[Openicc] color-policy vs. color-infrastructure

[Graeme Gill]

Kai-Uwe Behrmann wrote:
> Pyramidial images contain at the best antialiased versions. I dont think 
> this is of much use to determine the gamut extents of an image.

I don't see why not. The gamut of a natural reduced resolution image
is likely to be very similar to the full one. Pathological images
(likely to be synthetic images) may not of course.

Graeme Gill.