Spears & Munsil HD Benchmark Blu-Ray 2nd Edition - Page 16 - AVS Forum
Forum Jump: 
Reply
 
Thread Tools
post #451 of 671 Old 12-17-2013, 01:32 AM
Advanced Member
 
xvfx's Avatar
 
Join Date: Feb 2013
Location: UK
Posts: 579
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 39 Post(s)
Liked: 61
Quote:
Originally Posted by D Bone View Post

lol So 2.4 is darker than 2.2?

Quote:
Originally Posted by spacediver View Post

agreed.

Think how bad OS X originally looked before Leopard. It's gamma was 1.8. Quite cloudy grey.
xvfx is offline  
Sponsored Links
Advertisement
 
post #452 of 671 Old 12-17-2013, 04:58 AM
Advanced Member
 
PE06MCG's Avatar
 
Join Date: Jun 2010
Location: West Yorkshire, UK
Posts: 729
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 48
Quote:
Originally Posted by spacediver View Post

PE06, please read the replies carefully. A number of people in a number of different threads have patiently answered the same question repeatedly. Everything you are asking is answered in Don's post

Thanks for your advice.

I notice the same trait with yourself. Certainly if persistence is a fault I plead guilty.

May I suggest you read my post carefully in the same way you expect other people to read yours.

I'm sure the forum police do not need any help from either of us.
PE06MCG is offline  
post #453 of 671 Old 12-17-2013, 08:47 AM
AVS Special Member
 
dmunsil's Avatar
 
Join Date: Mar 2000
Location: Kirkland, WA, USA
Posts: 1,078
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 24
Quote:
Originally Posted by PE06MCG View Post

Obviously things may have changed since Chris's post on CC but you did say that you agree with Chris in your post?
Perhaps I have misread this part of his post?

I would call it a matter of degree. Yes, there are "mastering houses" setting things up poorly. However, the biggest houses like Deluxe and Technicolor and Warner have always used BVMs, and are now moving to flat panels calibrated to look like BVMs. I know this for sure. I know some of the people who do those calibrations. They have always mastered to 2.4.

Now, Chris's concern is that not all content is done at the big houses, and there are random small producers who are all over the map. And he would like everyone to be using the same standard, as would I. There is confusion out there about gamma because:

- Until recently there was no standard (other than the de-facto standard of "match a BVM") for video display gamma. It wasn't thought to be necessary, because CRTs have a natural gamma, so they all tended to look very similar.
- There has been for years a standard for computer displays (sRGB) that specifies an effective 2.2 gamma
- The encoding gamma for video cameras has been specified as approximately 1/2, and the inverse of 1/2 is 2.
- Digital Cinema uses a 2.6 gamma

You put all this together and there's confusion. The computer gamma has, I think, been the biggest issue. People doing home production are typically using computer monitors, which are usually going to default to 2.2. This means that unless they go out of their way to recalibrate, when their content is viewed on a television, it looks darker. There are some number of people who came out of the computer world who think the best answer is to get everyone to adopt the computer standard, not understanding that video has had a de facto standard for its entire history. Hopefully BT.1886 will help clear this up.

Also, keep in mind that that thread was in 2011, when BT.1886 wasn't yet finalized and confusion was perhaps higher. I like to hope that very few of the people on that thread would still be saying that video should be mastered at 2.2 in the US. I really don't even know where they got that idea, other than the whole computer thing.

Anyway, I return to my first assertion, which has not changed: big-studio movies are now and have always been mastered to 2.4. Television production at the major networks is mastered to 2.4. There are miscellaneous small producers who do all kinds of wacky things, but you can't adapt your system to deal with their foolishness, so you shouldn't even try.
dmunsil is offline  
post #454 of 671 Old 12-17-2013, 08:56 AM
AVS Special Member
 
dmunsil's Avatar
 
Join Date: Mar 2000
Location: Kirkland, WA, USA
Posts: 1,078
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 24
Quote:
Originally Posted by spacediver View Post

I hope this isn't too off topic, but what are your thoughts on the SMTPE-C primaries still being used after Rec 709 was introduced (while CRTs were still mainstream).

I don't think it's that big a deal. smile.gif The SMPTE-C primaries are very close to 709 primaries, and the primaries on consumer displays are not super perfect to begin with. Small deviations in the primaries are kind of par for the course. And it's getting better. Most of these high-end flat panels that the big video companies are moving to are either nominally 709 primaries or can be calibrated to 709 using the internal CMS. So eventually everyone will be using 709 primaries, which should happen right about the time people start mastering to conflicting wide-gamut standards. smile.gif

Actually, it looks like xvYCC has got the wide-gamut standard sewn up, but you never can tell. If some company sees it as advantageous to push a conflicting standard, they'll just go ahead and do it. It's the nature of the industry...
dmunsil is offline  
post #455 of 671 Old 12-17-2013, 09:05 AM
AVS Special Member
 
AustinJerry's Avatar
 
Join Date: Jan 2003
Location: Austin, Texas
Posts: 6,811
Mentioned: 3 Post(s)
Tagged: 0 Thread(s)
Quoted: 224 Post(s)
Liked: 646

Don,

 

I am somewhat confused between the recommended gamma of 2.4, and the results shown in the CalMAN Grayscale and Gamma Multipoint results screen.  Here are two results screens for my Sony XBR65-X900a 4K display:

 

 

 

 

 

The second result, with gamma of 2.2, seems to be clearly a better result than the first result, with gamma 2.38, based on DeltaE and RGB Balance. So, here is my question:  is the 2.2 gamma actually a better result for my display, or do the test results shown in the first screenshot indicate that I have done something wrong? 

AustinJerry is offline  
post #456 of 671 Old 12-17-2013, 09:55 AM
AVS Special Member
 
dmunsil's Avatar
 
Join Date: Mar 2000
Location: Kirkland, WA, USA
Posts: 1,078
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 24
Quote:
Originally Posted by AustinJerry View Post

The second result, with gamma of 2.2, seems to be clearly a better result than the first result, with gamma 2.38, based on DeltaE and RGB Balance. So, here is my question:  is the 2.2 gamma actually a better result for my display, or do the test results shown in the first screenshot indicate that I have done something wrong? 

I'm not understanding your question. The gamma of a BVM is approximately 2.4, which is why the video display gamma standard is 2.4. That's the standard. It remains the standard for watching video. You can certainly watch video on a display calibrated to 2.2 and it'll look fine. It's just brighter in the midtones, which wasn't what the filmmakers intended, but no small children will die. smile.gif

If you want to match a BVM, you want to get your gamma to as close to 2.4 as is feasible. If your display is able to do better grayscale tracking at 2.2, that tells you (perhaps) that the manufacturers of the display designed it around 2.2 as their primary test case. Maybe the display drive circuitry was designed for computer use and was repurposed for a video display, or one of the engineers at Sony just moved over from the computer display division and assumed that video displays use 2.2 gamma just like computer displays. Or it might mean you need to spend more time with the multipoint gamma adjustments to get a cleaner 2.4 gamma. I really don't know.

If what you're asking is whether it's better to have a consistent 2.2 gamma all the way across the range or a lumpy 2.4 gamma, I couldn't say. I'd personally go for the lumpy 2.4, especially given that the curve you're showing isn't really that lumpy, but it's going to be a judgment call. The preferred gamma curve would be a clean consistent 2.4, but I assume you already know that. smile.gif
dmunsil is offline  
post #457 of 671 Old 12-17-2013, 10:09 AM
Advanced Member
 
ElectronicTonic's Avatar
 
Join Date: Feb 2008
Location: Ventura, CA
Posts: 940
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 3 Post(s)
Liked: 58
Quote:
Originally Posted by AustinJerry View Post

So, here is my question:  is the 2.2 gamma actually a better result for my display, or do the test results shown in the first screenshot indicate that I have done something wrong? 
Did you change your target gamma in CalMAN? I can't tell from the screenshots.
ElectronicTonic is offline  
post #458 of 671 Old 12-17-2013, 10:33 AM
AVS Special Member
 
chunon's Avatar
 
Join Date: Dec 2007
Location: Wilmington Ohio
Posts: 4,923
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 76 Post(s)
Liked: 457
Quote:
Originally Posted by dmunsil View Post

I'm not understanding your question. The gamma of a BVM is approximately 2.4, which is why the video display gamma standard is 2.4. That's the standard. It remains the standard for watching video. You can certainly watch video on a display calibrated to 2.2 and it'll look fine. It's just brighter in the midtones, which wasn't what the filmmakers intended, but no small children will die. smile.gif

If you want to match a BVM, you want to get your gamma to as close to 2.4 as is feasible. If your display is able to do better grayscale tracking at 2.2, that tells you (perhaps) that the manufacturers of the display designed it around 2.2 as their primary test case. Maybe the display drive circuitry was designed for computer use and was repurposed for a video display, or one of the engineers at Sony just moved over from the computer display division and assumed that video displays use 2.2 gamma just like computer displays. Or it might mean you need to spend more time with the multipoint gamma adjustments to get a cleaner 2.4 gamma. I really don't know.

If what you're asking is whether it's better to have a consistent 2.2 gamma all the way across the range or a lumpy 2.4 gamma, I couldn't say. I'd personally go for the lumpy 2.4, especially given that the curve you're showing isn't really that lumpy, but it's going to be a judgment call. The preferred gamma curve would be a clean consistent 2.4, but I assume you already know that. smile.gif


Very interesting, how is it the 2.2 "standard" has gotten so much momementum from calibrators then ? It seems almost everyone calibrates to in the 2.2 to 2.25 range

65VT60
55ST60
Sony BDV-F7 3dbluray/soundbar
Calman Enthusiast
Dish Network with Hopper/Super Joey
chunon is offline  
post #459 of 671 Old 12-17-2013, 10:44 AM
AVS Special Member
 
AustinJerry's Avatar
 
Join Date: Jan 2003
Location: Austin, Texas
Posts: 6,811
Mentioned: 3 Post(s)
Tagged: 0 Thread(s)
Quoted: 224 Post(s)
Liked: 646
Quote:
Originally Posted by ElectronicTonic View Post

Did you change your target gamma in CalMAN? I can't tell from the screenshots.

Sorry, I don't see in my version of CalMAN where a target gamma can be set, can you clarify? Or perhaps I am not understanding the question properly.
AustinJerry is offline  
post #460 of 671 Old 12-17-2013, 10:56 AM
AVS Special Member
 
dmunsil's Avatar
 
Join Date: Mar 2000
Location: Kirkland, WA, USA
Posts: 1,078
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 24
Quote:
Originally Posted by chunon View Post

Very interesting, how is it the 2.2 "standard" has gotten so much momementum from calibrators then ? It seems almost everyone calibrates to in the 2.2 to 2.25 range

I know several of the top calibrators in the industry, who handle lots of professional studios and personal systems of studio bosses, directors, etc. They all calibrate to 2.4, except when the customer insists on doing it another way for some reason, which rarely happens. If people are calibrating to 2.2, then they're not matching a BVM, so I really have no idea what they're doing. I have to admit that I tend to keep my head down and work on the patterns and software, so if there's some kind of trend of people using 2.2, the only think I can think of is that they're following the computer gamma curve. I know there are some people who feel that 2.4 is too high because it makes the picture darker than they'd like. My reaction is that if you want to get the picture you like, go ahead. But don't call it "calibrating." smile.gif

The actual practice in the industry is 2.4, and has been for as long as television has existed, because while CRTs have a theoretical "true" gamma of 2.5, because of glare, dispersion, and a host of other issues, the actual measured gamma from a sample of properly calibrated BVMs turns out to be pretty close to 2.4. This has been well understood for a long time, but had not been standardized until BT.1886, though Charles Poynton, perhaps the most influential single engineer in the industry, has been beating the gamma drum for a lot of years.

In the end, there continues to be convergence between computers and video, and the computer industry is much bigger than the video industry and there are probably more computer monitors than video displays (and a ton of mixed-use displays). It would be very nice for the computer industry if video would just adopt computer practice, and so far they mostly don't. But that's not to say that the computer industry won't keep pushing the issue.

All we're trying to do is keep people informed. The monitors that Deluxe, Technicolor, etc. use to master movies are calibrated to 2.4. That is a hard, cold fact. Those big mastering houses do all of the first-tier movies in the industry, which is the content that most of us care about getting right. If you want things to look at home like it looks at the mastering house, you should calibrate to 2.4. If it's true that a lot of enthusiasts are using other gammas, then that is their right, and I am not going to tell them to stop. Freedom means you are allowed to watch movies on a slightly brighter-looking display without having goons crash through the windows and arrest you. smile.gif
dmunsil is offline  
post #461 of 671 Old 12-17-2013, 11:02 AM
Senior Member
 
CalWldLif's Avatar
 
Join Date: Feb 2011
Posts: 389
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 9 Post(s)
Liked: 23
Being a serious network programming watcher and being a late adopter
of HD tv all I know about what gamma the networks use that it appears to be a higher
number than what SD was. I noticed around 2007 that shows I regularly watched were
becoming darker over all. It coincided with when they were filming in HD and I had the black bars at top and bottom of my
SD 3x4 screen. What a poster said about editors/masters using the PC to encode the final show makes sense.
I am still not sure what the networks are using or if they even use the same standard.

I do know when I have 2.4+ a low percentage of shows are borderline too dark.

And with 2.2 a low percentage seem blown out.

Anywho, glad that there is info and people working on the issue.
CalWldLif is offline  
post #462 of 671 Old 12-17-2013, 11:29 AM
AVS Special Member
 
chunon's Avatar
 
Join Date: Dec 2007
Location: Wilmington Ohio
Posts: 4,923
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 76 Post(s)
Liked: 457
Quote:
Originally Posted by dmunsil View Post

I know several of the top calibrators in the industry, who handle lots of professional studios and personal systems of studio bosses, directors, etc. They all calibrate to 2.4, except when the customer insists on doing it another way for some reason, which rarely happens. If people are calibrating to 2.2, then they're not matching a BVM, so I really have no idea what they're doing. I have to admit that I tend to keep my head down and work on the patterns and software, so if there's some kind of trend of people using 2.2, the only think I can think of is that they're following the computer gamma curve. I know there are some people who feel that 2.4 is too high because it makes the picture darker than they'd like. My reaction is that if you want to get the picture you like, go ahead. But don't call it "calibrating." smile.gif

The actual practice in the industry is 2.4, and has been for as long as television has existed, because while CRTs have a theoretical "true" gamma of 2.5, because of glare, dispersion, and a host of other issues, the actual measured gamma from a sample of properly calibrated BVMs turns out to be pretty close to 2.4. This has been well understood for a long time, but had not been standardized until BT.1886, though Charles Poynton, perhaps the most influential single engineer in the industry, has been beating the gamma drum for a lot of years.

In the end, there continues to be convergence between computers and video, and the computer industry is much bigger than the video industry and there are probably more computer monitors than video displays (and a ton of mixed-use displays). It would be very nice for the computer industry if video would just adopt computer practice, and so far they mostly don't. But that's not to say that the computer industry won't keep pushing the issue.

All we're trying to do is keep people informed. The monitors that Deluxe, Technicolor, etc. use to master movies are calibrated to 2.4. That is a hard, cold fact. Those big mastering houses do all of the first-tier movies in the industry, which is the content that most of us care about getting right. If you want things to look at home like it looks at the mastering house, you should calibrate to 2.4. If it's true that a lot of enthusiasts are using other gammas, then that is their right, and I am not going to tell them to stop. Freedom means you are allowed to watch movies on a slightly brighter-looking display without having goons crash through the windows and arrest you. smile.gif

Don thanks for your thoughts and the info smile.gif My calibrator has always done 2.2 to 2.25 for my plasmas which I have never questioned, good to get some other viewpoints tho. Might try it out for myself just to see how it looks but I am probably one of the folks that prefers the more video like gamma of 2.2

65VT60
55ST60
Sony BDV-F7 3dbluray/soundbar
Calman Enthusiast
Dish Network with Hopper/Super Joey
chunon is offline  
post #463 of 671 Old 12-17-2013, 11:56 AM
AVS Special Member
 
dmunsil's Avatar
 
Join Date: Mar 2000
Location: Kirkland, WA, USA
Posts: 1,078
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 24
Quote:
Originally Posted by chunon View Post

Don thanks for your thoughts and the info smile.gif My calibrator has always done 2.2 to 2.25 for my plasmas which I have never questioned, good to get some other viewpoints tho. Might try it out for myself just to see how it looks but I am probably one of the folks that prefers the more video like gamma of 2.2

Video gamma is 2.4 - it's not just film mastering that's done at 2.4, it's all the broadcast networks. ABC, CBS, NBC, HBO, etc. 2.2 is only a standard for computer monitors.
dmunsil is offline  
post #464 of 671 Old 12-17-2013, 12:00 PM
AVS Special Member
 
chunon's Avatar
 
Join Date: Dec 2007
Location: Wilmington Ohio
Posts: 4,923
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 76 Post(s)
Liked: 457
I understand that Don wasn't implying otherwise, perhaps a poor choice of words. So I guess I should revise to say I prefer the 2.2 computer monitor gamma smile.gif

65VT60
55ST60
Sony BDV-F7 3dbluray/soundbar
Calman Enthusiast
Dish Network with Hopper/Super Joey
chunon is offline  
post #465 of 671 Old 12-17-2013, 12:00 PM
Advanced Member
 
ElectronicTonic's Avatar
 
Join Date: Feb 2008
Location: Ventura, CA
Posts: 940
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 3 Post(s)
Liked: 58
Quote:
Originally Posted by AustinJerry View Post

Sorry, I don't see in my version of CalMAN where a target gamma can be set, can you clarify? Or perhaps I am not understanding the question properly.
Select settings, on the upper right, next to the "?":

> Workflow Basic Options > Gamma Formula: Power > Target Exponent: You can input your target here.
ElectronicTonic is offline  
post #466 of 671 Old 12-17-2013, 12:01 PM
Advanced Member
 
cadett's Avatar
 
Join Date: Aug 2009
Location: Houma, La.
Posts: 809
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 1 Post(s)
Liked: 83
Can we just agree on 2.3 & call it a day smile.gif
cadett is offline  
post #467 of 671 Old 12-17-2013, 12:03 PM
AVS Special Member
 
chunon's Avatar
 
Join Date: Dec 2007
Location: Wilmington Ohio
Posts: 4,923
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 76 Post(s)
Liked: 457
^^^Touche

65VT60
55ST60
Sony BDV-F7 3dbluray/soundbar
Calman Enthusiast
Dish Network with Hopper/Super Joey
chunon is offline  
post #468 of 671 Old 12-17-2013, 12:14 PM
AVS Special Member
 
sotti's Avatar
 
Join Date: Aug 2004
Location: Seattle, WA
Posts: 6,585
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 12 Post(s)
Liked: 164
Quote:
Originally Posted by dmunsil View Post

I would call it a matter of degree. Yes, there are "mastering houses" setting things up poorly. However, the biggest houses like Deluxe and Technicolor and Warner have always used BVMs, and are now moving to flat panels calibrated to look like BVMs. I know this for sure. I know some of the people who do those calibrations. They have always mastered to 2.4.r thing.

Anyway, I return to my first assertion, which has not changed: big-studio movies are now and have always been mastered to 2.4. Television production at the major networks is mastered to 2.4. There are miscellaneous small producers who do all kinds of wacky things, but you can't adapt your system to deal with their foolishness, so you shouldn't even try.

Don,

I hate to disagree with you, but I was just talking to one professional calibrator (PM me if you want specifics) who was dealing with a big client who was screening the blu-ray of an academy award winning film. Played back on the Christie projector, it looked too dark. The Calibrator eventually found out that the projector was correctly calibrated at 2.4, but the film itself was mastered at 2.2. So I know of at least one big budget, award winning film that was done at 2.2. Also I went in to demo at a major post production/color grading facility and their principal color scientist confided in me that while he thought everything was button down, about half their screens were 2.2 and half were 2.4. I'm sure that's been rectified at this point.

But I do agree prior to flat panels, everything was BVMs. BVMs didn't have gamma controls. That's why BT.1886 specifically calls out trying to emulate BVMs. Which is why we recommend using BT.1886 for any flat panel calibration.

Once again sorry to disagree, but those are my experiences. I wish that it was more cut and dry myself.

Thanks,
Joel

Joel Barsotti
SpectraCal
CalMAN Lead Developer
sotti is offline  
post #469 of 671 Old 12-17-2013, 12:35 PM
Advanced Member
 
cadett's Avatar
 
Join Date: Aug 2009
Location: Houma, La.
Posts: 809
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 1 Post(s)
Liked: 83
Isn't BT.1886 close to 2.4 for a typical calibration
cadett is offline  
post #470 of 671 Old 12-17-2013, 01:07 PM
AVS Special Member
 
dmunsil's Avatar
 
Join Date: Mar 2000
Location: Kirkland, WA, USA
Posts: 1,078
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 24
Quote:
Originally Posted by sotti View Post

Don,

I hate to disagree with you, but I was just talking to one professional calibrator (PM me if you want specifics) who was dealing with a big client who was screening the blu-ray of an academy award winning film. Played back on the Christie projector, it looked too dark. The Calibrator eventually found out that the projector was correctly calibrated at 2.4, but the film itself was mastered at 2.2. So I know of at least one big budget, award winning film that was done at 2.2. Also I went in to demo at a major post production/color grading facility and their principal color scientist confided in me that while he thought everything was button down, about half their screens were 2.2 and half were 2.4. I'm sure that's been rectified at this point.

But I do agree prior to flat panels, everything was BVMs. BVMs didn't have gamma controls. That's why BT.1886 specifically calls out trying to emulate BVMs. Which is why we recommend using BT.1886 for any flat panel calibration.

Once again sorry to disagree, but those are my experiences. I wish that it was more cut and dry myself.

Thanks,
Joel

Well, that's just depressing. I still can't understand why there has been such a profusion of gamma standards over the years. It's been reasonably well understood what the natural gamma of CRTs is, so why did sRGB set the gamma curve at 2.2? It required a LUT to get that gamma on a CRT computer monitor. You wouldn't think the industry would go in for a needless expense unless they had some specific reason, but I'm at a loss as to what that reason is. Perhaps it's slightly closer to the perceptual just-noticeable-difference curve for normal lighting conditions. I wouldn't think that would be worth adding a LUT to millions and millions of monitors for, but what do I know?
dmunsil is offline  
post #471 of 671 Old 12-17-2013, 01:10 PM
AVS Special Member
 
sotti's Avatar
 
Join Date: Aug 2004
Location: Seattle, WA
Posts: 6,585
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 12 Post(s)
Liked: 164
Quote:
Originally Posted by dmunsil View Post

Well, that's just depressing. I still can't understand why there has been such a profusion of gamma standards over the years. It's been reasonably well understood what the natural gamma of CRTs is, so why did sRGB set the gamma curve at 2.2?

I think that is another thing that is quite misunderstood.

The sRGB gamma curve is a linear tail with an exponent of 2.4. It's much more similar to BT.1886 than it is to a normal gamma 2.2.
Quote:
Originally Posted by dmunsil View Post

It required a LUT to get that gamma on a CRT computer monitor. ]You wouldn't think the industry would go in for a needless expense unless they had some specific reason, but I'm at a loss as to what that reason is.

I don't have any CRTs left, but when you look at the original spec from the w3.org http://www.w3.org/Graphics/Color/sRGB.html, while they do describe it as a 2.2 gamma on average, I think that is because sRGB is lower than 2.2 for a bit, and then more around 2.3 so it averages around 2.2. But then goes on to specify the actual formula. They also talk extensively how that gamma was chosen because it was the natural response of a CRT.

Also you can look at the formula at the bottom of this sRGB spec at color.org( the curators of ICC) http://www.color.org/srgb.pdf, and one again the describe the gamma as 2.2, but then go on to specify a formula that is a linear tail with an exponent of 2.4.

Also every graphics card does have a LUT, so there is that.
Quote:
Originally Posted by dmunsil View Post

Perhaps it's slightly closer to the perceptual just-noticeable-difference curve for normal lighting conditions. I wouldn't think that would be worth adding a LUT to millions and millions of monitors for, but what do I know?

I do know that the sRGB spec does come with a specified viewing environment of 200lux, so it is target at brighter environments.

Joel Barsotti
SpectraCal
CalMAN Lead Developer
sotti is offline  
post #472 of 671 Old 12-17-2013, 01:12 PM
AVS Special Member
 
dmunsil's Avatar
 
Join Date: Mar 2000
Location: Kirkland, WA, USA
Posts: 1,078
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 24
Quote:
Originally Posted by cadett View Post

Isn't BT.1886 close to 2.4 for a typical calibration

BT.1886 specifies a gamma of exactly 2.4, and it does not have compensation for varying baseline "natural" black levels. That was on the table for a while, but it was generally considered better to just codify existing behavior of CRTs.
dmunsil is offline  
post #473 of 671 Old 12-17-2013, 01:16 PM
AVS Special Member
 
dmunsil's Avatar
 
Join Date: Mar 2000
Location: Kirkland, WA, USA
Posts: 1,078
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 24
Quote:
Originally Posted by sotti View Post

I think that is another thing that is quite misunderstood.

The sRGB gamma curve is a linear tail with an exponent of 2.4. It's much more similar to BT.1886 than it is to a normal gamma 2.2.

The exponent on the power curve portion is 2.4, but the linear tail and the scale and offset of the power curve make the curve (by design) extremely close to 2.2. You can see this if you plot the two curves.

This is similar to the curve for 709 encoding, which also has a linear tail. It has an exponent of 1/2.2, but the final curve after adjustment is very close to 1/2. Again, this is by design, and not a mistake.

BT.1886 does not have a linear tail or any scaling or shifting, so it's a straight 2.4 curve.

Edit: I was going to plot the curves so you could see, but I see that Wikipedia has a good plot showing how the sRGB curve matches a 2.2 curve very closely:

http://en.wikipedia.org/wiki/File:SRGB_gamma.svg

The red line is the sRGB curve, with the linear tail and the scaled 2.4 power curve, and the dotted black line behind it is a straight 2.2 curve. You can see they're almost identical.
dmunsil is offline  
post #474 of 671 Old 12-17-2013, 01:37 PM
AVS Special Member
 
AustinJerry's Avatar
 
Join Date: Jan 2003
Location: Austin, Texas
Posts: 6,811
Mentioned: 3 Post(s)
Tagged: 0 Thread(s)
Quoted: 224 Post(s)
Liked: 646
Quote:
Originally Posted by ElectronicTonic View Post


Select settings, on the upper right, next to the "?":

> Workflow Basic Options > Gamma Formula: Power > Target Exponent: You can input your target here.

 

As a relatively new CalMAN user, I always welcome the opportunity of learning something new, and I thank you for pointing out this setting.  I have made a renewed commitment to read the entire CalMAN help file, and then go back and see if I can get a slightly less "lumpy" gamma curve with a 2.4 setting.

 

Having said that, and don't take offense at this, Don, but it is difficult for my eyes to see the difference between 2.2 and 2.4, especially since adjusting the Backlight setting can compensate for a darker screen.  But, I respect that 2.4 is the recommended target, and will attempt to get there once again...

ElectronicTonic likes this.
AustinJerry is offline  
post #475 of 671 Old 12-17-2013, 01:45 PM
AVS Special Member
 
sotti's Avatar
 
Join Date: Aug 2004
Location: Seattle, WA
Posts: 6,585
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 12 Post(s)
Liked: 164
Quote:
Originally Posted by dmunsil View Post

BT.1886 does not have a linear tail or any scaling or shifting, so it's a straight 2.4 curve.

From everything I've read the recommendation for BT.1886 is the formula in Annex1, which does absolutely have an offset based on relative difference of the black and white levels.

Quote:
Originally Posted by dmunsil View Post

Edit: I was going to plot the curves so you could see, but I see that Wikipedia has a good plot showing how the sRGB curve matches a 2.2 curve very closely:

http://en.wikipedia.org/wiki/File:SRGB_gamma.svg

So here is sRGB (grey line) v 2.4 (yellow)


Here is sRGB ( grey line) v BT.1886 (yellow) with a Lb= 0.05cd/m^2, Lw=100cd/m^2.



So both sRGB and BT.1886 purport to emulate CRT and both are extremely similiar on black and white levels common to quality LCD and plasma displays.

Joel Barsotti
SpectraCal
CalMAN Lead Developer
sotti is offline  
post #476 of 671 Old 12-17-2013, 02:03 PM
Advanced Member
 
sawfish's Avatar
 
Join Date: Feb 2013
Posts: 753
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 8 Post(s)
Liked: 102
Quote:
Originally Posted by dmunsil View Post

BT.1886 specifies a gamma of exactly 2.4, and it does not have compensation for varying baseline "natural" black levels. That was on the table for a while, but it was generally considered better to just codify existing behavior of CRTs.

Then I totally don't get what Calman is doing, because it appears to me it does exactly that for BT.1886.
sawfish is online now  
post #477 of 671 Old 12-17-2013, 02:21 PM
AVS Special Member
 
PlasmaPZ80U's Avatar
 
Join Date: Feb 2009
Posts: 7,047
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 8 Post(s)
Liked: 186
Quote:
Originally Posted by chunon View Post

Don thanks for your thoughts and the info smile.gif My calibrator has always done 2.2 to 2.25 for my plasmas which I have never questioned, good to get some other viewpoints tho. Might try it out for myself just to see how it looks but I am probably one of the folks that prefers the more video like gamma of 2.2

with plasmas, it's a little less clear because (for example) on my S60 with the "2.4" setting I measure 2.2-2.25 avg gamma with standard 10-11% windows but if I look at other sizes, smaller and larger (and also fixed APL patterns), I can get some of those patterns to show a 2.4 avg gamma with the same "2.4" setting on the TV
PlasmaPZ80U is offline  
post #478 of 671 Old 12-17-2013, 02:53 PM
Advanced Member
 
spacediver's Avatar
 
Join Date: May 2013
Location: Toronto
Posts: 772
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 39 Post(s)
Liked: 65
Quote:
Originally Posted by PE06MCG View Post

Thanks for your advice.

I notice the same trait with yourself. Certainly if persistence is a fault I plead guilty.

May I suggest you read my post carefully in the same way you expect other people to read yours.

I'm sure the forum police do not need any help from either of us.

I apologize.
spacediver is offline  
post #479 of 671 Old 12-17-2013, 04:34 PM
Advanced Member
 
gwgill's Avatar
 
Join Date: Jan 2013
Location: Melbourne, Australia
Posts: 561
Mentioned: 3 Post(s)
Tagged: 0 Thread(s)
Quoted: 24 Post(s)
Liked: 70
Quote:
Originally Posted by dmunsil View Post


BT.1886 does not have a linear tail or any scaling or shifting, so it's a straight 2.4 curve.
The BT.1886 is a 2.4 power with input shifting to allow for a non zero black.
Please look at the equations in the spec.
gwgill is online now  
post #480 of 671 Old 12-17-2013, 04:42 PM
AVS Special Member
 
dmunsil's Avatar
 
Join Date: Mar 2000
Location: Kirkland, WA, USA
Posts: 1,078
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 24
Quote:
Originally Posted by sotti View Post

From everything I've read the recommendation for BT.1886 is the formula in Annex1, which does absolutely have an offset based on relative difference of the black and white levels.

I was unclear. Let me back up a little. Gamma has generally been expressed as the electrical transfer function over and above the baseline light coming from the display when it is at its blackest. Some display technologies could (and can) get pretty near absolute black, while others are still producing light when they're displaying "black", sometimes quite a bit of light, relatively speaking. Some people wanted to define the transfer function for video displays explicitly taking that into account (or maybe from another way of looking, pretending it's not actually true), so a specific input code value or voltage would correspond to a specific percentage of the maximum light output. This would mean that if you called for 0.0001% of max light, a display capable of going that low would display that absolute amount, while another conforming display with a higher absolute black level would show the minimum it could handle. This would result in some displays being at a big disadvantage, because they'd have to collapse all the levels near black to whatever their minimum is. So that essentially was considered impractical. But it was definitely discussed; I talked with some of the people who discussed it.

The problem with absolute gamma curves is that it implies that on certain displays, there is no visible difference between low code values until the specified output gets above the baseline that the display can produce. That's not optimal for shadow detail, obviously, and makes certain display technologies essentially non-viable. And CRTs don't act that way - they do get brighter as you raise the input voltage. For perfect reproducibility between monitors (especially different display technologies), the "absolute" gamma is desirable. For emulation of CRTs and practical expediency, the "relative" gamma is preferable.

(snipped graphs so as not to post them twice)

It seems to me that you're saying sRGB needs to be treated as an absolute gamma curve, which it has not been in the past, because it would have resulted in astonishingly large shadow clipping on lots of real-world monitors. Actual displays have always treated the sRGB curve as being relative to an implied black offset, so the L output from the sRGB function was (in effect) added to Lb, the minimum the display could produce. So you're comparing BT.1886, with its explicit display-dependent offset, to sRGB without any display-dependent offset. I'm not sure that's kosher, but I admit that I had never considered that interpretation of sRGB. If you look at the original sRGB white paper, they kind of handwave the black offset; they acknowledge that it exists and varies from monitor to monitor, but essentially then stop talking about it. I can ask some of the people who wrote that paper if their intent was for sRGB to be considered absolute. It's not completely impossible. Heaven knows (and this thread proves) that gamma is one of the least well understood topics in graphics. smile.gif

If in fact the theoretical model for sRGB is to see the 2.2 curve as the absolute curve for the monitor, rather than a relative curve ignoring the offset, then I agree - that explains the mismatch. In practice when you set a "2.2" or "2.4" gamma on a monitor, it's treated more like the BT.1886 calculation. There's a display-dependent offset, and then the gamma curve is calculated on the output values between the min and max values.
dmunsil is offline  
Reply Display Calibration

Tags
Spears Munsil , Spears Munsil High Definition Benchmark Blu Ray Disc Edition Blu Ray , Spears Munsil Hd Benchmark And Calibration Disc 2nd Edition
Gear in this thread

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