Re: [lojban] color

[Nathaniel Smith <njs@uclink4.berkeley.edu>]

On Thu, Mar 14, 2002 at 10:17:29PM -0500, John Cowan wrote:
> Jay Kominek scripsit:
> 
> > x3 is exactly what you're asking for. Yes, RGB is borked, but I believe it
> > still makes an acceptable default, as most people are familiar with it,
> > and do not need the kind of precision required.
> 
> What's wrong with RGB at gamma = 1?

</lurk>

There are two main problems:
1) No precision. Every monitor in the world has a different space
called "RGB", and the same RGB triple can look very very
different on different devices. (Fixing gamma does help here,
somewhat, but not that much -- there are still differences in
phosphors, etc. And let's not get into environmental lighting
conditions...)
2) Poor gamut. There are lots of colors which aren't expressible
using RGB (at least, not without serious cheating). This
includes all the most intense (saturated) ones.
Nice picture (look for "gamut"):
http://www.adobe.com/support/techguides/color/cms_glossary/G.html
(The solution to both of these problems is to use one of the CIE
spaces, like CIEXYZ or CIEL*a*b* or CIEL*u*v* or ... These spaces
often tend to be "more intuitive" as well, ie, correspond more
closely to perceived color.)

These problems, of course, are likely irrelevant to this discussion
:-). (Though, honestly, I can't imagine that people will ever use
this selbri aside from describing how to set a web-page background or
something. Most color space coordinates really aren't intuitive --
if you don't believe me, then quick, what's the difference between
tan and brown in RGB? And what are the coordinates for purple?
Would you be able to answer questions like that usefully while
half-way through a sentence?)

<lurk>

-- Nathaniel

-- 
So let us espouse a less contested notion of truth and falsehood, even
if it is philosophically debatable (if we listen to philosophers, we
must debate everything, and there would be no end to the discussion).
-- Serendipities, Umberto Eco

This email may be read aloud.