I haven't read all the thread so maybe somebody have said the same thing.

I have read a nvidia paper (i am pretty sure that the link came from
this list). They give internal data type used by there shader. It's
quite udge. There is more than 20 type : pixel in RGBA 8 bits per
channel, RGB 16 bits floatting point, 32 bits, a "compressed" pixel
format, etc...

I think if nvidia use those different format internally it's to
minimise the memory bandwith consumption. Maybe YUV conversion should
be done at the very end of the pipeline. So you will need RGB->YUV, if
you can't make filter directly in YUV format (mainly used a YUV image
as texture).

Nicolas Boulay
_______________________________________________
Open-graphics mailing list
[email protected]
http://lists.duskglow.com/mailman/listinfo/open-graphics
List service provided by Duskglow Consulting, LLC (www.duskglow.com)

Reply via email to