or Connect
AVS › AVS Forum › Display Devices › CRT Projectors › Gamma values used on your systems?
New Posts  All Forums:Forum Nav:

# Gamma values used on your systems? - Page 3

Quote:
Originally Posted by SOWK

so if i were to do this test with an i1 pro and calman, I would/should get a gamma "curve" of 1.00?

not 1.8 or 2.0 or 2.2 or 2.5?

Sowk did you note my example and how it would correlate with your 2' x 2' wooden reflectivity board analogy? These different chip chart values correlate with log film density plots.

### AVS Top Picks

Quote:
Originally Posted by SOWK

so if i were to do this test with an i1 pro and calman, I would/should get a gamma "curve" of 1.00?

not 1.8 or 2.0 or 2.2 or 2.5?

It completely depends on what you meant by the 10% increment in:
Quote:
Originally Posted by SOWK

The paints should ranging from Black to white in 10% increments.

If the 10% increments are the light power each reflects then the gamma curve should measure 1.0 because the steps would map out a straight line on a linear graph of 10%, 20%, 30%, ..., 100%. If however you meant that they should look like 10% increments to average humans then somebody would have had to have decided what shade of gray looks like 10% of white and the curve you got would depend on how much light power they mixed each step to reflect (the shade they mixed). An 18% gray card is generally accepted as the midpoint (50% mark) for video from what I've seen (and should look around half as bright as white, although it can get a little complicated). Using a strict .5^x = .18 yields a number of 2.47. Or 1/2.47 (.405) from .18^y = .5.

--Darin
Quote:
Originally Posted by tbrunet

If a given display has an ideal transfer function thats rec 709 (2.2 response @ D65) then a linear luminance condition exist.

The Rec. 709 curve is closest to 1.96 gamma (1/0.51) not 2.2, as shown here.
If you don't mind I'll go with Dr.Soneira's measurements and also the Director of Engineering of Ikegami. Both have stated clearly that a SMPTE phosphor Rec. 709 direct view reference CRT will measure when calibrated perfectly flat 2.20, all the way down to 1% of white calibration level.
Quote:
Originally Posted by SOWK

so if i were to do this test with an i1 pro and calman, I would/should get a gamma "curve" of 1.00?

not 1.8 or 2.0 or 2.2 or 2.5?

If each wooden board was labeled in IRE i.e. and you have 100 separate boards representing 1-100 IRE with the last board representing rec 709 reference white, that would be a logarithmic progression of luminance for your meter, but made into a linear progression or (Y’) video signal voltage. Per SMPTE standards for consumer (YPbPr) digital video:

(Y,Cb,Cr) 16,128,128 = 0V = 0 IRE black
(Y,Cb,Cr) 235,128,128 = .7 Volts = 100 IRE white

1 IRE (gray wooden board) or 1% stim (8-bit digital coding ~18,128,128)
100 IRE (White) exactly 235,128,128

each progressive step (~1/2 IRE), or 2.19 per 1%

The gamma corrected video grayscale would progress linearly (0-.7 volts) from blk-wht due to gamma compression.
Quote:
Originally Posted by tbrunet

If you don't mind I'll go with Dr.Soneira's measurements and also the Director of Engineering of Ikegami. Both have stated clearly that a SMPTE phosphor Rec. 709 direct view reference CRT will measure when calibrated perfectly flat 2.20, all the way down to 1% of white calibration level.

Wait, are you now arguing that measurements from a professional CRT monitor are more accurate than the official spec?

2.2 gamma is not equivalent to the Rec. 709 curve.

Section 1.2, on page two of the document detail the Rec. 709 curve.

You are allowed to get three free papers from the ITU a year, so it won't cost you anything: http://www.itu.int/publications/book...-buy.html#free
Quote:
Originally Posted by andrewfee

Wait, are you now arguing that measurements from a professional CRT monitor are more accurate than the official spec?

What I am saying, if one wants to “emulate” a Pro Sony/Ikegami direct view SMPTE mastering reference CRT, you will indeed measure after calibration 2.20 gamma expansion. It will measure that way over practically 99% of its response.

If the content encoded is 2.2 and nominal balanced conditions and the monitor is calibrated 2.2 @ D65 then camera-eyeball is indeed 1 (uniform) not 1.125
Quote:
Originally Posted by tbrunet

If you don't mind I'll go with Dr.Soneira's measurements and also the Director of Engineering of Ikegami. Both have stated clearly that a SMPTE phosphor Rec. 709 direct view reference CRT will measure when calibrated perfectly flat 2.20, all the way down to 1% of white calibration level.

Dr. Soneira's own measurements here:

http://www.displaymate.com/ShootOut_Part_2.htm

show 20% input at about 2.4% of white in Figure 2. A strict 2.2 gamma of .2^x = y would result in 2.9% of white, not 2.4%. Even 2.3 gamma would result in slightly more light than he measured. And at 10% a 2.2 from the same equation would result in about .63% of white, but according to his Figure 1 he only got about .47% of 100% video level (from .83/176), so higher gamma than 2.2 and even a little higher than 2.3 there. That isn't even counting that he did not set the display up as it would be setup for mastering and hadn't measured one in a mastering environment (and setup for mastering) when I asked him.

And the information I have is that Pixar sets up their in house DLPs to 2.4-2.45 to match their BVMs.

--Darin
http://media.panavision.com/Referenc..._explained.pdf

This document is from Panavision, it backs up my conjecture 100%. Rec 709 camera gamma is ".45" and not 1.96.

If possible please note where I've misrepresented any info from this (Panavision) manufacturer reference.
Quote:

For a number of years, video camera manufacturers have provided a non-standard extension to the standard video gamma function known as “knee” and “slope” control. The concept is that the basic 0.45 power function could be modified by creating a new point gamma function or slope that could be initiated at a particular video level or “knee point”. Unfortunately, this deviation from the 0.45 gamma function introduced an unpredictable change in the camera transfer characteristic that usually resulted in transitions in video levels that were often at the nominal skin tone levels and appeared unnatural
See the sig...oops, too late.
Quote:
Originally Posted by tbrunet

This document is from Panavision, it backs up my conjecture 100%.

It doesn't back up your conjecture about gamma being 2.2 for the display. You've quoted Poynton a lot. Do you disagree with him that the end-to-end gamma should be higher than 1.0 because people tend to prefer that? If the camera gamma is basically the inverse of 2.2 (the .45 you just posted) then it takes higher than 2.2 at the display to do what Poynton said. So, what is your position? Is it that the camera is the inverse of 2.2, the display is 2.2 and so the end-to-end is 1.0 and not the higher than 1.0 that Poynton claims?

--Darin
I quoted Poynton already in this thread and Mr. Poynton, Sony, Barco, HP and Microsoft all agreeing on the manufactured tolerances for CRT and the predictability of their phosphors basically make 2.2 @ D65 the target. This is the reference the colorist was experiencing until recently.

I quoted Mr.D stating the exact same target.
fwiw here it is for your reference
Quote:
Originally Posted by tbrunet

HP and Microsoft recognized the need for a common internal color space and also recognized the need to create a common exchange space. Several principles that guided that development also provide good guidelines for making design trade-offs:
Respect legacy content and practices.
Ensure compatibility with professional, well-calibrated CRTs and HDTVs.

Sony provided studies that proved optimally calibrated CRTs in their native state are 2.2 gamma and have HDTV primaries. [Katoh and Deguchi, "Reconsideration of CRT Monitor Characteristics"]

Charles Poynton showed in HDTV development that the HDTV phosphors were a reasonable compromise between CRT manufacturers and within the manufacturer tolerances. [Poynton, A Technical Introduction to Digital Video]
*Barco, Sony, and other high-end CRT manufacturers confirmed these results.

Hewlett-Packard provided studies and computations to prove that gamma 2.2 is very close to a perceptually uniform lightness scale for display viewing conditions. The mathematics illustrating this can be found on the Hewlett-Packard sRGB web site at http://www.srgb.com.
Quote:
Originally Posted by SOWK

so if i were to do this test with an i1 pro and calman, I would/should get a gamma "curve" of 1.00?

not 1.8 or 2.0 or 2.2 or 2.5?

There is no curve at all. As I said, your question doesn't make any sense. If you are measuring light directly, you'll get the measurement of that light directly. There isn't any other relationship involved to which you can compare two different quantities. If you measure some amount of light, it just is what it is. There is no other signal value to which you can compare that in order to characterize that relationship with some mathematical function. Gamma is just a math function that relates two different scales, or you can think of it as creating a new scale with some set of values. If you've only got the light to measure, that just is what it is, there isn't any function involved. Even saying that it will come out as a gamma of 1 is rather nonsensical.
Quote:
Originally Posted by tbrunet

What I am saying, if one wants to emulate a Pro Sony/Ikegami direct view SMPTE mastering reference CRT, you will indeed measure after calibration 2.20 gamma expansion. It will measure that way over practically 99% of its response.

If the content encoded is 2.2 and nominal balanced conditions and the monitor is calibrated 2.2 @ D65 then camera-eyeball is indeed 1 (uniform) not 1.125

That's hogwash. A BVM has a natural gamma higher than 2.2.

Go freaking measure one.
Interestingly, going back to the what is real world gamma question, I've just added Lstar (our eye's gamma) to the spreadsheet I created for this sort of thing.

It seems that the best fit curve is the Rec. 709 curve with an end-to-end gamma of 1.25.

Its equivalent as a power curve would be 2.45. (though the power curve is too dark towards black)

Quote:
Originally Posted by tbrunet

I quoted Poynton already in this thread and Mr. Poynton, Sony, Barco, HP and Microsoft all agreeing on the manufactured tolerances for CRT and the predictability of their phosphors basically make 2.2 @ D65 the target.

No, Poynton did not as far as the 2.2 part in what you quoted. You quoted a reference from 2001 from somebody else which did not include Poynton saying anything about 2.2 for the display and Poynton elsewhere has talked about 2.2 for a brighter environment, but also made it clear himself that he does not believe there is a standard for the display and that he does not believe studio monitors in their environments are 2.2. He even has this as one of the first things on his own website at www.poynton.com where it says:
Quote:
I continue to be flummoxed by the absence of any viable, modern, realistic standard for the "gamma" — or properly, electro-optical conversion function (EOCF) — of studio reference displays for video and HDTV. With the demise of CRTs, and with the imminent introduction of reference-grade flat panel monitors, I believe that such a standard is now critically important. More information is available.

and his paper after the link says on page 3:
Quote:
In a CRT display, the electrostatic characteristics of the electron gun cause the CRT to impose an electro-optical conversion function (EOCF) that is approximately a 2.4-power function from voltage input to light output. The symbol γ (gamma) represents the exponent at the display: A studio reference display is said to have gamma of about 2.4.

and near the end:
Quote:
I have concluded that professional studio displays in use today for SDTV and HDTV, as configured for studio control room environments, have exponents very close to 2.4.

I'm not saying I know Poynton to be right, but lets not misrepresent his stated position.

Back to my question:
Quote:
Originally Posted by darinp2

You've quoted Poynton a lot. Do you disagree with him that the end-to-end gamma should be higher than 1.0 because people tend to prefer that? If the camera gamma is basically the inverse of 2.2 (the .45 you just posted) then it takes higher than 2.2 at the display to do what Poynton said. So, what is your position? Is it that the camera is the inverse of 2.2, the display is 2.2 and so the end-to-end is 1.0 and not the higher than 1.0 that Poynton claims?

Do you disagree with Poynton's position about greater than 1.0 total or not? The one stated on his website at http://www.poynton.com/notes/colour_...aFAQ.html#NTSC
Quote:
If an images's correct physical luminance is reproduced in a dim surround, a subjective effect called simultaneous contrast causes the reproduced image to appear lacking in contrast, as demonstrated above. The effect can be overcome by applying an end-to-end power function whose exponent is about 1.1 or 1.2. Rather than having each receiver provide this correction, the assumed 2.5-power at the CRT is under-corrected at the camera by using an exponent of about 1/2.2 instead of 1/2.5. The assumption of a dim viewing environment is built into video coding.

And based on what he says in section 13 about reducing the end-to-end gamma if viewing in a brighter environment, and his position about 2.4 or 2.5 in a dim surround, it would be logical to assume that in a completely black environment Poynton would be in favor of at least 2.4 or 2.5 for the display and probably a bit higher (reduce display gamma if going brighter for the room and increase display gamma if going dimmer for the room).

--Darin
One other thing I was thinking about, if a movie is mastered for DCI with 2.6 assumed for the display gamma and then compressed for Blu-ray without modifying gamma characteristics, something close to 2.6 could get the closest to the way it was mastered. Of course there are the bit depth issues, but that doesn't mean that anybody went in and modified things specifically for a different gamma. I don't know how often care is taken in this area when going from DCI to Blu-ray versus when it isn't though.

--Darin
I may be mistaken here but I’m the engineer here with the more sensitive test equipment like my state of the art Tektronix/Leader SMPTE reference analyzers. I’m a 30 year veteran of this discipline as well.. and I have the camera.

Now I do my work in an educational environment .

Above I’ve attempted to illustrate the standard rec 709 gamma and how a tv camera RGB balance is calibrated under known light condition which btw is extremely linear at 2000 lux but we deviate “on purpose” and by design.. the pre emphasis done for nearly 100% major motion picture are not rec 709. Above I have fundamentally and scientifically color science wise described how “ideal perceptual uniformity” is propagated and ‘baked in’

camera gamma (.45) & CRT @ 2.2 results in a linear grayscale system or rec 709 unity

Test Patterns are nominal .45 grayscale representations
Grayscale ramp linear 16-235 luminance Y'(0-.7V)

Thomas
Oh, man... there's nothing like a good gamma flamewar. Except maybe a volts-vs-amps-vs-watts flamewar populated by 50 guys all claiming to be electrical engineers, and who all have totally different explanations for everything...
Forgive me for my ignorance but So far from what I can gather in this thread ,

According to Thomas (30 years in this industry)
There is A defined Gamma standard for setting up our Displays/Projectors correctly D65/709, which is 2.2,

But on the other hand Poynton? Darin Chris W and others are saying for D65/709, A gamma target of 2.4-2.5 is correct.. no?

At the moment for Blu-Ray movie play back I aim for A gamma target of 2.35-2.4/ D65 709 for my CRT projector (CalMan/ i1pro) which looks right to my untrained eyes...

I woud be very interested to here thoughts from some one like D6500Ken/ Ken Whitcomb
on what Gamma target he aims for when calibrating A highEnd CRT like A Colour corrected G90 or Cine9...

Cheers...
Quote:
Originally Posted by tbrunet

Above I've attempted to illustrate the standard rec 709 gamma and how a tv camera RGB balance is calibrated under known light condition which btw is extremely linear at 2000 lux but we deviate on purpose and by design.. the pre emphasis done for nearly 100% major motion picture are not rec 709. Above I have fundamentally and scientifically color science wise described how ideal perceptual uniformity is propagated and baked in'

camera gamma (.45) & CRT @ 2.2 results in a linear grayscale system or rec 709 unity

Test Patterns are nominal .45 grayscale representations
Grayscale ramp linear 16-235 luminance Y'(0-.7V)

Thomas

So why does the Rec. 709 paper claim that the transfer function is:

If it's really just 0.45 ?

And why is 2.2 gamma at the display ideal perceptual uniformity when it doesn't match up that closely to Lstar?

Why do so many sources claim that the target is an end-to-end gamma greater than one, and not 1.0? (linear)
Quote:
Originally Posted by tbrunet

camera gamma (.45) & CRT @ 2.2 results in a linear grayscale system or rec 709 unity

Why are you refusing to answer my question about whether you agree or disagree with Poynton that linear is not what people want? It is funny how you claim I am wrong about something I didn't say. One of the main questions is whether linear end-to-end is what people should shoot for. Poynton has made it very clear that he does not agree with that and you refuse to say whether you agree or disagree with him, but instead try to tag the 2.2 to him as if it is his position because somebody else wrote a paper in 2001 who said that 2.2 was correct and had a supporting sentence about Poynton, but not about 2.2, and Poynton didn't say that he supported 2.2 for the video and TV in dim environments (in fact he has flat out said that he disagreed with 2.2 on the display side).

Seriously Thomas, you should be able to read and since you say that you haven't been purposely misleading anybody here you should be man enough to look at the evidence (including what Poynton himself said) and not try to make people believe that his position is the 2.2 is correct.

Just wanting to stick to something you had interpreted a certain way before more evidence came in just because you want to be stubborn is not a good excuse for pushing something and goes against your claim that you haven't purposely tried to mislead anybody here. Sticking with something for no good reason other than you held that position at some point is purposely misleading others when it goes against all the evidence.
Quote:
Originally Posted by tbrunet

I may be mistaken here but I’m the engineer here with the more sensitive test equipment like my state of the art Tektronix/Leader SMPTE reference analyzers. I’m a 30 year veteran of this discipline as well.. and I have the camera.

This isn't about the camera, it is about the display. And interesting that you say here that you have the more sensitive test equipment when the excuse you have stuck with for years for telling me that you would measure on/off CR from your CRT to see if you were right and then never doing it was that you didn't have measurement equipment sensitive enough. So, do you now magically have measurement equipment sensitive enough to measure gamma for a display? If not, then please don't push that people should believe you because you have the test equipment when it is test equipment for the camera side, which people weren't arguing about. It comes down to that fairly basic thing that you seem to want to avoid answering now. That is, should the end-to-end be linear. Poynton says no. If you think Poynton is wrong then say so.

As far as:
Quote:
Originally Posted by tbrunet

I’m a 30 year veteran of this discipline as well.

some people on the CRT forum may not be familiar with tbrunet's tactics. He uses this thing pushing that people should believe him because of how many years of experience he has in this discipline even though he has posted some of the most ignorant stuff around here about subject matters related to this and when things were pointed out to him he would stick with those ignorant positions for a while, using some pretty low tactics in order to try to get people to believe that he was right. It is kind of like somebody saying that everybody should believe them about some math thing because they have 30 years experience in math when in the last few years they spent months trying to get people to believe that 9 times 9 is 18 and that anybody who said it was 81 didn't know what they were talking about, all while refusing to answer what 3 times 3 is or actually count some things. Like here, I don't think a person is going to go from not being able to figure out what 9 times 9 is to understanding calculus in a small amount of time.

And tbrunet still hasn't shown that he understands on/off CR. He never has apologized to Greg Rogers for claiming in post after post that Greg was wrong for thinking it was possible for on/off CR to be higher than 219:1 or 980:1 with consumer video. tbrunet would avoid the simplest questions about the subject matter that would have made it more clear that he was wrong, yet we are supposed to believe that he is right here because he has experience (experience which wasn't enough to get him to understand in less than 6 months how video from a display could have higher on/off CR than 219:1 or 980:1).

It was almost exactly a year ago where tbrunet was backing up that gamma doesn't apply to digital displays (to support somebody saying that if 100% video level was 10 ft-lamberts then 50% video level should be 5 ft-lamberts) and according to tbrunet he must have had way over 20 years experience at the time he did that. Yet a year later we are supposed to believe that he now knows not only that gamma does apply to digital displays, but that it should be 2.2.

tbrunet doesn't like it when I provide actual links to things he has claimed and stuck with even after they were explained to him, like after he claimed about 6 months ago that he had been right about everything he had posted about a subject matter related to this one even though he knew he had pushed for a long time that consumer video couldn't provide an on/off CR higher than 219:1.

Here is a link where I have included links to some of the things that tbrunet has said over the last 4 years and some of his claims:

http://www.avsforum.com/avs-vb/showt...6#post16411186
Quote:
Originally Posted by darinp2

For a short (partial) synopsis, here have been some of his claims:

11/2/05: The highest CR possible with 8 bit video (video black at 16) is 219:1.
11/3/05: If white is 30 cd/m2 then the darkest black that a CRT could do with 8 bit video is .15 cd/m2.
7/30/06: Claimed that Dr. Soneira had validated his position (which Dr. Soneira responded to 2 days later, telling tbrunet that this wasn't true).
7/12/07: After telling us that he understood this subject matter and we didn't and avoiding my question about how bright a 10%stim (or 10 IRE with 0 base) level should be if white is 10 ft-lamberts, answered that 10 IRE would be 1 ft-lambert in that case.
8/1/08: Backed up somebody else claiming that gamma doesn't apply to digital and so a 50% video level (50%stim or 50 IRE) would be 5 ft-lamberts if white is 10 ft-lamberts.
1/29/09: Claimed that John Luff of SMPTE had validated everything he had ever posted on the subject.

If people follow that stuff they will see that even this year tbrunet refused to answer simple questions about that subject matter and in the end tried to change the subject from on/off CR in order to try to keep people from realizing that he had been wrong about it. Still wouldn't be a man and admit he was wrong, but instead tried to make it look like others had been wrong while refusing to discuss that actual subject matter.

While some may think it isn't fair to actually point out a person's history, I think it is when they are trying to use their past (their experience) to get people to believe that they must be the ones that are right and have done this before.

And if tbrunet follows his pattern he will refuse to answer simple questions and push whatever it was he thought at some point, regardless of whether it is true or not. Like here, I don't expect him to actually answer whether he thinks Poynton is right or wrong about the greater than 1.0 end-to-end thing, or even if he doesn't know whether Poynton is right or wrong, which would contradict taking a position that things should be exactly 1.0 end-to-end. If tbrunet thinks that 1.0 end-to-end is correct then he should stand up and say that Poynton is wrong instead of trying to make people believe that Poynton's position is that the CRTs are 2.2.

--Darin
Quote:

According to Thomas (30 years in this industry)
There is A defined Gamma standard for setting up our Displays/Projectors correctly D65/709, which is 2.2,

But on the other hand Poynton? Darin Chris W and others are saying for D65/709, A gamma target of 2.4-2.5 is correct.. no?

I'm not going to claim I know 2.2 to be wrong and 2.4-2.5 to be right for the display side. I haven't been able to measure gamma for a CRT as setup for mastering and while I suspect that it is higher than 2.2 in general based on the things I have gathered I can't say for sure. And the data tbrunet has gathered doesn't mean that 2.2 is correct. Especially given that he thought Poynton supported that (or at least tried to make it look like Poynton did) when Poynton's own words have contradicted that. I have talked to Charles Poynton and tried to get more information about the gammas as setup for mastering, but he hasn't gotten data from mastering in Hollywood either (at least he hadn't at the time). He had some other data that supports the higher than 2.2, but not from a mastering house in Hollywood with a CRT monitor setup for mastering. I asked Dr. Soneira about his measurements in an email and he told me he hadn't set the CRT up for mastering, which goes along with his 176 cd/m2 measurement for white (since that is much higher than is standard for mastering). His measurements also have closer to 2.3 down low, not 2.2. They are 2.2 around 40% video level and higher, but down lower it looks like they are a hair above 2.3 gamma to me. I don't know how they would change if he set the monitor up for mastering.

Also, as far as gamma for mastering it depends on whether only the display itself is considered or the ambient light also as far as what is natural for a CRT. Raising the black floor off the display with other lighting changes the calculated gamma if that light is included in the measurements. Things can get a little complicated as far as what to do if then watching with less ambient light at lower light levels, even if a person had one of the mastering monitors. Few people here are going to be watching at the 30 ft-lamberts plus that is standard for mastering and I think most people here are going to have less room lighting on.

--Darin
Quote:

Forgive me for my ignorance but So far from what I can gather in this thread ,

According to Thomas (30 years in this industry)
There is A defined Gamma standard for setting up our Displays/Projectors correctly D65/709, which is 2.2,

Yes, according to a man who holds more absolutely ridiculous ideas that I can even list here.

His claim is very easy to test. He claims that there is a "defined gamma standard" as you put it, for displays of 2.2. Well, if that were the case, it would be fairly simple to reference what that standard is, provide a link, an excerpt, or even the title of that standard. He has provided no such reference because such a standard does not exist.

The closest he came was claiming that Rec709 provided a display gamma in the spec, but it does not, he was lying as usual. Rec709 is silent on the subject of the display gamma. It only describes the encode gamma, which is effectively ~.5, and was just posted above by andrewfee.

Quote:

But on the other hand Poynton? Darin Chris W and others are saying for D65/709, A gamma target of 2.4-2.5 is correct.. no?

At the moment for Blu-Ray movie play back I aim for A gamma target of 2.35-2.4/ D65 709 for my CRT projector (CalMan/ i1pro) which looks right to my untrained eyes...

I woud be very interested to here thoughts from some one like D6500Ken/ Ken Whitcomb
on what Gamma target he aims for when calibrating A highEnd CRT like A Colour corrected G90 or Cine9...

Cheers...
I also want to be very clear, as I think I've said in other gamma-related threads, that I do not have a problem with a gamma of 2.2, and it can be an appropriate choice depending on the viewing environment, the display, and viewer preference. However, that is very different than saying that the standard display gamma should be 2.2 or is stated as such in any video standards, when that is patently false.
Quote:
Originally Posted by ChrisWiggles

I also want to be very clear, as I think I've said in other gamma-related threads, that I do not have a problem with a gamma of 2.2, and it can be an appropriate choice depending on the viewing environment, the display, and viewer preference. However, that is very different than saying that the standard display gamma should be 2.2 or is stated as such in any video standards, when that is patently false.

I agree with everything you've said there. I would go one step further though and say that 2.2 is the minimum acceptable gamma for good image quality, but not necessarily the ideal target.

Anything above 2.5 is too high in my opinion, and I'd even go as far as to say that 2.5 is too high as well if you're using a simple power curve.

I also wanted to point out that, for anyone using CalMAN (I noticed it was mentioned here) if you switch to the ITU/EBU function as your target gamma (the Rec.709 curve) you have to change your exponent to compensate for the way they calculate it. You have to multiply the target by 1.133 for the equivalent to a power curve and ignore values below 9% grey. (E.g. if you wanted to use the equivalent of 2.2 power, you would need to set the target as 2.5)
Andrew my question is why should we trust your interpretation if you don't comprehend the digital (color space) workflow.

When a digital editing platform linearize the Y'R-Y, B-Y luminance data for processing it's model and transform is based on a .45 gamma profile exactly. For nearly 100% of the data excursion (1-100 IRE)

The perceptual model is scaled so that a given camera's matrix profile will align with SMPTE/ITU-R Recommended Practice.

The Rec 709 profile crushes exposure latitudes I've have shown this to be a fact per Panavision's perspective anyway.

No one makes use of the standard default 709 gamma profile for motion pic releases, for real world cinema productions. One emulates a log gamma profile aka a filmlike look' within a Rec 709 space. The recording does not represent .51 unity Rec Camera default gamma conditions.

DVD/Blu Ray disc have different recording gamma. Andrew I can explain how and more importantly 'why'.
Quote:

I agree with everything you've said there. I would go one step further though and say that 2.2 is the minimum acceptable gamma for good image quality, but not necessarily the ideal target.

Anything above 2.5 is too high in my opinion, and I'd even go as far as to say that 2.5 is too high as well if you're using a simple power curve.

Right, I think 2.2-2.5 is an appropriate range for good display gamma, with lower gamma preferred for viewing situations with ambient light, and higher gammas closer to 2.5 preferred for light-controlled rooms with displays of significant on/off CR capabilities that can better support a higher gamma, and also enough system ANSI CR so that shadow details are not overly obscured.

But I would say that the effective standard, implicit, is on the higher end of that range based on my understanding of mastering bays, a BVM I measured, and what others have told me.
http://www.poynton.com/PDFs/Rehabilitation_of_gamma.pdf

The Rec. 709 transfer function is standard for 525/59.94 and 625/60 conventional video, and for HDTV. The Rec. 709 function is based on a power function exponent of 0.45

I’ll attempt to paraphrase Mr. Poynton’s comment here:

Quote:

Poynton:
“Rec. 709 appears to strictly define the transfer function at the camera. However, real video cameras have controls that can alter the transfer function. “

“These controls are routinely used by cinematographers and videographers to achieve their artistic intents. Obviously the artistic intention of the cinematographer must be imposed at the camera, not at the display – it ought to be the displays that are standardized, not the cameras! But there is no mechanism to impose standards on displays, so we standardize the reference transfer function at the camera instead. In effect, Rec. 709 is standardized so as to produce acceptable reproduction on a conventional display. Despite the lack of standards, CRT displays are tacitly considered to have similar response.”

Andrew should we redesign software and editing suite? So that a .51 model is the default perceptual model? Is .51 pre emphasis the same as a Rec 709 profile? .45 or .51 “average gamma” over the largest part of the defined dynamic range?
New Posts  All Forums:Forum Nav:
Return Home
Back to Forum: CRT Projectors

### AVS Top Picks

AVS › AVS Forum › Display Devices › CRT Projectors › Gamma values used on your systems?