From: <nav...@gm...> - 2003-03-09 22:21:06
|
hi all, downloaded lprf109w32 plus the patched qtscannerprofiler1.exe and tried profiling my nikon ls-2000 filmscanner and a fuji digitalcamera. i have a basic understanding of a profiling workflow and followed the tutorial, but my problem is that only a very good (ie. no color cast, good dynamic range) input image of the target leads to an acceptable icc profile in the end, which is sort of pointless. in more verbose terms: if i load an image of my wolf faust reflective target into qtmeasurementtool, pick rgb values, fire up qtscannerprofiler, load the apropriate reference file, generate the icc profile (various parameter settings tried) and proof the image again in qtmeasurementtool i get quirks ranging from jumps in gradients to black spots to completely black images, depending on the quality of the input file (also big differences in dE). i understand that a better input image should lead to better interpolation between color patches, but shouldn't the test target patches be proof displayed correctly even on suboptimal input images? after all, it's the patches that define the rgb->l*a*b conversion. any help/general enlightenment would be appreciated. thanks, chris -- +++ GMX - Mail, Messaging & more http://www.gmx.net +++ Bitte lächeln! Fotogalerie online mit GMX ohne eigene Homepage! |