W3C home > Mailing lists > Public > w3c-wai-ig@w3.org > October to December 2002

Re: Colors and the WCAG

From: Charles McCathieNevile <charles@w3.org>
Date: Sun, 20 Oct 2002 11:50:47 -0400 (EDT)
To: Joe Clark <joeclark@joeclark.org>
cc: WAI-IG <w3c-wai-ig@w3.org>
Message-ID: <Pine.LNX.4.30.0210201058570.21572-100000@tux.w3.org>

On Sat, 19 Oct 2002, Joe Clark wrote:

>>>..which is virtually impossible for a layperson to estimate.
>>>Quickie question to the WAI: What exactly do other people see?

and then
>The WAI must exercise much more caution on this question. WCAG
>readers should not be encouraged to jump to the conclusion that
>certain colour combinations are guaranteed to be safe, a claim it is
>empirically impossible to prove.

Correct. It is possible to empirically prove that someone has a problme with
things in many more cases than it is possible to prove that something is
perfect for everybody.

>>  I believe (reaching into my memory, which tells me someone posted a
>>proper reference to this list once which I think is the basis for
>>what I am saying) that having something like a 50% or greater
>>difference in values of saturation and brightness, combined with a
>>90-degree difference in hue, provides enough contrast that most
>>sighted people can usefully distinguish things.
>Can you tell me where in Photoshop or Fireworks I can set those
>things up? Just real quick?

In Dreamweaver and Graphic Converter, you can select colours on a colour
wheel - "HSV" or "HSL"  picker in Dreamweaver, HSV in Graphic Converter
(other methods for selecting colours involve adjusting levels of different
colours that are mixed, or picking from a set of colours).

The saturation is the distance from the centre of the circle. The hue is the
angle around the circle. There is a slider for "value" or "lightness" - these
are different ways of measuring what I called "brightness", but with this
rough technique use either of them.

I'm pretty sure photoshop has this functionality, and I believe Fireworks
does too (I don't really use the copy of Fireworks Macromedia kindly gave me
and I recently uninstalled it to save disk space, but I am sure someone can
verify this. If you buy me a copy of photoshop I could check that for certain

As I pointed out in the following sentence this isn't enough for everybody,
but it is a first approximation.

>>  Of course there are some people who use black and white only,
>I'd like to know where they're getting their monitors. Ever tried
>buying a greyscale monitor these days? (Not from an antique store, I
>mean.) Unless of course you mean Lynx or equivalent, which can and do
>run in full colour. (I could give you a screenshot of my own usage.)

The people I know in this situation buy ordinary monitors, and set everything
to black and white. (This is an option in various systems and screen
magnifiers, as you have pointed out before).

>>  and large size (at least there are some friends of mine who do
>>this. I suspect they are not the only ones.)
>"Large size" and "colour" have little correlation.

In practice that isn't true (which is how people can print colour images
affordably). The issue is not colour per se, but contrast.

>In practice, even confusable colours (a spectrum that extends beyond
>red and green) can be used together if other factors make confusion
>impossible or, since we cannot definitively establish what other
>people see, manifestly unlikely.

Well, we can establish that there are things people cannot see. Testing for
common forms of colour-blindness isn't infallible, of course, but it does
tend to work in practice. It is manifestly unlikely that the use of a couple
of contrasting colours (or even a handful) will cause problems for more than
a few people. It is clear that those few people will rely on other features
to identify differences. (For example, flattening text colours to foreground
and background, then using font-style and layout clues to identify what the
text does, or using different patterns in monochrome images like some maps

It is true that WAI work should be based on extensive research. It is also
true that people are looking for guidance to build pages today (and were in
1998 and in 1994) that are as accessible as they can make them.  That's why
the Techniques documents are living documents that get updated. It was not
helpful in 1998 to say "we won't tell you anything about colours until 2002
when Joe Clark's book is published" so we put together the best information
we had available, and we work on updating it as we learn more.


Received on Sunday, 20 October 2002 11:50:48 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 13 October 2015 16:21:21 UTC