Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Path: utzoo!mnetor!uunet!lll-winken!lll-lcc!ames!aurora!jbm
From: jbm@aurora.UUCP (Jeffrey Mulligan)
Newsgroups: comp.graphics
Subject: Re: Gamma Correction of Monitors
Message-ID: <1314@aurora.UUCP>
Date: Wed, 18-Nov-87 22:44:00 EST
Article-I.D.: aurora.1314
Posted: Wed Nov 18 22:44:00 1987
Date-Received: Sat, 21-Nov-87 11:54:26 EST
References: <880002@hpcehfe.HP.COM>
Organization: NASA Ames Research Center, Mt. View, Ca.
Lines: 50

in article <880002@hpcehfe.HP.COM>, avi@hpcehfe.HP.COM (Avi Naiman) says:
> 
> I am interested in finding out what techniques people use (or know about)
> to perform monitor 'gamma correction'.  In particular, does anyone have a
> method to linearize screen luminance without a photometer (or similar
> measuring device), but simply with some visual task which is independent
> of the display hardware.  A matching task using true gray in one field
> and a dither pattern of black and white pixels in the other will not
> necessarily work, since there's no gaurantee that turning on half the
> pixels (for example) will generate a luminance reading halfway between
> min and max.  In fact, different layouts of the black and white pixels
> can lead to significantly varying luminance readings.
> 

Indeed.  However, it would seem that this failure of spatial independence
is a monitor defect.  For example, an isolated "on" pixel may be
dimmer that a similar pixel in the middle of a large "on" field
because of inadequate video bandwidth.  If this is the case,
then a simple gamma correction table (say, from photometer
measurements of large uniform fields) isn't enough to insure you
get what you want.

I have tried doing things like you propose; instead of doing a brightness
match, you can interleave frames of the white/black dither
pattern with a uniform gray, and adjust the level of the gray to
minimize the flicker.  Another approach is to do a brightness match
between rapid black/white flicker and steady uniform gray.
I tried both of these earlier this year on my monitor
(brightness match of b/w flicker and gray, and flicker minimization
of b/w stripes and uniform gray, with both vertical and horizontal
stripes).  I got a different answer with each method, and none
of the answers agreed with the photometer data from a large field.
The different results may have straddled the photometer data,
I don't remember.  We are getting a new monitor.

A problem related to the failure of spatial independence arises
in conjunction with dithering gray scale pictures for printing.
The problem is that the size of a single ink dot is often
larger that the positional resolution for drawing a dot;
thus two adjacent black dots give something less than twice as
much black area.  The is a guy at IBM Almaden Research labs
(whose name escapes me) who has worked on correction schemes
for printing accurate gray scale images.  Presumably these techniques
could be applied equally well to monitors exibiting failures
of spatial independence.

-- 

	Jeff Mulligan (jbm@ames-aurora.arpa)
	NASA/Ames Research Ctr., Mail Stop 239-3, Moffet Field CA, 94035
	(415) 694-5150