Re: Action Item - Background sounds

 To Whom It May Concern -

I read with interest the note from Mr. Joe Clark.  It is clear to this
individual that just looking at a page and making assumptions is not enough
to assume that there is adequate accessibility.  Mr. Clark points to:
http://www.wpdfd.com/ as an example of a page that puts text over graphics
and stiill remains accessible - not the caser for me.  

Mr Clark sites: GRAPHIC-DESIGN BACKGROUNDS
I assure the Web Accessibility Initiative that Web designers are 
perfectly able to specify a background image that works well with 
foreground text. Example: <http://www.wpdfd.com/>.

It just goes to show that those without significant color blindness/contrast
issues are NOT in a position to say what is the best for those of us who
have this issue.  The red text on the blue background presented a real
problem for me - this was NOT an easy read!   

Doyle Burnett


 
-------Original Message-------
 
From: Joe Clark
Date: Sunday, April 20, 2003 03:22:59 PM
To: WAI-GL
Subject: Re: Action Item - Background sounds
 
>At the teleconf call it was pointed out that it looked unbalanced to 
>ask that background images be removable but not background sounds 
>(at the upper levels)

I don't understand the principle here.

My assumption is that people are worried that background 
images will interfere with the readability of text.

But has the GL group bothered to investigate the issue, or 
have they simply decided that their own vague mental image is 
sufficient to ban an entire practice?

The range of practice includes:

"CLASSIC" UNREADABLE FOREGROUND/BACKGROUND COMBOS
I suppose the example of <http://www.michaelkelly.fsnet.co.uk/>-- the 
sort of thing that is now very very hard to find-- is what the 
guideline was originally intended to address. Admittedly hard to read 
the text here.

GRAPHIC-DESIGN BACKGROUNDS
I assure the Web Accessibility Initiative that Web designers are 
perfectly able to specify a background image that works well with 
foreground text. Example: <http://www.wpdfd.com/>.

BACKGROUNDS with CUTOUTS
As seen in <http://www.chromewaves.org/chromeblog.php3>, a CSS <div> 
or even a table cell cuts a legible slice out of the background 
image. Eric Meyer knows everything there is to know about CSS and 
uses backgrounds himself: 
<http://www.meyerweb.com/eric/css/edge/complexspiral/demo2.html>. I 
would hardly place him in the category of "person impairing 
accessibility for others."

DECORATIVE USES
A subset of the others, really-- headlines and such. 
<http://rebelprince.com/> has a small background image behind 
selectable text. Really, what is the harm? (I saw an even better 
example last week and bookmarked it, but where is it now?)


Anyway, why must the page author go to all the trouble to code a 
capacity to remove background images when this is properly the role 
of the user agent and the user stylesheet?

I would appreciate it if WAI leaders could provide convincing 
reassurance that:

* This is a known accessibility issue that has affected real 
people and is not a theoretical worry.

* Forcing authors to code custom workarounds is the 
least-invasive method of solving the problem.

* This is in no way a reflexive action designed to outlaw a 
feature that makes Web sites look nice. (WAI has a habit of bandying 
about the prohibition of practices that contribute to good graphic 
design.) I am looking for proof that the guideline does not exist to 
sabotage good visual design-- and that even if such was not the 
motive, it will not end up as the result.


>It was pointed out the guidelines was "text presented over graphics" .

So you can't put words on top of pictures now?

There goes any accessible retrospective of Gran Fury's work 
for ACT UP in the 1980s. Donald Moffett and his cohorts layered text 
over images all the time.

And doesn't this mean Dean Allen can't run a photo of a puppy 
named Hugo with the word HUGO superimposed?
<http://www.textism.com/photos/?s=38>
<http://www.textism.com/photos/?s=37>

Please explain how this does not constitute an assault on 
graphic design, and how alt, title, and longdesc cannot make the 
general case ("text presented over graphics") accessible.

>Our decision was therefore to not put requirements in the higher 
>levels for speech to be separable from background sound - nor to 
>require that there be specific separation between speech and 
>background. There are captions that can be used for those who can 
>see and cannot hear well.

Good.

>1.5 Benefit
>
>#2 Speech over background sounds
>
>Because speech is often naturally mixed with background sounds and 
>cannot be easily removed or separated, captions are provided (under 
>checkpoint guideline 1.2) to make dialog understandable. However not 
>all people can see or read the captions. Where speech is mixed or 
>recorded so that it is at least 20 db above any background sounds 
>people do not need to rely on captions to understand the dialog.

What does this mean, if anything?

Suddenly you can sweet-talk your audio engineer to arrange a 
20-dB differential between speech and background and the file no 
longer has to be captioned?

The guideline, as written, implies that a 20-dB sound 
difference eliminates deafness. It eliminates the need for captioning.

I question the *entire basis* of this discussion-- with text 
and with graphics-- and urge WAI to retract both. It's another 
example of nonexperts trying to overthrow the Web as we know it.
-- 

Joe Clark | joeclark@joeclark.org
Accessibility <http://joeclark.org/access/>
Author, _Building Accessible Websites_
<http://joeclark.org/book/>

.