or Connect
AVS › AVS Forum › Display Devices › Display Calibration › 2.5 is display gamma, NOT 2.2
New Posts  All Forums:Forum Nav:

2.5 is display gamma, NOT 2.2 - Page 2

post #31 of 161
Quote:
Originally Posted by Drakaal View Post

2.5 would assume that the end user took it out of the box, and never used it or calibrated it. Durring the first 300 hours of its life it is 2.5, but it won't stay that for long.

Properly calibrated it should be 2.2, which also leaves room to compensate for aging, so that the display can be re-calibrated over time.

2.2 is the right number.

I was not aware that the gamma curve will change over time from 2.2 to 2.5 kind of a self calibrating display nice. Overall output yes, maybe a slight color xy shift as the phosphor or bulbs age. Well I guess we need to inform all the industry professionals about this.

BTW, you may want to remove the link from your signature before you get your account suspended for a clear violation of AVS posting rules.
post #32 of 161
Icaillo and Mr.D, thanks for your feedback.

The gamma on my Panasonic plasma display is typically 2.0-2.1, which is too bright for my taste. If I reduce Brightness, gamma goes up, but then I lose black detail. What I've been doing is an undesirable trade off between those 2 scenarios, and that's why I mentioned it would be very nice to tweak gamma (upwards) without having to reduce Brightness (and Contrast). I'll play with the settings this coming weekend and post any relevant result.

Thanks again,

Fernando
post #33 of 161
Thread Starter 
Quote:
Originally Posted by Drakaal View Post

2.5 would assume that the end user took it out of the box, and never used it or calibrated it. Durring the first 300 hours of its life it is 2.5, but it won't stay that for long.

Properly calibrated it should be 2.2, which also leaves room to compensate for aging, so that the display can be re-calibrated over time.

2.2 is the right number.

Are you just making this up? What is your reference for any of these claims?

If you are referring to a CRT, almost every CRT I've ever seen in my life will generally rise in gamma as the phosphor wears, it generally has much greater difficulty coming out of black and rendering shadow detail as it ages rather than the other way around.

As for lack of calibration, Poynton also expressly includes a long description of mistaken gamma assumptions and how it relates to lack of calibration and the lower numbers often 'cited' for many CRT displays.
post #34 of 161
Here's what I measured for a generic Chinese 21" direct view CRT. the ref line is 2.22 Not to say what is should be but rather what I measured one to be. Seems consistent with what Tom has measured the Sonys at.


Dave
post #35 of 161
Quote:
Originally Posted by Drakaal View Post

2.5 would assume that the end user took it out of the box, and never used it or calibrated it. Durring the first 300 hours of its life it is 2.5, but it won't stay that for long.

Properly calibrated it should be 2.2, which also leaves room to compensate for aging, so that the display can be re-calibrated over time.

2.2 is the right number.

In case Derek's sarcasm or Chris' outright challenge is missed, let me add this point lest others get confused:

You say authoritatively that 2.2 is the right number. If you really can be authoritative, then you also know enough that this number changes depending upon the viewing environment. Care to amend your statement a bit?

Bill
post #36 of 161
Thread Starter 
Quote:
Originally Posted by dlarsen View Post

Here's what I measured for a generic Chinese 21" direct view CRT. the ref line is 2.22 Not to say what is should be but rather what I measured one to be. Seems consistent with what Tom has measured the Sonys at.

I'm not sure why this is at all relevant. Mastering studios are not using "generic Chinese" CRT monitors for professional monitoring.
post #37 of 161
Quote:
Originally Posted by FernandoF View Post

Icaillo and Mr.D, thanks for your feedback.

The gamma on my Panasonic plasma display is typically 2.0-2.1, which is too bright for my taste. If I reduce Brightness, gamma goes up, but then I lose black detail. What I've been doing is an undesirable trade off between those 2 scenarios, and that's why I mentioned it would be very nice to tweak gamma (upwards) without having to reduce Brightness (and Contrast). I'll play with the settings this coming weekend and post any relevant result.

Thanks again,

Fernando

I calibrated mine again last night: getting the brightness and contrast correct on a panny plasma is very important before you calibrate grayscale and gamma. Too high a contrast gives you quite a nasty bulge from around 40IRE that you won't be able to correct for with the cuts and drives.

I find Calman quite useful for suggesting how best to adjust Y for gamma .The normalised Y reading after an initial metering run is useful as is the actual plot of the gamma.

I had to drop my Y around the 80IRE mark to pull the upper part of the curve down to an idealised 2.5. It was pretty much bang on at the lower ranges though.

This fixed the washed out picture I was getting courtesy of my new HTPC...now if I can just fix the color primaries...which are frankly nuts.
post #38 of 161
Quote:
Originally Posted by ChrisWiggles View Post

I'm not sure why this is at all relevant. Mastering studios are not using "generic Chinese" CRT monitors for professional monitoring.

Is the following relevant enough for your professional standards

http://www.extremetech.com/article2/...1747580,00.asp
Dr. Raymond Soneira
“The key to using all of this material was the simultaneous viewing and comparison of all the displays against one another and against a reference standard, which was a direct-view professional High Definition CRT studio monitor”

Our reference CRT was a Sony PVM-20L5:
http://www.cev.ca/production/Sony_PV...re-english.pdf
which is a direct-view professional High Definition studio monitor and was chosen as the reference standard for image and picture quality because it delivers virtually perfect performance.

Gamma 2.20 (perfect, the standard)
Black Level .01 cd/m^2
Peak Brightness 176 cd/m^2
Dynamic Range 17,600:1
4x4 checkerboard contrast 219:1

CRT Strongest Points:
• Best black-level and Dynamic Range of all the display technologies
• Highest color and gray-scale accuracy
Most accurate Gamma
• Perfectly smooth gray-scale with no false contouring
• Excellent accuracy and low noise at the dark-end of the gray-scale
• Supports a wide range of resolutions
• Image rescaling not necessary
• No motion artifacts
• Widest viewing angles
• Least artifacts of all the display technologies
• Gaussian beam profile produces a very smooth image.
post #39 of 161
http://www.extremetech.com/article2/...1736943,00.asp
"Gamma is the numerical value of the slope (steepness) of the gray scale when plotted on a logarithmic "log-log" graph. While there are reasons a gamma of 3.0 might be considered optimum (based on the eye's specific power-law response), it is more important to have a standard gamma value defined. Television, DVD, Web, and computer-based photographic content are generally color balanced on professional CRT studio monitors adjusted to a standard gamma of 2.20, so you'll get the most accurate images with this value"
post #40 of 161
Thread Starter 
I've asked you not to continue posting in my threads thomas. Please leave me alone.
post #41 of 161
Thread Starter 
I also should link this old thread by Guy Kuo while I'm at it, it also contains some more gathered information on this topic:

http://archive.avsforum.com/avs-vb/s...hlight=scaling

His first post in its entirety:

Quote:


Display Gamma should be Higher than 2.2
I'm cross posting this because it pertains even more to digital displays. CRT's when correctly set up inherently have a gamma response which is about right. Digitals need to emulate that behavior to make their images correct, but we keep seeing all sorts of problems. The most glaring I think is the recurrent idea that 2.2 is the ideal display gamma. It actually should be higher to properly emulate what is seen on the CRT monitor of the telecine operator.



quote:
--------------------------------------------------------------------------------
Originally posted by sspears
A Sony BVM is usually the display of choice during color correction with the film to video transfer. The Sony BVM24 HD monitor actually has a gamma of ~2.7. It also has SMPTE C phospor.
--------------------------------------------------------------------------------



The roughly gamma 2.5 ideal CRT response is such that it slighly over degammas the 0.45 gamma of the source signal. The overall system gamma response is supposed to be about 1.1 to 1.2 to allow dark surround viewing. If you make the display response too linear (lower gamma) the scenes will have their midrange values too bright and the image will appear too "flat."

----------------------------------------------------

I quote some useful material from Poynton http://www.poynton.com/notes/colour_and_gamma/

The actual value of gamma for a particular CRT may range from about 2.3 to 2.6. Practitioners of computer graphics often claim numerical values of gamma quite different from 2.5. But the largest source of variation in the nonlinearity of a monitor is caused by careless setting of the Black Level (or Brightness ) control of your monitor. Make sure that this control is adjusted so that black elements in the picture are reproduced correctly before you devote any effort to determining or setting gamma . display transfer characteristic (gamma, ?),” EBU Technical Review 257 (Autumn 1993), 32–40..... "

"......Now, here’s a surprise. If a film system is designed and processed to produce exactly linear reproduction of intensity, reflection prints look fine. But projected transparencies – slides and movies – look flat, apparently lacking in contrast! The reason for this involves another aspect of human visual perception: the surround effect.

As explained in Adaptation, on page 85, human vision adapts to an extremely wide range of viewing conditions. One of the mechanisms involved in adaptation increases our sensitivity to small brightness variations when the area of interest is surrounded by bright elements. Intuitively, light from a bright surround can be thought of as spilling or scattering into all areas of our vision, including the area of interest, reducing its apparent contrast. Loosely speaking, the vision system compensates for this effect by “stretching” its contrast range to increase the visibility of dark elements in the presence of a bright surround. Conversely, when the region of interest is surrounded by relative darkness, the contrast range of the vision system decreases: Our ability to discern dark elements in the scene decreases. The effect is demonstrated in Figure 6.4 above, from DeMarsh and Giorgianni.

The surround effect has implications for the display of images in dark areas, such as projection of movies in a cinema, projection of 35 mm slides, or viewing of television in your living room. If an image is viewed in a dark or dim surround, and the intensity of the scene is reproduced with correct physical intensity, the image will appear lacking in contrast.LeRoy E. DeMarsh and Edward J. Giorgianni, “Color Science for Imaging Systems,” in Physics Today, September 1989, 44–52. "

"As explained in Gamma in film, on page 96, it is important for perceptual reasons to “stretch” the contrast ratio of a reproduced image when viewed in a dim surround. The dim surround condition is characteristic

of television viewing. In video, the “stretching” is accomplished at the camera by slightly undercompensating the actual power function of the CRT to obtain an end-to-end power function with an exponent of 1.1 or 1.2. This achieves pictures that are more subjectively pleasing than would be produced by a mathematically correct linear system.

Rec. 709 specifies a power function exponent of 0.45. The product of the 0.45 exponent at the camera and the 2.5 exponent at the display produces the desired end-to-end exponent of about 1.13. An exponent of 0.45 is a good match for both CRTs and for perception."

Figure 6.6 above illustrates the transfer function defined by the international Rec. 709 standard for high-definition television (HDTV). It is basically a power function with an exponent of 0.45. Theoretically a pure power function suffices for gamma correction; however, the slope of a pure power function is infinite at zero. In a practical system such as a television camera, in order to minimize noise in the dark regions of the picture it is necessary to limit the slope (gain) of the function near black. Rec. 709 specifies a slope of 4.5 below a tristimulus value of +0.018, and stretches the remainder of the curve to maintain function and tangent continuity at the breakpoint. In this equation the red tristimulus (linear light) component is denoted R, and the resulting gamma-corrected video signal is denoted with a prime symbol, R’709. The computation is identical for the other two components"

The difference between the SMPTE 240M and Rec. 709 transfer functions is negligible for real images. It is a shame that international agreement could not have been reached on the SMPTE 240M parameters that were widely implemented at the time the CCIR (now ITU-R) discussions were taking place. The Rec. 709 values are closely representative of current studio practice, and should be used for all but very unusual conditions.

An idealized monitor inverts the transform: Real monitors are not as exact as this equation suggests, and have no linear segment, but the precise definition is necessary for accurate intermediate processing in the linear-light domain. In a color system, an identical transfer function is applied to each of the three tristimulus (linearlight) RGB components. See Frequently Asked Questions about Color . Incidentally, the nonlinearity of a CRT is a function of the electrostatics of the cathode and the grid of an electron gun; it has nothing to do with the phosphor. Also, the nonlinearity is a power function [which has the form f( x ) = x a ], not an exponential function [which has the form f( x ) = a x ]. For more detail, read the Gamma chapter in Poynton’s book [3].

Does NTSC use a gamma of 2.2? Television is usually viewed in a dim environment. If an images’s correct physical intensity 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.

--------- end quotes

In other words, an ideal display gamma is actually higher than 2.2 but closer to 2.5 so the overall system has a positive gamma. This keeps getting messed up in digital implementations because designers see the 2.2 in the spec and forget that 2.2 isn't the actual display gamma but the inverse which if used makes the system (incorrectly) linear. A real CRT display's transfer characteristics should be emulated, not a 2.2 power function. It's close to a 2.5 power function, but most accurately given by the inverse of the rec 709 as below.




(Now let's watch this drop like a rock out of the bottom of the forum. This is the sort of stuff that Stacey and I mean when we say that some our best postings go ignored on the boards.)

Attachment: rec 709 for monitor.jpg
This has been downloaded 231 time(s).


__________________
Guy Kuo
Director - Imaging Science Foundation Research Lab
Video Test Design - Ovation Multimedia / Home of OpticONE Colorimeter, AVIA and Avia PRO
post #42 of 161
Rather than arguing over which is correct, here's a more practical question for you: without resorting to lowering the brightness and clipping shadow details, what flat panels out there are actually capable of a flat 2.5 gamma (rather than one that averages at 2.5) without an external processor?

Based on the discussion in this topic, I calibrated my computer screen to 2.5 last week and I must say that the difference was bigger than I expected. Despite it being a laptop LCD with an 800:1 (calibrated) contrast ratio calibrating to 2.5 made a massive improvement to how things look.

No longer are images/videos "washed out" and skintones look far more natural than before, especially in low-light scenes. (I transferred a few recordings off my PVR for testing) It's a lot closer to the "CRT-look" that I prefer.

Unfortunately, my CRTs are on their way out and I'll need to replace them very soon, and flat-panels are the only option available here these days. I calibrated my laptop screen to a 2.0 gamma tonight, as I believe that's roughly what all the latest BRAVIAs measure, and it's unwatchable in my opinion - completely flat/washed-out. Now, I know that the Kuros measure around 2.2 (I believe they can go slightly higher, I can't remember now) but most HDTVs I've encountered seem to be around 2.0 or lower.

Now, I know that around these parts, Plasmas are considered to be the far superior displays, and I agree that they are in many respects, but I plan on doing a lot of time gaming with the display, and I am sensitive to seeing the phosphor lag which I find very distracting so I would prefer an LCD, despite the issues they have.

These days it seems like you're lucky to find something that gets as high as 2.2 let alone 2.5. Is there actually anything on sale that can do it without losing detail?
post #43 of 161
Very interesting post, Andrew.

I was under the impression that manufacturers/engineers had more control over gamma, but it looks like it is rather limited by the adopted display technology. I look forward to seeing more comments.
post #44 of 161
They do have a great deal of control over gamma if they choose to exercise it. There are lots of ways to argue the matter, but in practice, I have used the Accucal software with a target gamma of 2.22 for some time. I try to get the best match and find that somewhere slightly higher than that is better in most cases. I typically err on the higher side of 2.22, with most ending up in the 2.3-2.4 range and getting excellent results. Far more important than the actual number is to get the smoothest tracking over the range of the image, without clipping white nor crushing blacks, and getting the black level just right and preserving dark detail. Generally, in the field, you have somewhat limited control and work with what the set gives you. Regardless, a single number is much less meaningful than looking at the response curve.
post #45 of 161
I don't see how a display with 800:1 on/off CR could be capable of a 2.5 gamma without losing shadow detail.

For example, assume peak output of, say, 75 fL. That means at 800:1 the display's black level would be 0.094 fL. However, with a linear 2.5 gamma the output at various levels of stimulus would be the following:

10% - 0.237
9% - 0.182
8% - 0.136
7% - 0.097
6% - 0.066
5% - 0.042
4% - 0.024
3% - 0.012
2% - 0.004

which means that all blacks would be crushed below 7% stimulus. What this suggests is that a display gamma of 2.5 is fine if the display is a CRT with very high on/off contrast (20,000:1 or better), but displays with lower CR simply cannot take advantage of a gamma that high without losing shadow detail.


Quote:
Originally Posted by andrewfee View Post

Rather than arguing over which is correct, here's a more practical question for you: without resorting to lowering the brightness and clipping shadow details, what flat panels out there are actually capable of a flat 2.5 gamma (rather than one that averages at 2.5) without an external processor?

Based on the discussion in this topic, I calibrated my computer screen to 2.5 last week and I must say that the difference was bigger than I expected. Despite it being a laptop LCD with an 800:1 (calibrated) contrast ratio calibrating to 2.5 made a massive improvement to how things look.

No longer are images/videos "washed out" and skintones look far more natural than before, especially in low-light scenes. (I transferred a few recordings off my PVR for testing) It's a lot closer to the "CRT-look" that I prefer.

Unfortunately, my CRTs are on their way out and I'll need to replace them very soon, and flat-panels are the only option available here these days. I calibrated my laptop screen to a 2.0 gamma tonight, as I believe that's roughly what all the latest BRAVIAs measure, and it's unwatchable in my opinion - completely flat/washed-out. Now, I know that the Kuros measure around 2.2 (I believe they can go slightly higher, I can't remember now) but most HDTVs I've encountered seem to be around 2.0 or lower.

Now, I know that around these parts, Plasmas are considered to be the far superior displays, and I agree that they are in many respects, but I plan on doing a lot of time gaming with the display, and I am sensitive to seeing the phosphor lag which I find very distracting so I would prefer an LCD, despite the issues they have.

These days it seems like you're lucky to find something that gets as high as 2.2 let alone 2.5. Is there actually anything on sale that can do it without losing detail?
post #46 of 161
My question.. is there any reason to doubt the Director of Engineering from Ikegami?

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


The Gamma for the Sony CRT agrees perfectly with the 2.20 standard value. CRT monitors from Ikegami, another major brand of professional studio monitors, also have a Gamma of 2.20 according to their Director of Engineering.
post #47 of 161
Quote:
Originally Posted by FernandoF View Post

Very interesting post, Andrew.

I was under the impression that manufacturers/engineers had more control over gamma, but it looks like it is rather limited by the adopted display technology. I look forward to seeing more comments.

Perhaps I worded that badly. When I say "capable of a flat 2.5 gamma" I really mean one which has the controls to allow you to get anywhere near a 2.5 gamma without crushing all shadow detail.

Quote:
Originally Posted by TomHuffman View Post

I don't see how a display with 800:1 on/off CR could be capable of a 2.5 gamma without losing shadow detail.

...

Well, I can see everything down to 2% grey (5,5,5) but anything below that does clip to black in unmanaged applications. In colour-managed programs, I can see all the way down to 1,1,1.

The profile was created in ColorEyes Display Pro and the target was 2.5 Gamma, D65 temperature, maximum brightness (for that backlight setting) and relative black point rather than absolute. (I think this is important) In the end, it was at 87cd/m2 for white, and 0.110 cd/m2 for black which works out at ~790:1.

I should point out that I fully expect this to be adding posterisation, but as it is a laptop display which suffers from it anyway, the effect really isn't that noticeable. On the other-hand, find me any non-CRT display that doesn't posterise the image to some degree. (though this is obviously worse than the best performers)

Perhaps it isn't a completely flat 2.5 gamma, but the fact that it is at least close, looks far better than anything else I've calibrated the screen to before. At anything else, images look flat/washed out, but with this 2.5 profile things are significantly improved.



Now, I realise that this is not what you would do in a HT environment, but as this is a laptop display, I would not have expected great results anyway. However, my question wasn't really about whether or not this display is showing a completely flat 2.5 gamma, but rather, whether there are actually any flat-panel HDTVs on the market right now that can get anywhere close to it. Most that I have seen either have S-shaped gammas, or are well below 2.2, let alone 2.5.
post #48 of 161
This is an excerpt from a short email that I recieved from Charles Poynton in response to a question. I'll probably post the entire email later. I'm waiting to see if he answers another question I asked of him. Extremely nice guy. I thought this would give another point of view in this thread, although my question to him is in relation to another thread.

Quote:


Ideally, the display transfer function should follow its inverse gamma curve from reference white (interface 235) up to peak white (interface 254). Inverse gamma should be a 2.4-power for studio viewing conditions (100 nit display, 64 lux ambient). A substantially brighter display of a substantially brighter ambient calls for a lower value of gamma; for example, for typical desktop computer environments the 2.2 value of sRGB is appropriate.
post #49 of 161
We are really debating a very fine point. This sums it up. Get your gamma somewhere between 2.2 and 2.4. However, as I wrote before, the higher gamma figures really require a very high on/off contrast ratio to avoid loss of shadow detail.

Quote:
Originally Posted by hwjohn View Post

This is an excerpt from a short email that I recieved from Charles Poynton in response to a question. I'll probably post the entire email later. I'm waiting to see if he answers another question I asked of him. Extremely nice guy. I thought this would give another point of view in this thread, although my question to him is in relation to another thread.
post #50 of 161
Quote:
Originally Posted by TomHuffman View Post

I don't see how a display with 800:1 on/off CR could be capable of a 2.5 gamma without losing shadow detail.

For example, assume peak output of, say, 75 fL. That means at 800:1 the display's black level would be 0.094 fL. However, with a linear 2.5 gamma the output at various levels of stimulus would be the following:

10% - 0.237
9% - 0.182
8% - 0.136
7% - 0.097
6% - 0.066
5% - 0.042
4% - 0.024
3% - 0.012
2% - 0.004

which means that all blacks would be crushed below 7% stimulus. What this suggests is that a display gamma of 2.5 is fine if the display is a CRT with very high on/off contrast (20,000:1 or better), but displays with lower CR simply cannot take advantage of a gamma that high without losing shadow detail.

Actually, your analysis only captures black crush when measuring light. From a visual standpoint, you crush through the teens and into the twenties -- 22.2% would be 4 dE (really dL*) from the black point.

Bill
post #51 of 161
Quote:
Originally Posted by TomHuffman View Post

We are really debating a very fine point. This sums it up. Get your gamma somewhere between 2.2 and 2.4. However, as I wrote before, the higher gamma figures really require a very high on/off contrast ratio to avoid loss of shadow detail.


I'd say this is a sensible conclusion. BTW, Tom, I find the term "linear gamma" to be a poor choice. Most of us know that you mean constant over the range of output, but those who do not understand the matter may be confused. Obviously, gamma is not a linear function.
post #52 of 161
There is something wrong with this.

The difference between black and the output from 20% stimulus in my example would be about 1.25 fL which is way, way above the threshold of visibility. Differences as low as 0.08 fL are easily visible.

Quote:
Originally Posted by Bear5k View Post

Actually, your analysis only captures black crush when measuring light. From a visual standpoint, you crush through the teens and into the twenties -- 22.2% would be 4 dE (really dL*) from the black point.
post #53 of 161
Quote:
Originally Posted by TomHuffman View Post

There is something wrong with this.

The difference between black and the output from 20% stimulus in my example would be about 1.25 fL which is way, way above the threshold of visibility. Differences as low as 0.08 fL are easily visible.

It was, admittedly, an approximation, however let me bring in a couple of key points. Pardon the extended explanation since I want others who are reading this to follow the analysis.

The limit of visibility is nominally 1 dE for static images, but quite a bit of recent research questions this with the older formulae (we don't need to rehash that debate). For those who missed those discussions, the issue with having multiple revisions to the formula for calculating color error has quite a bit to do with zeroing in on just what is accurate for static items like textiles, paint, etc. The noticeability of color error is also more apparent when you have a reference within your field of view. Video/Film images do not generally have either of these, so limits of noticeability are generally held to be around 4 dE.

The other issue is the adaptability of the human eye to a white point. The eye adapts to a white point, and then how it perceives other amounts of light are "scaled" against this white point (pupil dilation and contraction is a good thing).

So, can one "see" a 0.08 ftL change in luminance? It depends entirely upon whether there is a reference image, whether the image is static and where the reference white point is at. I assure you that you will not notice a 0.08ftL fluctuation outside on a sunny summer day, no matter what the other conditions are.

I do acknowledge that 4dE is probably the maximum of what won't be visible on such a screen, but the thrust of my point absolutely still holds. Care to watch Dark City on an LCD to test the point?

Bill
post #54 of 161
Poynton quotes Rec. 709 in his Gamma FAQ, saying an "idealized monitor" uses the inverse 709 transform to de-gamma. This is the formula, with R'709 being the gamma corrected number normalized to a 0->1 range.

R'709/4.5 for R'709<=0.081 and
((R'709+0.099)/(1.099))^(1/0.45) for R'709 > 0.081

Using this method to calculate luma targets gives very different results than the power law commonly used. For example, assuming peak output is 75 ftL, 10% video yields 1.68 ftL with Rec. 709 and 0.47 ftL with power law 2.2. It gets more similar the brighter it gets, but the low end (including the tail) is rendered much brighter using Rec. 709.

All calibration software I've come across uses the power law approach.

Is this relevant for real-world displays, or only something that matters in the video processing intermediate steps? Should we try to follow Rec. 709 for brightness targets, or just try to emulate the response of a professional studio monitor? If rendering/directorial intent is what matters, following whatever a studio monitor implements should give the most accurate results, even if this substantially deviates from Rec. 709 and Poynton "idealized monitor".

Comments?
post #55 of 161
Quote:
Originally Posted by csundbom View Post


All calibration software I've come across uses the power law approach.


Not true...CalMAN offers a number of gamma functions and, in point of fact, the Power Function is *NOT* their default / standard curve...go to www.spetracal.com and play with the demo version for more information...

HTH
post #56 of 161
Quote:
Originally Posted by Joelc View Post

Not true...CalMAN offers a number of gamma functions and, in point of fact, the Power Function is *NOT* their default / standard curve...go to www.spetracal.com and play with the demo version for more information...

HTH

Actually, is is true; I've never used Calman. Thanks for the pointer.
post #57 of 161
Quote:
Originally Posted by reio-ta View Post

csundbom,

Doesn't lowest black level for a particular peak lumen reading also determine what to use? I still haven't found out what the next stim value after the 0 ftl lumen value is. I think that as long as your black level is above the next stim after 0 ftl, even if the 0 ftl lumen signal isn't 0 ftl, you're not crushing black. The higher the gamma the lower a particular % stim in lumens becomes. Your display has to be able to reach a low enough black level to peak lumens ratio, before you can go to the next gamma level. But gamma has a limit too? 2.6 is the maximum you can go, no matter how dark you black level is? Over 2.6 makes the image too "deep" and "contrasty" looking?

Yes, you're right. You have to compensate for any luminance at black. I think the standard approach is to add "luma at black" to all target values (expect 100% of course). This becomes important with displays with poor black levels and rendering in the very low IREs. My example assumes perfect blacks and is more about which transform to use.
post #58 of 161
Quote:
Originally Posted by csundbom View Post

It gets more similar the brighter it gets, but the low end (including the tail) is rendered much brighter using Rec. 709.
Is this relevant for real-world displays, or only something that matters in the video processing intermediate steps? Should we try to follow Rec. 709 for brightness targets, or just try to emulate the response of a professional studio monitor?

There are 219 digital levels per Y' component amplitude:

219 steps / 100IRE = 2.19 steps per encoded IRE.

With an ideal end-end gamma of "1" every 1% change in encoded luma (IRE) should be right at the (HVS) human visual system perceptual threshold.

The linear tail that is grafted onto the lowest region of the 709 transfer function is there to effectively mask noise by reducing gain for picture elements located there; it is implemented during capture and necessary for production and processing accuracy. The linear tail should not be part of your display response characteristic.

The perceptual model L* when correlated to the linear tails merging point, has a breakpoint that corresponds to the L* value of '8' and is considered the effective lower limit for (LDR) low dynamic range 8-bit image data.
post #59 of 161
Quote:
Originally Posted by tbrunet View Post

There are 219 digital levels per Y' component amplitude:

219 steps / 100IRE = 2.19 steps per encoded IRE.

With an ideal end-end gamma of "1" every 1% change in encoded luma (IRE) should be right at the (HVS) human visual system perceptual threshold.

The linear tail that is grafted onto the lowest region of the 709 transfer function is there to effectively mask noise by reducing gain for picture elements located there; it is implemented during capture and necessary for production and processing accuracy. The linear tail should not be part of your display response characteristic.

The perceptual model L* when correlated to the linear tails merging point, has a breakpoint that corresponds to the L* value of '8' and is considered the effective lower limit for (LDR) low dynamic range 8-bit image data.

Got it. So the Rec. 709 transfer function is useless in determining target luma values in real life situations?
post #60 of 161
Quote:
Originally Posted by Joelc View Post

Not true...CalMAN offers a number of gamma functions and, in point of fact, the Power Function is *NOT* their default / standard curve...go to www.spectracal.com and play with the demo version for more information...

HTH

For Gamma functions we support the ITU/EBU Standard, Power Function, Studio RGB and Dynamic Offset.
New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: Display Calibration
AVS › AVS Forum › Display Devices › Display Calibration › 2.5 is display gamma, NOT 2.2