Setting contrast with Spears and Munsil disc - Page 2

FYI calman v4.0 supports runs up to max white (109% in 8 bit, 109.5% in 10 bit).

That way you'll be able to visualize WTW clipping.

dsskid

This is a very detailed post and instructive post. Thank you for the time and information


Bringing this back around to my situation. I was not describing a hard clip at 235. I. My scenario was at some point lets say 245, green begans to roll off but it is not something you can see in a contrast screen. The white does not clip, sure from 245 there has to be a change in RGB balance. But at least to mey eyes it is not apparent either on the contrast screen in Spears and Munsil or in gray ramps. It is only on the pattern clipping test screen where you see that you do not have the ability to see full set of steps in green above 235. I guess I need to put a meter on 109% saturation which I believe is 254 and see what the DE is.

On the Mits set I am using, there is an advanced most with traditional RGB High and Low controls to set grayscale and a gamma control. Unfortunately the CMS in that mode is minimal and not very useful. In the non-advanced mode you have a very useful CMS but you only have access to single level control for each of RGB for grayscale and no gamma control.

I am using a DVDO Duo to adjust the grayscale and gamma. It is an 11 zone system and while you can get grayscale 0-100 to measure nearly perfect with very low Delta E and the gamma of you choice, you have no contol of what happens beyond 100% saturation. With the built in grayscale and gamma controls in the advanced mode I believe you woud maintain your gamma pretty much right out to the top white, but you would have color errors you could not fix. The Duo can set xy on primaries but can not do anything with secondaries.

I am using the non-advanced picture mode Natural as my starting point, setting the RGB white balance as closely as I can using the Service Menu setting, touching the grayscale and gamma up with the Duo. I then adjust primaries with the Duo. I finish the color with Mits CMS. Of course the problem with this is if the gamma, and grayscale go off track above 100 %, the only control you have is to reduce contrast. I am mesuring using the built in paterns in the Duo which do not include a 109% pattern.

I tried again last night to adjust my Gamut Delta Es. I was able to get Green under 3. The picture looks good, but I am still seeing the clipping in Spears.

I guess now the question becomes in calibration should I be able to not only make out the last box on a contrast pattern, but fully make out the full detail of the box and the number perfectly. I mean besides in an ideal world.

That might be useful for me. But unfortunately I am not sure how much control you have once you visualize it.

I have ChromaPure now. Is this functionality going to be in the Home version of 4.0?

If we take white clipping as an example:

What you see is rather than a smooth roll off up to the brightest parts of the image you get a flat bright "hole" where there is no intensity variation just a sudden step up to peak brightness (think mountain tops being chopped off flat).

If you take black clipping as an example:

Instead of ever darkening detail visible smoothly down to black you get a sudden drop off into flat featureless black.

Now your eye will eventually do this naturally , eventually you hit a point where you cannot differentiate black and a little bit above it or white and a little bit below it . (you see far more towards black than towards white...something to do with checking caves for sabre tooth tigers probably)

However the problem with synthetic images is they tend to get to this stage quicker than your eyes do and your visual system can perceive this artificiality ( aesthetically images that do not look clipped tend to be preferred).

Film gets much closer to giving your eye an intensity range that is comparable to real life than other forms of capture so aesthetically it tends to be viewed as looking more favourable in a wider range of exposure situations than video for example.

Individual color channels that clip ( if we are referring to peak levels) are usually down to display limitations but conversely they are useful in defining the true peak white performance of a given display if you assume that you don't want to set a clip point lower than 254 in terms of mapping a video signal.

So contrast should go as high as you can set it without clipping in any channel or burning out your eyeballs whatever comes first.

This is a good thing. If you can't detect errors visually, the error isn't that great to begin with. Fortunately, errors on the high end aren't as noticeable as errors on the low end.

It sounds as if your display maxs out fTls somewhere around 23-25 when properly calibrated, but due to your situation, you're looking to increase light output further, which is causing you to clip white. A lot of properly calibrated RP displays have to overwork to get to the 30 fTls range. There are limitations with displays and equipment.


Short of upgrading your display, which we established is not an option for you at this time, compromises need to be made.

If keeping contrast at a level where 254 is still slightly visible leaves your display too dark for enjoyable viewing, I recommend clipping some of the white level above 235, as long as your whites don't become noticeably blue or pink (or green, which doesn't seem to be a problem if you're clipping green).

You are still keeping the important range of 17-235, and you're still able to enjoy your RP display in a bright environment.

Is it perfect? No. But no display is absolutely perfect, and if you you can't notice any color shift when pitting up an 11 step grayscale ramp, chances are you won't have a problem viewing regular programming content.

Truthfully, noone is going to pop the spears and munsil disc pattern up on your screen and say, "Look, I can see your green gain is somehat clipped".

You have to accept the limitations of the display, maximize it's potential, and still be able to making watching it an enjoyable experience. You cannot calibrate your way to a new display.

When you're ready to upgrade, the technology will be newer and better, and the issues you are having won't be an issue anymore.

Bottom line, watching TV is supposed to be enjoyable, so pop some popcorn and enjoy it.
-Best,
John


P.S.

No, the last box (255) should blend in, and ideally you will slightly make out box 254.

All versions, but it will be dependant on your pattern source.

Contrast and RGB balance should help you change what's happening above white. Maybe you can do something about it, maybe not. But it will be a more educated decision than currently can make.

So when the patterns look pink that means you are clipping red at the extreme end.

Your eye locks onto the birghtest thing and says that is white. Since red is clipping and the brightest feild has a blue hue but your brain says that's white, everything else appears as if it has a red hue.

Turn down the contrast till the hue shift goes away.

You may have to go back and forth between contrast and clipping patterns until the pink hue is minimized if not totally eliminated, and the red in the clipping pattern does not remain solid and you can see the inner square.

the pinkish color isn't becasue <100% is red, it's an optical illusion because Max white (255) is too blue. Because red is clipping.

If you think the image without clipping is too dim, then you'll need to compromise.

Only you can decide how bright or how much clipping is acceptable.

I'm glad this thread came up. I've been having issues with clipping on all colors.

I am using the InFocus SP8602 and the spears and munsil Cal disc (BD). I clip white with the contrast image up at 234, no matter how low I set the contrast. Even contrast set all the way down is clipped at 234, the image is just darker. AND when I go to the clipping patterns image, all colors are clipped including white. not one square is visible in any of the W/G/B/R rectangles.

Ive set everything and retierd everything on all settings. The only way I can get bars up to 252 visible is to adjust picture through the blu ray player (pioneer elite 23fd) and by doing this causes banding.

Any thoughts on this? I also was in contact with spears through email but he quit emailing me. I'm guessing because he was at a loss.

Picture looks fine though.

When you say you tried everything, I assume you included Gamma adjustments as well?

is it going through a yamaha reciever?

There are a few devices that hard clip the signal.

What player are you using, the PS3 doesn't ouput >235 unless superwhite is turned on, and more than a few HTPC configurations won't either.

I would not use gamma to set black and white point.

It is going through a pioneer elite Sc27 paired with a pioneer elite 23fd bluray player. All connections through hi speed HDMI cables. I know there are no settings on the projector such as superwhite, head room, etc... to open the contrast range to reference white. I have looked through the manual several times on both the bluray player and reciever and still cannot find any of the settings that I listed above. Maybe it is in the projectors service menu, but I'm not sure and am not going to mess with anything in the service menu as I am not a professional.

I can however use the memory setting on the blurray player to adjust white level. by doing this, I can adjust contrast to see all of the contrast bars, but there is bad banding on the gradiant bar.

I'm at a loss.

If the picture looks fine then why are you concerned?

How much money are you willing to spend and how much work are you willing to do to diagnose (not fix, just diagnose) a defect which doesn't appear to be problem for you?

yes gamma adjustments do nothing for the clipping image.

Everything is under warranty so no money will be spent if something is in fact wrong. I'm mainly asking out of curiousity. I just find it odd that the projector is clipping so low (234).

I'm talking about using it for the color-clipping pattern. Where it can have an affect.

As Sotti mentioned, something in your video chain is hard clipping the signal. The first thing to try is to run the Blu-Ray player directly to the projector if possible and eliminate the receiver from the equation. See if you get the same results. You may also want to do a bit of research in the receiver forums to see if that Pioneer receiver has any combination of settings, etc that result in clipping of the signal. The next thing I'd try is another Blu-Ray or DVD player that has an HDMI output in place of the Pioneer and see if you get the same results. Basically, you need to solve the problem of which which device in your chain is causing the clipping. Once you know that then you can see if there is anyway to fix it. In the end, unless you change your setup, you may have to live with it.

hope this helps,


--tom

thanks tom,

I actually hooked my Panny bluray directly to the projector with the same results which makes me think it is a setting in the projectors service menu. this also is a brand new projector and the first version of firmware was just released. I will install the firmware to see if possibly this fixes it.

1) There's a difference between diagnose and repair. In warranty repairs may be free but collaborative diagnosis is not since it takes time if nothing else.
2) When people help you they're giving up their time and it's lost forever. Perhaps if you're just curious you should take the answers you've been given as sufficient.
3) Contact the vendor and see if it's expected. If not place a service call.

Whether you are setting luminance or individual RGB channels for black and white ref I would not use a gamma control to set them.

Doesn't seem like this is the issue anyway.

I tried going through some random commercial DVDs I own a couple times, and in the end I personally could not find anything that I could say was clearly intended to be brighter than 235 gray. Likewise in reading the forum I haven't ran across many posted examples. The standards do allow for information above reference white, but I'd have to say that it seems rarely intentionally used in commercial movies. If there are clear examples I'd still like to see them, but at this point I'm familiar with more practical considerations than any real-world video examples that might put much emphasis on maintaining a tight grayscale above reference white.

But you are setting up clipping in the RGB (RGB gains) space wich is very different than YCrCb space.

a YCrCb value of 215,225,128 is not only legal, but inside reference levels and evaluates to 255,170,215 RGB.

Not clipping from 16-255 is essential to getting a reference image.

 

RGB-YCC.zip 5.4130859375k . file

It may be legal, but not necessarily valid or in-gamut. If it evaluates to a 255, then it would be out of gamut or invalid for StudioRGB. 255 is not valid for payload data in StudioRGB. It is reserved.

Dave

+1

Forgive my ignorance, I'm focused mainly on the formula side of the equation. Where is the list of valid YCrCb that are 16-235 for Y and 16-240 for the chroma channel that are invalid?

Also I don't know what you mean by invalid and reserved for sRGB, every computer system I've ever seen goes 0-255 for R,G,B.