Ben Goren wrote: > > Graeme, I think the new extrapolation code needs a bit of work. It didn't > always render > input white as output white, and there were times when it seemed like there > might have > been some inversion going on in the highlights. It did do quite swimmingly > when fed a > profile with an added D50 patch. Oh -- and you hinted at this, but there > doesn't seem > to be any difference between profiles with and without -u. Hi, well, all input profiles are meant to render the chart white patch as D50 using relative colorimetric. With the current code there will be a difference between -u and non -u if the estimated device 1,1,1 has a higher Y than the white point patch. I'm now wondering if that is good enough - it should really be whatever color within the device gamut has the largest Y value, but given the extrapolated nature of it, is not correct to search the whole cLUT - just the Y headroom in the "white" direction would seem relevant. I'm not sure I can do much more without a specific example in regard to the extrapolation code - ie. a .ti3 and some indication of what you think is wrong with it. (ie. I spent some time tuning the extrapolation code, and I'm currently at a loss to know in what way it is lacking, given the dubious nature of extrapolation in the first place.) Graeme Gill.