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

Betreft: Contrast and grayscale

From: Eric Velleman <E.Velleman@bartimeus.nl>
Date: Tue, 18 Jun 2002 15:53:03 +0200
Message-Id: <sd0f578b.016@bartimeus.nl>
To: jmdamour@videotron.ca, w3c-wai-eo@w3.org


Just a very short reaction. You have a point there. Contrast sensitivity is a great topic in the world of measuring eyesight. And there are many color deficits with different intensities and combinations (to make thing more complicated). We probably should keep things a bit general and not dive to deep into science. It is my experience that it is mostly common sense. 

I looked into the different methods of measuring contrast sensitivity for paper and for screen a few years ago. For paper there are many tests for black and white as for color contrast. But screens act completely different, the most obvious beeing that they emanate light compared to paper (reflects light). This reflectioncoefficient is one of the most important things that is measured here at the Bartimeus institute when measuring contrast sensitivity as a function of the eye. 

On a screen contrast sensitivity is difficult to measure, for instance because of differences of calibration, differences of screentypes, the possibilities of screens like the number of colors they can show and how they are made and our eye can see more colors than a screen can show us, where are you measuring (external circumstances), motivation of the person doing the test etc.

Another thing i found was that when colors are 'translated' to black and white, it seems that there are differences between programs doing that. Within Adobe Photoshop there seem to be more than one ways to translate to grayscale leading to different results.

Tried to write some things down (in Dutch, sorry) at: http://www.accessibility.nl/informatie/kleuren/kleuren3.html#wieiskleurenblind
We used the Vischeck plugin for Photoshop to make the colordeficit pictures. This can also be done with a screenshot of a webpage.

Eric Velleman

>>> "Jean-Marie D'Amour" <jmdamour@videotron.ca> 18-6-2002 15:00:22 >>>


I have tested our recommandation about evaluating contrasts with grayscale 
in the context of a trainig.

I'm not satisfied about this because this kink of evaluation is very 

The checkpoint itself is rather subjective.

2.2 Ensure that foreground and background color combinations provide 
sufficient contrast when viewed by someone having color deficits or when 
viewed on a black and white screen.

What is a sufficient contrast? Advices differs and the grayscale test add 
nothing to this discussion. A relatively poor contrast that someone judged 
sufficient in full colours is also judged sufficient in grayscale.

A-Prompt has a dialog to objectively evaluate colour contrast, but this 
dialog appears only if the basic colours of the page (in the body element) 
are judged insufficient. Also, A-Prompt have choose to accept just clearly 
strong contrasts that limits a lot the range of possible colours.

I don't know how A-Prompt calculate colour contrast and I feel tha this 
evaluation is a little too severe.

Doe's someone know another objective test for colour contrast that can 
replace our subjective test with grayscale?



Jean-Marie D'Amour M.Éd.
CAMO pour personnes handicapées
Received on Tuesday, 18 June 2002 09:56:22 UTC

This archive was generated by hypermail 2.4.0 : Friday, 17 January 2020 20:29:32 UTC