Gamma values used on your systems?

I use an HTPC to playback DVD's on my NEC 110LC CRT and have gamma set to 1.40 for all.

Craig

Craigo87 says for all, so he could mean "end-to-end gamma". While you are quoting "display gamma".

Like Craig, I have my HTPC set to 1.25 (B), 1.35 (G) and 1.45 (R).

But this is a completely different Gamma value to the one you mean nidi, which is the end display gamma.

end-to-end gamma = [camera gamma]*[display gamma]

Camera gamma has a 0.45 exponent but is simplified to 0.51(correction thanks to Andrewfee see his post below)
The highlights are effectively gamma 0.5, going from highlights to shadows the camera goes from 0.5 to 1.0 gamma, and at the same time the camera gain factor slowly increases from 1 to 4.5.

Reference monitors use during mastering use gamma 2.2 (PAL uses 2.35). But consumers are assumed to be using crt gamma 2.5 (PAL is mastered to look acceptable 2.2 to 2.5). Color might look best at 2.2 (PAL 2.35), but should look ok at 2.5, using gamma higher than 2.5 may result in problems with color on some material.

Perception of contrast is effected by the viewing condition surround effect so different gamma is usually recommended for different viewing conditions.

Gamma of 2.5 for dim surround conditions, dim is defined as being between 0% and 20% of the luminance of white in the image).
Gamma of 2 for bright surround conditions
Gamma of 3 for dark surround conditions

Video gamma is used in tv, dvd, blu-ray, etc... including for films. High gamma = more contrast in lighter part of image, less contrast in darker part. Low gamma = more contrast in darker part of image, less contrast in brighter part. The transition point between raised and lowered contrast is in the middle intensitys, not hidden in near black. Always above middle 18% grey (perceived as middle 50% brightness). The average image brightness will also be lower with higher gamma. So it comes down to personal choice visibility of shadow detail and a brighter image vs more image depth in bright scenes.
Also video has increased color saturation over film, because expected to be displayed on bright crt. Gamma effects color saturation and hue as well. As gamma is increased, dark colors at the bottom end of the gamma curve get dimmer, while brighter colors at the top end of the gamma curve get brighter, so a mix of colors with each at different points in the gamma curve, shift in ratio/proption to each other, the color mix gets more saturated since the lowest out of the rgb colors determines how much white - desaturation is present, and the hue shifts towards the color highest in the mix.

Film gamma is used only in commercial cinemas, not consumer versions like dvd or blu-ray, it uses a lazy s-curve. The s-curve transition point is darker, and gets darker the less gamma is increased, well below middle grey, 11% grey at most. The middle point gamma is typically 3.0. So for film higher gamma = more contrast, lower gamma = lower contrast. Sadly dvds, blu-rays are remastered for video altering their gamma and color so may not look acceptable displayed using film gamma that would otherwise be better for home cinema.

I tweaked my CRTs down to about 2.3. With higher ANSI CR capabilities, and enough on/off CR, I prefer 2.3-2.5ish. Many people are under the wrong impression that 2.2 is the standard gamma, when it is not. I think anywhere between 2.2-2.5 is appropriate, depending on viewing conditions, display, and user preference.

I prefer to tweak my gamma to be lower at the low brightness range and progressing to higher gamma at higher brightness levels. I use 2.0-2.1 at the low end gradually transitioning to 2.4-2.5 at the higher brightness levels. I use a Lumagen RadianceXE to shape the gamma curve.

If you're simplifying it, camera gamma is 1/0.51, not 1/0.45:



(don't know why the grid has disappeared off that graph, must be a bug in the latest openoffice)

FWIW practically 100% (DVD/Blu ray) of motion picture releases are NOT engineered with the standard ITU-R 709 pre-emphasis.

The only thing in the post production chain that is Rec 709 is the video monitor the colorist was referencing. Until recently this was an industry standard Sony or Ikegami direct view CRT. This monitor (~$25K) was designed by the manufacturer to have a ruler flat 2.20 (OETF) transfer function all the way down to 1% of reference white "when calibrated".

andrewfee
"If you're simplifying it, camera gamma is 1/0.51, not 1/0.45"

Ok thanks, how do you figure out what is simplifies to. I have seen it refered to before as being simplified to 0.5, but do not understand how you get the simplified figure, I guess it is what ever plots closest to it on a graph.

Craigo87 & Mark_A_W

With your end-to-end gamma figures, are you counting camera/source gamma as 0.45 or 0.51 ?

Citation4444
"I prefer to tweak my gamma to be lower at the low brightness range and progressing to higher gamma at higher brightness levels"

I too prefer a tweaked gamma curve. I have a dlp projector so guess I should not really be posting here, using a sony dvd recorder I currently end up with a odd shape, about gamma 1.5 at 0-20, gamma 2.8 at 20-40, gamma 2.5 at 40-60, gamma 3.1 at 60-80, gamma 2.1 at 80-100. So a lumpy s-curve, still it looks better than a normal 2.2 gamma curve which is the highest gamma curve the projector can manage on its own.

tbrunet
"FWIW practically 100% (DVD/Blu ray) of motion picture releases are NOT engineered with the standard ITU-R 709 pre-emphasis."

You lost me, are you saying only the color is altered when it is remastered. No change to gamma pre-emphasis from the cinema lazy s-curve gamma?

The pre-emphasis is aka "gamma correction". No two colorist use the same emphasis.

SMPTE industry standard ITU-R Rec 709 gamma is rarely used in capturing (film or digital video) the image nor is it adhered to in the production phase.

Gamma itself is subjective.

That doesn't match up to what the EBU and ARIB say. The EBU's measurements have shown that a CRT's native' gamma is 2.35. (not 2.5 or 2.2 as some people believe)

The ARIB's measurements of Sony/Ikegami monitors showed that the gamma value is dependant on black level (0.1cd/m² ≈2.2, 0.01cd/m² ≈2.4) and that they no not follow a power curve all the way down to black:



From ARIB TR-B28 v1.1, page 95 in the PDF.


It's mentioned in a few EBU technical documents, but yes it's basically what plots closest to it on a graph.

The problem is that the Rec.709 curve (camera gamma) includes an exponent of 0.45.

For whatever reason, many people simplify this by ignoring the rest of the function and use 1/0.45, when the actual equivalent simplified curve is 1/0.51.

Great an industry "professional" including the manufacturer (Ikegami and Sony) in this case don't know how to design nor measure a display?? The info below is based on comprehensive laboratory measurements and analysis using the advanced DisplayMate Multimedia Editions together with a Spectroradiometer. The articles are written by DisplayMate Technologies President, Dr. Raymond Soneira.

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

"Television, DVD, Web and computer based photographic content are generally color balanced on professional CRT studio monitors that are electronically adjusted to have a standard Gamma of 2.20, so you’ll get the most accurate images if your display also has this value."

"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.)"

andrewfee

Thanks for the reply. I live and learn.

I believe a crt native gamma, depends on which crts you take to be typical and how you measure them. The gamma in the center of the curve or the overall average gamma, the gamma for the green phosphor or the average of all the phosphors, etc. USA says 2.2 or 2.5, and used to say PAL was 2.8 (I have no idea how the got that figure), Europe says 2.35 (and uses a different green phosphor which might have an effect or not, I have no idea) So the native gamma of crt seems to depend on which expert you ask.


tbrunet

Thanks for the reply. So do you use 2.2 as being reference, true to what the colorist saw on the professional CRT studio monitor. Or adjust gamma to suit your viewing conditions. Would you change your gamma for material mastered in Europe to 2.35.
I assume if they are doing a good job the colorist be they in USA or Europe check the image robustness, so the image should look ok at 2.2 to 2.5 gamma.
I am not a purist so go for what I prefer the look of. But have never seen a reference quality display, so do not know what I am missing.

If I'm reading this correctly, the CRT was set to almost 200cd/m² with black above 0.7cd/m² in that test?

Even if it was 200 and 0.7, that's not even 300:1 contrast!

I can get a flat 2.20 with my old CRT monitor here as well if I don't care what peak white or my black level is.


I believe SMPTE requires monitors to be set to ≈100cd/m² (30fL) and the EBU specifies 80cd/m² for video white (100cd/m² for peak white) with a minimum of 1000:1 CR for grade 1 broadcast monitors.

here's an old thread on Gamma, lots of good info here.

http://www.avsforum.com/avs-vb/showthread.php?t=836259

Nashou

The same paradigm exists for audio. If the industry does not have a "reference" then tweaks made in the production phase will not be reproduced in a predictable manner. Audio EQ or Video tonal emphasis is correlated to the real time feedback the operator is getting from his/her monitors.

Change the reference and the operator makes different adjustments.

For me at home, factory presets and calibration are fine. The manufacturer is using a profile (OETF) that compliments there design. Whether it's 1.9, 2, 2.2, 2.5 ect. is irrelevant..once calibrated the average measured gamma is what it is.

The .7 cd/m2 was for 10% video level. If you look at part 1 you will see that the measurements for that CRT as setup were 176 cd/m2 for white and .01 cd/m2 for video black (although the .01 would be down at the limits of the lght meter).

Also, from looking at the graph and calculating some things it looks like the gamma was more like 2.3 gamma down low and 2.2 up higher (like above 30% video level).

You are right about that CRT not being setup as they would be for mastering. I ask Dr. Soneira about that in email and he said that each of the monitors was first calibrated for optimum peak luminance and gray-scale at D65. So, it wouldn't have been as setup for mastering because mastering doesn't call for 176 cd/m2. I don't know how setting up the white for mastering would have affected the measurements as far as gamma.

--Darin

My mistakeI misread it and didn't realise it was starting at 10% grey there, not 0%.

I'm just not convinced that two standards bodies (EBU/ARIB) would manage to get things wrong and end up with very similar results. (2.35 / 2.4 gamma as being native' for CRTs)

Originally Posted by Raymond Soneira
"Whenever I have measured the Transfer Function of these monitors (Sony/Ikegami) it has always been a virtually perfect power-law (straight line on a log-log graph) as far down as I measured with accurate photometers, which is below 1 percent of the luminance for Reference White. See, for example, Figure 1 in Part B."

I would like to know how you can for certain say this...

All Film Cameras/Digital Video Cameras/Digital devices such as computers internally are gamma 2.2

All video games are gamma 2.2

All Digital Movies like Toy Story are gamma 2.2

I am not going to set my display to anything other then 2.2 for a small number of movie studios that don't follow the real standards during post production.


*
Gamma Curves

A set of gamma curves represent how the image data in an RGB image file translate
into measured screen brightness for each color. Commonly, an ideal gamma curve
is represented as an exponential relationship which approximates the response of
the human visual system. In this case, the gamma curve is described by a single
number which is the exponent value. In Microsoft Windows, a standard gamma of
2.22 is the most commonly used value, and this is also the standard used for the
web, video, most monitors, and Photo CD. The prepress standard is gamma 1.8
which approximates the characteristics of a printing press.

http://ftp2.bmtmicro.com/dlc/Color%20Management.pdf

Why not try it and see, you might even prefer a higher gamma, I certainly do.

I have and losing shadow detail is not worth it. You can create some cool 3D effects, but there's more stuff wrong to my eyes then correct. Things in real life don't even have that much of a contrasty look to them.

In the article you linked, he's using a meter that's only specified down to 0.01cd/m², and is working at the very limits of that. (which is generally a bad idea)

The meter used for the ARIB measurements was specified down to 0.005cd/m² and black was set to 0.01cd/m².

Gamma is fixed in CRT projectors for the most part.
Set up the gray scale for best tracking across the board and then tweak the gamma via external device.

No matter what pre-set gamma value you use there is no pj that can track the theoretical gamma cure properly. Unless you use a multi point gamma correction outboard.

Source gamma and display gamma aren't the same thing. Why do you say video games are 2.2? Are you saying the source is 2.2? If so, that doesn't mean the display is in that case even while the games are being developed.

What standards? The CRTs have a natural gamma and they have been used for years as the standard even if not written down explicitly. Choosing 2.2 for the source does not mean the CRTs were the inverse of that, especially if the whole reason for using 2.2 for the source was to end up with slightly higher than 1.0 gamma end to end. The only way to get higher than 1.0 end to end was to use a lower standard gamma for the source than the display (at least as far as this one number that gets inverted for the source and so would mean higher where I said lower if using the inverted number).

Not that I think Poynton is right about everything (and from discussing things with him I would say he is a nice guy who also doesn't claim he is right about everything), but he gave one point of his opinion here:

http://www.poynton.com/notes/colour_...FAQ.html#gamma

"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."

There are people at Microsoft who's jobs involved dealing with mastering for video and I would trust them more than the Windows group as far as things like that. But even they haven't done a study of what Hollywood's CRT monitors were doing as far as gamma last I heard. One of them mentioned about 2.4 or 2.45 for the display, but that was after talking to Poynton. And Joe Kane has said that his projectors are designed to match the gamma of CRTs and his projectors are set to around 2.5 gamma from what I have heard.

--Darin

Losing shadow detail, I believe some people with crt projectors tweak their gamma as refered to in the thread nashou66 provided a link to. So they preserve the black level while increasing the visibility of shadow detail. With digital projectors like mine that lack the huge on/off contrast ratios of crt, to track high gamma in the middle it has to have lower gamma at the bottom, which brightness up shadow detail anyway.

Overly contrasty look to them, perception of contrast is dependent on the surround effect, darker surround = less contrast perceived and I think the brightness of the white level of the projector - dark adaption of the eyes. A relatively dim projector 12ftL reference white or like mine less than 8ftL reference white in a batcave looks better with more contrast / higher gamma, since you perceive less contrast and the eyes are more dark adapted so shadow details are still visible.
A flatscreen (30+ftL white level) in a living room can have a lower gamma and look like it has the same contrast, it also needs the brighter shadow details that lower gamma gives it as the eyes are adapted to its higher screen brightness and the shadows have to overcome any ambient lighting.