Still confused on gamma...
-
LIVE: AVSForum Tech Talk Podcast with Scott Wilkinson, Episode 7 Click here for details.
- Status
1 - 15 of 15 Posts
Or... if they're coming from a film print, might they simply leave the film's original ~1.5 viewing gamma intact and pass that directly through to the DVD to approximate something close to 1.25-1.3 viewing gamma on a system with 2.2 display gamma?
Hmm... this link seems to suggest that most CRTs (VGA computer monitors anyway) have more in the neighborhood of a 2.2 gamma (+/- 0.2) on average, rather than 2.5. Which would make the PC's viewing gamma ~1.0 on average, rather than ~.88. If that applies to TV CRTs as well, then the viewing gamma with ~0.5 corrected video sources still ends up only being about 1.1 (0.5 camera gamma * 2.2 CRT/display gamma = 1.1 viewing gamma), rather than the target of ~1.25 for a dim surround... assuming no other correction has been applied.
Perhaps this isn't such a bad thing though, if you don't have much contrast to work with.
FWIW, my Sony 34XBR800 CRT appears to have only about a 2.25 gamma.
Perhaps this isn't such a bad thing though, if you don't have much contrast to work with.
FWIW, my Sony 34XBR800 CRT appears to have only about a 2.25 gamma.
Some related info...
Updated Gamma Tutorial from PNG spec:
http://www.libpng.org/pub/png/spec/1...aAppendix.html
Note that in this newer revision of the PNG gamma tutorial the estimated gamma of CRTs has been changed from the previous version , from 2.5 to 2.2. And the resulting estimated viewing gamma (or end-to-end exponent) for TVs in a dim surround adjusted from 1.25 to 1.14. Perhaps due to the influence of this document: http://www.color.org/sRGB.html
So apparently I'm not the only one who's still a bit confused... Or perhaps the video guys don't quite agree with the Internet guys on exactly what the average gamma of a CRT is.
Updated Gamma Tutorial from PNG spec:
http://www.libpng.org/pub/png/spec/1...aAppendix.html
Note that in this newer revision of the PNG gamma tutorial the estimated gamma of CRTs has been changed from the previous version , from 2.5 to 2.2. And the resulting estimated viewing gamma (or end-to-end exponent) for TVs in a dim surround adjusted from 1.25 to 1.14. Perhaps due to the influence of this document: http://www.color.org/sRGB.html
So apparently I'm not the only one who's still a bit confused... Or perhaps the video guys don't quite agree with the Internet guys on exactly what the average gamma of a CRT is.
Joined
·
3,340 Posts
(without having read any of your links) I imagine there's considerable difference between the gamma of "television" CRTs built for viewing broadcasts/DVD vs. CRTs built for use as computer monitors...
That thought occured to me as well. However the last link above suggests that both computer and TV CRTs have about the same gamma of 2.2 (+/- 0.2) and that differences in measurement are due mostly to "user error" in Brightness and Contrast adjustment... which I can somewhat believe since changing Brightness gives a different gamma reading on my 34XBR800. (FWIW, in typical viewing conditions, with Brightness/black level adjusted for low ambient lighting, my 34XBR800 gives a gamma reading in Pro mode of about 2.25 which is well within the average of 2.2 +/- 0.2.)
The main reason I was interested in average CRT gamma is I was hoping it might provide some more insight into the correction used on DVDs, and the difference between DVD video in an overlay and other software applications on my PC desktop. If I had to estimate the difference visually, I would have guessed that it was somewhere around 1.25, or possibly higher, rather than just 1.14. Most mainstream DVDs seem to have pretty consistent correction though. So they seem to be following some kind of standard, maybe something like the ~0.5112 camera correction standard. And I guess other apps on my PC are probably using 1/2.2 or ~.4545 correction. (2.2 is the monitor gamma setting I use in Photoshop to match other desktop applications, so that seems about right.) And the difference between those two correction levels (~.5112 & ~.4545) works out to something in the neighborhood of 1.125 (pretty close to 1.14).
The main reason I was interested in average CRT gamma is I was hoping it might provide some more insight into the correction used on DVDs, and the difference between DVD video in an overlay and other software applications on my PC desktop. If I had to estimate the difference visually, I would have guessed that it was somewhere around 1.25, or possibly higher, rather than just 1.14. Most mainstream DVDs seem to have pretty consistent correction though. So they seem to be following some kind of standard, maybe something like the ~0.5112 camera correction standard. And I guess other apps on my PC are probably using 1/2.2 or ~.4545 correction. (2.2 is the monitor gamma setting I use in Photoshop to match other desktop applications, so that seems about right.) And the difference between those two correction levels (~.5112 & ~.4545) works out to something in the neighborhood of 1.125 (pretty close to 1.14).
I guess that's my real question here... What gamma correction is typical on mainstream film-based DVDs?
Is it ~0.51 (the current standard for in-camera video correction), which apparently yields a viewing gamma of about 1.125?
Or maybe something a little higher to approximate the viewing gamma of film in a darker environment? (If my math is correct, then the correction needed to achieve a "film-like" 1.5 viewing gamma on a 2.2 display is on the order of ~0.68.)
Or is it somewhere in-between these extremes, and designed to acheive something more in the neighborhood of a 1.25 viewing gamma?
Anybody have a guess on this?
Is it ~0.51 (the current standard for in-camera video correction), which apparently yields a viewing gamma of about 1.125?
Or maybe something a little higher to approximate the viewing gamma of film in a darker environment? (If my math is correct, then the correction needed to achieve a "film-like" 1.5 viewing gamma on a 2.2 display is on the order of ~0.68.)
Or is it somewhere in-between these extremes, and designed to acheive something more in the neighborhood of a 1.25 viewing gamma?
Anybody have a guess on this?
Why would they have been different? At least for the past half-decade they've been using video cards from manufacturers born and bred in the PC world. There isn't anything original in hardware for the computers, except for the planned-for upcoming EDID chip on the Apple-Intel Personal Computer. 
I'm pretty sure gamma has been discuessed in great detail on the $3500 & Up Digital projector forum. Its all very interesting, and handling different displays when their gammas are different should be too difficult provided the gammas in the rest of the video chain are held constant.
I'm pretty sure gamma has been discuessed in great detail on the $3500 & Up Digital projector forum. Its all very interesting, and handling different displays when their gammas are different should be too difficult provided the gammas in the rest of the video chain are held constant.
Quote:
Much of this may apply whether you're using a CRT or some other type of display technology. I don't know as much about FPDs, but I believe they also have to emulate the ~2.2 power curve of a CRT to work with the same video content. If you have a 2.2 PC connected to your display at it's native progressive resolution you may be able to get an idea of the display's gamma from a pattern like this . There are alot of incidental factors that may effect/skew the reading though, including angle of view, Brightness/black level if not correctly set, dynamic contrast/black enhancement in the display, pixel interpolation, etc.
Quote:
I believe that's because the Mac uses a hardware Look-up table (LUT) that handles part of the gamma compensation, and the PC doesn't. I think Mac users can now choose from a variety of LUTs/display gammas though, including a 2.2 display gamma. There's a bit more about that in the links above.
Still hoping someone can shed some light on what level of correction is common in movie titles on DVD. Having some idea about this might simplify the gamma and contrast adjustment of a display.
| Originally Posted by Kid Red But I no longer have a CRT so it's hard to say. |
Quote:
| Mac gamma and PC gamma are (were) different, I think mac used 1.8 not 2.2 |
Still hoping someone can shed some light on what level of correction is common in movie titles on DVD. Having some idea about this might simplify the gamma and contrast adjustment of a display.
A correction to something I said earlier (and already deleted) after digesting some of this some more...
If your display gamma is ~2.2, then ostensibly your "viewing gamma" (or "end-to-end exponent") will probably be about 1.0 with normal desktop apps on a PC, and about 1.125 with standard .51 corrected video content.
If your display gamma is closer to 2.5, then your viewing gamma will probably be about 1.136 (not .88 as previously stated) on most PC apps, and about 1.28 on standard .51 corrected video. Meaning it will look a bit darker and more contrasty than on a 2.2 display, which may be a desireable thing from a HT standpoint.
If your display gamma is ~2.2, then ostensibly your "viewing gamma" (or "end-to-end exponent") will probably be about 1.0 with normal desktop apps on a PC, and about 1.125 with standard .51 corrected video content.
If your display gamma is closer to 2.5, then your viewing gamma will probably be about 1.136 (not .88 as previously stated) on most PC apps, and about 1.28 on standard .51 corrected video. Meaning it will look a bit darker and more contrasty than on a 2.2 display, which may be a desireable thing from a HT standpoint.
Joined
·
9,908 Posts
Quote:
I'm no gamma expert, but Kid Red is right. There is (or at least was) a difference depending on which computer platform one used. If I'm remembering correctly, video originating on a Mac (to be displayed on a PC) needed to be tweaked to look acceptable, otherwise it would be too dark.
| Originally Posted by TheFerret Why would they have been different? At least for the past half-decade they've been using video cards from manufacturers born and bred in the PC world. There isn't anything original in hardware for the computers, except for the planned-for upcoming EDID chip on the Apple-Intel Personal Computer. |
Re 2.2 vs. 2.5... I did a little experiment to see what difference black level makes in the gamma reading with the above pattern on my 34XBR800.
if I use RGB=0 as my black reference, then I get a gamma reading from the above pattern of about 2.25.
If I use RGB=16 as my black reference, then I get a somewhat higher gamma reading closer to 2.5.
Perhaps that's the source of some of the confusion re the average CRT gamma, because many of the video guys may be using something like RGB=16 for their black reference, while the Internet guys are using RGB=0. I don't think patterns like this were designed with a 16-235 "video palette"* (and it's lower Brightness setting) in mind though. So the 2.25 reading is probably a more accurate measurement of my CRT's true gamma curve.
*I put "video palette" in quotes btw because some video sources such as an HTPC may output video with either a 16-235 palette, or a 0-255 palette.
if I use RGB=0 as my black reference, then I get a gamma reading from the above pattern of about 2.25.
If I use RGB=16 as my black reference, then I get a somewhat higher gamma reading closer to 2.5.
Perhaps that's the source of some of the confusion re the average CRT gamma, because many of the video guys may be using something like RGB=16 for their black reference, while the Internet guys are using RGB=0. I don't think patterns like this were designed with a 16-235 "video palette"* (and it's lower Brightness setting) in mind though. So the 2.25 reading is probably a more accurate measurement of my CRT's true gamma curve.
*I put "video palette" in quotes btw because some video sources such as an HTPC may output video with either a 16-235 palette, or a 0-255 palette.
1 - 15 of 15 Posts
- Status
Join the discussion
