the DVDO iScan Pro output and also the DVD-player. Umm, how does one do this? Can a simple multimeter be used, or does one actually need a scope to do this?
post #78 of 84
It is MUCH easier to analyzing things with a scope. With AVIA waveforms you can see the little black level bars riding near black and the white level bars hanging just below white. With Avia PRO, put up the Double Cross Ramps and you see a diamond shaped waveform whose tips indicate clipping and the slopes show linearity.

Oops forgot AVIA in the first part.---
post #79 of 84
Avia Pro is for a professional, of which I am not.
post #80 of 84
Wanman- Multimeters are for DC and RMS values of low freq AC (mains).
A video waveform needs a 'scope, if you're going to do it that way.
post #81 of 84
Boy, twenty-years have passed and my memory is SOL.
post #82 of 84
For what it's worth- on my PJ with 0V across RGB port 3 inputs, R G and B tubes are very,very faintly lit, if I look in a pitch black room at the phosphours. No chance of playing shadow puppets.
post #83 of 84
John Alison...

Yes, the Marquee designs are very flexible. They have 80 (or is it 99? I forget...) channels to store complete setups in, and each one stores all the data for everything from which of the four color temp alignments to use all the way to brightness, contrast, tint, detail, masking settings, and full convergence data, all for each individual channel. The only hard part is getting used to the idea of actually going to the right channel before you start tweaking your settings. If your setting is perfect for a given source but a little off for another source, go to that second source's own channel rather than try to adjust your first channel to it, or you'll have to do it all over again.

post #84 of 84
CJ- Sounds interesting, from an academic point of view- I could be wrong, but Marquees are pretty thin on the ground over here. Good second-hand Barcos are the thing. I'm not familiar with the versions that have a built in scaler- they might offer a similar option.
New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: CRT Projectors
This thread is locked  
AVS › AVS Forum › Display Devices › CRT Projectors › Black bars? Well, not quite, more dark gray.