W3C home > Mailing lists > Public > www-style@w3.org > December 2013

Re: [css-color] . 12-Bit deep color support.

From: Tab Atkins Jr. <jackalmage@gmail.com>
Date: Tue, 10 Dec 2013 09:33:44 +1100
Message-ID: <CAAWBYDBb7+QgGzSOpQjsQOx-xwYFEkJAEYbm8tTTwoNB0Audbg@mail.gmail.com>
To: Daniel Beckstein <daniel@beckstein-rehau.de>
Cc: www-style list <www-style@w3.org>
On Mon, Dec 9, 2013 at 8:04 PM, Daniel Beckstein
<daniel@beckstein-rehau.de> wrote:
> With upcoming 4K and OLED monitors, with 10 Bit support, it eventually would
> make sense to extend the web-colors to 12 Bit.
>
> Currently webcolors are defined 8 Bit – ‘web’ sRGB (like: #FFFFFF or
> rgba(255,0,0,1) )
>
> So I propose adding something like 12-Bit ‘extended-web’ sRGB where you
> define colors like:
>
> X#FFF FFF FFF - instead of #FF FF FF
> Xrgba(255*16,0,0,1) - instead of rgba(255,0,0,1)
>
> Supporting just 10 Bit (what is actually just needed by upcoming displays),
> would be quite difficult because of the poor 8+2 Bit format,
>
> so the color white would be X#8FF - instead of X#FFF or #FF

As others have said, CSS colors don't define the underlying storage
format.  Implementations can use however many bits they want.

Some of our syntax formats (hex, integer rgb()) lean toward 8-bit
colors, but you can always use rgb() with percentages instead, or
hsl().


On Tue, Dec 10, 2013 at 3:43 AM, Marat Tanalin <mtanalin@yandex.ru> wrote:
> A sort of advantage of existing hex notation and its direct [in fact] mapping to 8-bit-per-channel color depth is that web-developer can be specify an _exact_ color, not some approximate one. It may be nice to still have ability to specify an exact color for more deep color spaces.

Once you get past 8-bit, the ability to specify a color *exactly*
becomes far less valuable, since you start hitting the limit of human
perception.


But yeah, if we do end up needing to do something about this, we can
always extend the syntax in the future.  I don't think it's necessary,
but we'll see!

~TJ
Received on Monday, 9 December 2013 22:34:31 UTC

This archive was generated by hypermail 2.3.1 : Monday, 2 May 2016 14:39:17 UTC