On 12 May, Keith Packard wrote:
> 
> I wasn't planning on doing any colorspace conversion, just leaving the 
> operands in RGB space and using that.  Given that RGB<->XYZ is a linear 
> transformation, additive results in one space are equivalent to additive 
> results in the other.

I'm not following your intent.  You earlier said:

> This conversion would use tables with the same content as 
> the ICCCM XDCCC table, essentially an arbitrary array for each
primary.

Isn't that the conversion to XYZ?  (I may be misunderstanding what XDCCC
does.)

Alternate similar approaches are to convert from R'G'B' to RGB, do
arithmetic, and convert back to R'G'B'.  That might be what you meant.
This also deals with the interaction between alpha channel and the CRT
gamma inherent in R'G'B'.  (I guess we really should be properly
pedantic and refer to the component RGB in the pixel values as being in
the R'G'B' space to indicate that it is in the non-linear voltage
space.)

_______________________________________________
Render mailing list
[EMAIL PROTECTED]
http://XFree86.Org/mailman/listinfo/render

Reply via email to