Re: Gamma settings for monitors
Re: Gamma settings for monitors
- Subject: Re: Gamma settings for monitors
- From: "Bruce J. Lindbloom" <email@hidden>
- Date: Thu, 19 Jul 2001 09:29:57 -0500
Roger Breton wrote:
>
One last tought. I would surmise, and Bruce Lindblom and the others would
>
correct me if I am wrong, that the choice of gamma used to be more important
>
in the pre-Photoshop 5.x era where digital counts were going straight to the
>
display. Then, I agree with the notion that 2.2 was more linearly related to
>
psychometrics than 1.8.
Most CRT monitors have an inherent gamma of about 2.2. The 1.8 gamma
attributed to Macs comes from the hardware driving the monitor (e.g. video
LUTs), and therefore this 1.8 behavior comes from the computer, not the
monitor.
Since the days of matching Macs to LaserWriters are gone (this is where the
1.8 gamma supposedly came from), I can see no good reason to use 1.8 for
anything at all, except perhaps historical compatibility.
It can be shown that a gamma of 2.2 models the uniform perceptual grayscale
(CIE L*) much better than 1.8, so I would recommend its use pretty much
universally, both for RGB reference color spaces and for monitor calibration
programs.
Roger continued:
>
Now, if anyone has pointers or tips on how to measure gamma from a set of
>
measured xyY values, I would be most interested to hear about it.
I would suggest converting xyY to XYZ and then calculating the gammas using
a least squares fit of a power function to the X, Y and Z data (for R, G and
B gammas respectively). You do not say what colors you have measurements
for, but it would probably be best to have ramps going from black to full
red, black to full green and black to full blue. Contact me off-list if you
wish to discuss in more detail.
--
Bruce J. Lindbloom, Pictographics Intl. Corp.