Ben Goren wrote:
Hmmm...Graeme, would that make sense for an addition to Argyll? Maybe something like an option to tiffgamut that creates a new working space profile just for that image? One could then (for example) re-process the RAW file and do a gamut-mapped conversion directly to the new space. Tedious to do by hand, but probably not that bad as part of a script that's in turn part of a batch process.
It's certainly possible, but I'm not sure that it's really tackling the right problem. The are really two distinct situations, depending on whether imagery has been rendered into a restrictive gamut or not. If the imagery has been fitted into a restricted device space (ie. it's been rendered into that space, making the best use of it), then the colorspace gamut is a good representation of the gamut of the imagery, and can be used to guide re-rendering to some other output media. If the imagery is input referred (ie. it encodes real world colors) or if it has been rendered into an output colorspace that is noticeably larger than that of the image, then the encoding colorspace doesn't represent the imagery gamut, and you'll get bad results if you use it as if it does. To get good results you have to determine a gamut that does represent the imagery. Argyll can support this sort of workflow using tiffgamut. [ Note that ProPhoto is actually the basis of the RIMM and ROMM standards for input referred and output referred encoding. See <http://www.colour.org/tc8-05/Docs/colorspace/PICS2000_RIMM-ROMM.pdf> ] Graeme Gill.