nome cognome schrieb:
I did a test with a real photo, here it is: http://darkbasic.homelinux.com/images/sv2/090714_1_1200_adobergb_16bit.png http://darkbasic.homelinux.com/images/sv2/090714_1_1200_sRGB-relative_16bit.png http://darkbasic.homelinux.com/images/sv2/090714_1_1200_sRGB-perceptual-argyll_16bit.png tiffgamut -d1.0 -pj -cmt 090714_1_1200_adobergb_16bit.tif 090714_1_1200_adobergb_16bit.tif collink -qh -G 090714_1_1200_adobergb_16bit.gam -ip -cmt -dmt 090714_1_1200_adobergb_16bit.tif sRGB_Color_Space_Profile.icm link.icc cctiff -dsRGB_Color_Space_Profile.icm link.icc 090714_1_1200_adobergb_16bit.tif 090714_1_1200_sRGB-perceptual-argyll_16bit.tif The relative intent it's really, really better with this photo.
... but there is pretty much clipping in the red channel! here are two GaMapICC 0.3b results (50% and 100% filtering in tiffgamut): http://digitalproof.info/gamapicc/090714_1_1200_adobergb_16bit_sRGB_IEC61966-2.1_is50.tif http://digitalproof.info/gamapicc/090714_1_1200_adobergb_16bit_sRGB_IEC61966-2.1_is100.tif BTW: how do you asses the original AdobeRGB file? Klaus
