W3C home > Mailing lists > Public > www-style@w3.org > June 2012

Re: [css4-color] unclamped values for RGB

From: Rik Cabanier <cabanier@gmail.com>
Date: Tue, 19 Jun 2012 00:44:25 +0200
Message-ID: <CAGN7qDC7wyiiTeiRwtnCwwtKf1LddL45R8Dp74U23iFfh_bfkw@mail.gmail.com>
To: Alan Gresley <alan@css-class.com>
Cc: www-style list <www-style@w3.org>
Hi Alan,

On Mon, Jun 18, 2012 at 4:43 AM, Alan Gresley <alan@css-class.com> wrote:

> On 18/06/2012 5:18 AM, Rik Cabanier wrote:
>  The problem I see with the current spec, is that it is wrong and that it
>> will make decent color management in the future impossible. CMS is
>> primarily about mapping gamuts from one colorspace to another [1]. By
>> removing the clamping, the gamut is made infinite so this becomes
>> impossible.
>> I think if CSS implements the SVG color-profile property [2] (without the
>> <name>  keyword), that would be a great step forward. Old implementations
>> would ignore it, so their colors will look slightly off but would most
>> likely still be reasonable.
>> Lab would be really great since it corresponds to how we perceive color.
> Are you referring to CIELAB [3]?


> Also, the way we perceive color is not fully understood by academia (nor
> by myself). Please view this test.
> http://css-class.com/test/css/**3/colors/experiments/**
> perceptional-color.htm<http://css-class.com/test/css/3/colors/experiments/perceptional-color.htm>
> Firstly, 'example 1' is very large representation of 'example 2'.
> In 'example 3' (which is identical to 'example 2') one can perceive on the
> left of this whitish (since it's over black) region a red band and then a
> yellow band. With the same example, one can perceive on the right of this
> whitish region an aqua blue band.
> Here is a photograph of this whitish region of 'example 3'.
> http://css-class.com/test/css/**3/colors/experiments/rgb-**black-color.png<http://css-class.com/test/css/3/colors/experiments/rgb-black-color.png>
> How is this possible?

I think those are more tricks where our eyes introduce artifacts because
they're not designed to deal with such high contrast.
This has nothing to do with the perception of color.

>  This makes gradients and meshes more pleasing to the eye and has no need
>> for profiles from the user's side.
>> It's not unthinkable that in the near future there will be a race for best
>> color displays (like there now is for resolution) so it would be good to
>> be
>> prepared.
>> Rik
>> [1] http://en.wikipedia.org/wiki/**Color_management#Gamut_mapping<http://en.wikipedia.org/wiki/Color_management#Gamut_mapping>
>> [2] http://www.w3.org/TR/SVG/**color.html#**ColorProfileProperty<http://www.w3.org/TR/SVG/color.html#ColorProfileProperty>
> This is my major concern. A race for the best color displays (along with
> some specs or standards) where perception of color is not understood (or
> understood in terms of the 'classical theory of light').

not so. Color perception is well understood and Lab is designed to
approximate it. [5]

> I once stated on this mailing list that color could be generated by use of
> nanometers [4]. For this to even be done, it needs to mapped correctly to
> RGB colorspace.

I don't think there's a need to invent new color representations. People
have studied this for over a 100 years, applications have successfully
implemented it and there are many books on the subject...

> 3. http://en.wikipedia.org/wiki/**Lab_color_space<http://en.wikipedia.org/wiki/Lab_color_space>
> 4. http://lists.w3.org/Archives/**Public/www-style/2010Sep/0488.**html<http://lists.w3.org/Archives/Public/www-style/2010Sep/0488.html>
5. http://en.wikipedia.org/wiki/Lab_color_space#Advantages
Received on Monday, 18 June 2012 22:44:54 UTC

This archive was generated by hypermail 2.3.1 : Monday, 2 May 2016 14:39:00 UTC