RE: colors

Yup, I agree that this is a good problem statement.
What about a multi stage approach -they often work out quicker in the
end...
 
step 1, Formalize what we are intending to solve in general terms
step 2, Develop use cases - a range of scenarios that we are trying to
solve for... -this would be the place to decide which variations of
disabilities we are considering top priority
step 3, Build the algorithm
step 4, Test against use cases
 
 
 
 

All the best

Lisa Seeman

 

Visit us at the UB  <http://www.ubaccess.com/> Access website

UB Access - Moving internet accessibility

 

-----Original Message-----
From: Gregg Vanderheiden [mailto:gv@trace.wisc.edu] 
Sent: Monday, March 15, 2004 6:08 PM
To: 'lisa seeman'; w3c-wai-gl@w3.org
Subject: RE: colors



Around color, we have two different problems as I understand it

 

1 - information that is CODED in color.    (e.g.  please correct the
field below that is shown in red.) or ( the numbers in green are to your
credit, the red must be addressed) or ( the items in the picture that
are red must be fixed). 

 

And 

 

2 - places where color is used to separate figure from background (e.g.
red letters on a green background). 

 

 

The first one requires that some other form of coding is provided as
well so that the person does not need to understand color in order to
understand which items are marked as such-and-such a category.
Sometimes the color coding can be deciphered from text formatting - but
the coding may be lines or embedded in a picture such that it is not
easy to do this.     Also, many people who are color blind do not know
how to query the color coding of text. 

 

The second requires that contrast be maintained even if one does not
know for sure which cones are weak or inoperative.  Some are more common
than others, but the easiest way to avoid all this is to look at
darkness - not color.   If you use  dark color A   over light color B it
doesn't matter what the colors (hues) are.  They can in fact be the SAME
color (hue).     Using a black and white test  is NOT to see how it
would look to people with no color perception.   It is an easy way to
test for DARKNESS and LIGHTNESS irrespective of color.    If a person
doesn't have some cone - the black and white test would not be an
accurate picture of what they would see (unless they had no cones).  But
it would give you a pretty good test of darkness and lightness.   

 

We need however to 

 

1 - come up with some specific tests

2 -  determine their affect / benefit on the different types of color
blindness.

 

I do not like the idea of just picking the highest frequency color
blindness though and addressing it in isolation - forgetting the rest.
Yes we want to start there.  Yes we want to keep it simple.  But if we
can keep it simple and address the problem generically then we should.
I don't want to only solve it for some unless we can show that it is a
lot of work to address it generically and only a little to solve it for
most.

 

I think the darkness/lightness approach may help us here.  

 

Thougths?

 


Gregg

 -- ------------------------------ 
Gregg C Vanderheiden Ph.D. 
Professor - Ind. Engr. & BioMed Engr.
Director - Trace R & D Center 
University of Wisconsin-Madison 


  _____  


From: w3c-wai-gl-request@w3.org [mailto:w3c-wai-gl-request@w3.org] On
Behalf Of lisa seeman
Sent: Sunday, March 14, 2004 11:33 PM
To: w3c-wai-gl@w3.org
Subject: colors

 

re - forwarding some of my old emails / action item to the list on color


http://lists.w3.org/Archives/Public/w3c-wai-gl/2002AprJun/0371.html

http://lists.w3.org/Archives/Public/w3c-wai-gl/2001JanMar/0323.html

 

Note: It does not look like the algorithms  currently proposed will help
a lot of these user groups

 

If someone twists my arm I could work on a formula...

 

 

Important bits from past emails (just incase someone does not follow
through the links..):

 

...........................................


The conclusion must be that we do not know enough about what are real
disorders to write them off as a continuum. What we should do is inlist
help from people who do know what they are talking about. I wrote of to
Achromatopsia.org. If anyone knows more people to write to, please do.

However that said, I will give you the fruits of my labor.

Two types of photoreceptors in the eye: rods and cones. Rods, which
provide vision in dim light, have no ability to distinguish between
colors. Cones are responsible for color vision
Color vision deficiencies result from either a lack of one or more cone
types, or cones that behave differently from average

To have rod monochromacy, or congenital achromatopsia is totally
colorblind or almost totally colorblind, and they have poor visual
acuity. This is very rare.  Without normal cone vision, their eyes are
not able to adapt normally to higher levels of illumination. There are
many variations in the severity of these symptoms. There are complete
rod monochromats, incomplete rod monochromats, and blue cone
monochromats. Complete rod monochromats have the most severely impaired
vision of all achromats

They cannot distinguish any hues (e.g., blue, green, yellow and red.
They also have poor visual acuity, aversion to bright light and
nystagmus (an involuntary, rapid movement of the eyes).

Cerebral achromatopsia (unlike other  achromatopsia ) report that they
see a monochromatic world, all in shades of gray. (not accompanied by
severely impaired vision, extreme light sensitivity, or any abnormality
in the photoreceptors of the retina)

Dichromacy is a less severe form of color defect than monochromacy It is
much much more common. Dichromats can tell some hues apart. Dichromacy
is divided into three types: protanopia, deuteranopia and tritanopia. 

Protanopia and deuteranopia are red-green defects. Persons with
red-green defects have difficulty distinguishing between reds, greens
and yellows but can discriminate between blues and yellows. Protanopes
often can name red and green correctly because green looks lighter to
them than red.

Hereditary tritanopia is a blue-yellow defect. Persons with blue-yellow
defects cannot see the difference between blues and yellows but can
distinguish between reds and greens. (Less common.) 

Anomalous Trichromacy-The ability of anomalous trichromats to
distinguish between hues is better than dichromats but still not normal.
Red-green anomalous trichromacy is subdivided into protanomaly and
deuteranomaly. A third type of anomalous trichromacy is tritanomaly. In
fact, those suffering from any of these conditions do experience color,
but not in the sense that a color "normal" observer does.



Some sites claim (and seem reputable) that with color deficits, ability
to discriminate colors on the basis of all three attributes - hue,
lightness and saturation - is reduced. Designers can help to compensate
for these deficits by making colors differ more dramatically in all
three attributes


To conclude with an opinion ( It is not ready for a proposal but could
be the basis for one) 

For Dichromacy:
Avoid colors that depend on being able to differentiate:
    a.. Red-green 
    b.. blues and yellows
    c.. reds, greens and yellows 

So using colors that mix a blue with a red and a green with a yellow
would work

Or hand control over to the user (so this is more important in an image
then as a text color scheme) or provide alternative renderings...

For achromatopsia: 
Maintain high contrast in all color schemes.(Especially were control is
not handed over to the user.)
Hand control over to the user.

How does that sound?

Needs more work..

 

All the best

Lisa Seeman

 

Visit us at the UB  <http://www.ubaccess.com/> Access website

UB Access - Moving internet accessibility