Est, in filters.hpp to mitchell_filter. Bicubic filter is a

common name for bspline or photoshop style filters and mitchell is different.

OK, I'll take a look at this - I still need to get some proper filtering culture.

I think rgba32f_pixel_t is a good model for hdr pixels. I don't see the 0..1 range in floating point channels as limits. And GIL does not enforce them in any way. They only define a mapping between integer and floating point channels. In this case (rescale), clamping is an "artifact" of using integer channels and filters with possibly negative weights.

That's effectively a point of view of the current state of things. I'm still not at ease with it. I believe I had problems with some algorithms that relied on channel::max_value (being 1 for bits32f channels) and values > 1 - but I need to look at all this before being more definitive :) Does anyone else out there have experience manipulating HDR images using GIL ?

I did some informal profiling, on a dual core machine, using the bicubic filter, I got this

Thanks for the information, 75-80% performance gain is not so bad for this simple code transform ! :) Cheers, Nicolas.