i've done a lot of work with storing xml meta-data in the comment fields of images so i've got a little experience with this. i'm not sure if photoshop uses the standard comment field of jpgs. if it does, i've found that the unix program 'rdjpgcom' will extract it much faster than Image::Magick.

it was a while ago, but i'm thinking that photoshop actually used some other mechanism for storing comments though. i think it was some weird system of multiple numbered slots nearer to the end of the file (the standard comments field is right at the beginning).

as a course of experimentation, i'd recommend making a very small jpg in photoshop and adding some comments to it. then run it through 'rdjpgcom' and 'identify -verbose' (identify comes with ImageMagick) and lastly open it up in a hex editor. if 'identify -verbose' can't show you the info you're looking for, you might be out of luck as far as using Image::Magick.

anders pearson