W3C home > Mailing lists > Public > w3c-wai-gl@w3.org > April to June 2002

Re: Color & contrast- response to Lee's message

From: Al Gilman <asgilman@iamdigex.net>
Date: Wed, 17 Apr 2002 10:58:02 -0400
Message-Id: <200204171458.KAA2502518@smtp2.mail.iamworld.net>
To: "'Web Content Guidelines'" <w3c-wai-gl@w3.org>
John's message is right on the mark.

There are good reasons why RNIB recommends yellow text against a black background as the ultimate low-vision styling -- the one that will get to the people just short of where there is no know way.  [If I recall correctly w.r.t. RNIB allegation]

Here is the way I wrote it up just the other day for another person 


4.  Remember that the basic high contrast mode for text is light text on a dark background.  In fact the conventional wisdom is yellow on black.  Stick figures light on dark will preserve recognizability through more low pass filtering than will more equal light-dark areas with less border length.  A plus sign is more different from a circular disk than is a solid hexagon or octagon.  Think in terms of how far the light has to fuzz to blend indistinguishably into the fuzz it is meeting from the next light area.  Light diffuses, dark doesn't.  One of the factors or interface transformations you should be thinking about is "at what point does the interface go through a light-dark inversion phase boundary" between dark on light and light on dark.  You won't want to deny light on dark to those who need it and you won't want to burden those who don't need it with this mode.  And it's a discrete leap, not something that happens continuously.


At 10:00 AM 2002-04-17 , john_slatin wrote: 
I meant to send this to the list yesterday and not just to Lee, but evidently hit 'Reply' instead of 'Reply-all'.
==Message begins==
Lee, thanks for the Microsoft reference.  I assume that the snippet below is what you're referring to as grounds for the statement that text should always
be darker than the background:
"It is clear to everybody that black text on a white background is extremely easy to see, which is part of the reason why most of the printed material you
read is black text on a white background. Likewise, white on black produces high contrast, but it is more difficult to read because black is perceived
as being heavier than white and, thus, squeezes it out a little. "
I'm sure this is true for most people, especially those with so-called normal vision. I am equally certain, however, that *some* people with limited vision
find it far easier to read light text against a dark background.  If for example you look at CCTV (closed circuit TV) systems that display magnified images
of print material on computer and/or TV screens, you'll find that many of them have switches that allow users to choose between dark on light and light
on dark, as well as choosing different color combinations.  The same is true for many software based screen magnifiers such as AI Squared's ZoomText. 
And Kara Pernice Coyne and Jakob Nielsen report in _Beyond ALT Text_ (Nielsen-Norman Group, 2001) that users with low vision often switched back and forth
between "normal" and "inverted" video (i.e., between dark on light and light on dark) to ease eyestrain and fatigue.  I used to do that when I could still
see enough for ZoomText and CCTV to be worth the trouble.  And I used to use MS Word's settings to display white text on a blude background-- the option
was there to ensnare Word Perfect users for whom those were the default colors, but it was good for me, too.  I finally had to stop because JAWS couldn't
handle it any more....
So perhaps the success critierion should involve giving users the ability to choose between light on dark and dark on light, and to make that choice readily
available rather than a one-shot deal.
Sorry for length of message.
John Slatin, Ph.D.
Director, Institute for Technology & Learning
University of Texas at Austin
FAC 248C, Mail code G9600
Austin, TX 78712
ph 512-495-4288, f 512-495-4524
email jslatin@mail.utexas.edu
web http://www.ital.utexas.edu
Received on Wednesday, 17 April 2002 10:58:06 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 16 January 2018 15:33:41 UTC