Hi :) Graeme Gill wrote:
In fact I've already modified tiff gamut somewhat from the last release to accommodate Lab colorspace tiffs, making the profile optional in the case the TIFF is in that colorspace, so it would quite naturally accommodate using any embedded profile too.
Excellent - that's good to know.
In general though I was toying with the idea of simply allowing an tiff file with embedded profile to be provided wherever an ICC profile could normally be provided, simply because this doesn't introduce any command line logical issues, although it means the filename has to be provided twice.
Yes, that's a possibility. It's trivially easy to extract the profile from a TIFF if one is present, and from the point of view of an application calling Argyll's utilities to do its job, repeating the filename isn't an issue.
The only thing that occurs - and my patch didn't do things this way anyway, so this is merely a "what if" - it might be useful, for batch-mode usage, to be able to provide both a profile and a TIFF file, such that if the TIFF file *doesn't* have an embedded profile, the provided profile is used as a default.
I considered doing it this way with my patch, but instead went with the option of always using the provided profile, if present, and warning if an embedded profile is being ignored as a result - but I was torn between the two approaches!
I might do a bit of a survey first of the utilities to see what will work best before implementing a general change.
OK :)I was toying with the idea of writing a GUI wrapper for argyll's utilities, to create an optimal transform for a specific image and printer profile, for the purposes of achieving the best possible photographic prints, so I'll watch your changes in this area with interest :)
All the best, -- Alastair M. Robinson