This needs a higher profile – a service to show you what any website looks like to someone with various sorts of colourblindness. (Example: compare WAI and colorblind WAI.) If you use the grayscale filter, not only will you know whether any bits of your page would be unreadable to one or more colourblind people, but it would also make it obvious where you haven’t explicitly specified colours (because the default colour will remain).
This is obviously useful to web designers, but perhaps it would also be useful to colourblind people too. When they complain about a website, they could send a URL saying “this is what it looks like to me”, to give the webmaster an idea of the problem.
It does look very useful, but unfortunately is painfully slow. Are there any offline tools or at least reliable methods we can use to replicate the functionality (for example, in a graphic editor)?
Here you get Photoshop and other plugins
http://vischeck.com/vischeck/
I have also coded an Extension, but not released it at this time.
As someone who is colorblind (or colourblind), I find stuff like this interesting. I believe I have the most common form of colorblindness (Deuteranomaly… a reduced sensitivity to green) and I didn’t know I had it until I finally took a true dot test around age 30. This, of course, pleased my wife because she’d always insisted I was colorblind. Suddenly she had proof. But given the fact that I am a serious photographer, programmer, and occasional web-designer, I keep wondering when they’ll come up with something that helps me compensate for my reduced color sensitivity. I need something that shows me what a non-colorblind person sees… sort of a calibrated monitor to MY vision.
@ScW
Unfortunately it is impossible to simulate normal vision for colorblind(Dichromacy) people.
There is an algorithm to get a better contrast on images for colorblind people called Daltonize(see vischeck).
River: Are you willing to make the source code for what you’ve done so far available? Or would you consider working on it as a Summer of Code project?
Daltonize looks promising. However, that (as well as the “how that would look to someone who’s colorblind” thingy) is something that an OS or graphics device driver should be capable of, not a browser.
@Gerv
I have informed myself over the Summer of Code Project.
The Deadline for adding a Project is 26. March
http://code.google.com/support/bin/answer.py?answer=60325&topic=10729
Is there a chance to get a “application”( here ) ready before this deadline?
River: You can apply any time up to the deadline. To apply, use the form on Google’s Summer of Code site.
They appear not to be able to handle the additional exposure of a higher profile, within the limits of their available resources on the server.
This is the sort of thing that I’d like to see the Web Developer toolbar extension do. All web developers should have this capability at their disposal.
ScW: software can’t do that for you, any more than it can show a person with “normal” vision what a tetrachromat sees. It’s not like converting between colorspaces (RGB => CMYK or whatever). The conversion is strictly one-way, because a colorblind person’s eye throws away some of the information that a normal eye keeps, in a way that’s strictly not recoverable. (Yes, there can be some minor information loss going RGB=>CMYK and back, because of bit depth and rounding errors, but that’s different and, in comparison, minor. Colorblindness throws away a whole channel, more or less. There’s no way to get it back.)
Actually, if ScW has only got a reduced sensitivity to green (as he says), then you could theoretically boost the greenness of the monitor output by a compensating amount. It’s only if you have no or almost no vision whatsoever in the relevant channel that compensation is impossible.
Of course, I doubt very much that the greenness boost would be a simple “multiply green channel by 1.3”. One would need to do experiments to see how much of a boost was needed to make the response in each channel broadly the same.
@Gery
This isn’t so easy.
I’ll try to explain. Your eye have three different receptors for color vision. One for light of a long wavelength(mostly red), one for light with medium wavelength(mostly green), and one for light with a small wavenlenght(mostly blue).
Due to overlapping of the absorption spectra of the receptors you can’t say that R=L G=M B=S. For example an RGB stimulus of R=x% G=0% B=0% produces a respond from your receptors like L=y% M=z% S~0%.
So boosting the Green Channel isn’t a good Idea. But if you transform the RGB value to the LMS colorspace, maybe it is possible to “boost” now the “M”-Channel, but I don’t think that you get an accurate effect, without scientific research.
I only know algorithms that simulate total colorblindness (Dichromasie)
River: Thanks for the explanation. I still think it must be possible, if you have an accurate graph of a particular person’s LMS response and assuming one set of sensors isn’t completely broken, to produce a mapping of standard colours to their colours across some portion of the colour space. But you are right, it would be specific to a particular person and would take a lot of research.