Rev and Photoshop

[Alex Tweedly]

Sivakatirswami wrote:

> Does anyone know of a way to get Revolution to read and write to Adobe 
> Photoshop's per image metadata?
>
> real world scenario:  our photographer in Nepal (world famous, doesn't 
> just work for us) Thomas Kelly... we just sent him on assignment to 
> Sri Lanka. He's back now and has a kaJillion digital images that he 
> wants to caption inside PhotoShop. Opening the file info for each one 
> of these in Photoshop is a "killer" process. I tried to have our 
> managing editor consider having the metadata as a separate 
> fooPhoto.psd /fooPhoto.txt document pair,  for which I have many Rev 
> apps that can easily preview thumbnailis and allow for data entry very 
> quickly. And though one can see lot's of advantages of having this 
> data as a separate file... the concept of the caption being part of 
> the photo file itself seems to be "ingrained" as the only way to go...

The Photoshop metadata is EXIF tags attached to JPEG (or RAW) files.
There are 2 distinct versions of it - traditional EXIF data in a binary 
format, and new (Adobe specified) XML-based version.

Currently, the new version is mostly supported by Adobe applications - 
so they generally leave the older format info in there as well.

You can extract basic EXIF info using the EXIF library I posted back in 
November (available in RevOnline - under alextweedly called libEXIF). 
Note it doesn't do many manufacturer specific data, does not handle the 
new Adobe format and doesn't handle updating the info. Changing to write 
some data would be easy - but writing exif data in general is quite hard 
(and of course carries the danger of file corruption).

I actually stopped work on that library, and switched to using a set of 
tools and utilities via shell. The ones I use are at
http://home.arcor.de/ahuggel/exiv2/   and  reportedly work (or can 
easily be made to work) on Mac ( as well as Win and Linux which is what 
is directly supported). What I do is use them to bulk-extract the info I 
want from a directory at a time, then process those separate files in 
Rev, and then use the tools to re-insert the data.

I do however agree - having the metadata as part of the file is the 
right way to do it (until we get metadata aware file systems which will 
ensure that the metadata goes everywhere the file goes). I use this 
extract-modify-reinsert as a purely transient state, I try not to leave 
significant data in the additional files for any length of time,

-- 
Alex Tweedly       http://www.tweedly.net



-- 
No virus found in this outgoing message.
Checked by AVG Anti-Virus.
Version: 7.0.308 / Virus Database: 266.11.5 - Release Date: 04/05/2005