Hi Matt,
 
> So now we can begin compressing our TIFFs? The main reason I ask is that I
have heard a lot of really
> bad arguments against compression in the past, and a lot of non-argument
(for example: "it's just not
> done"). I want to know what the good reasons against it are.

> For a small museum especially, that might save hundreds of thousands of
dollars in storage by reversibly
> compressing RAW and TIFF files, and therefore be able to preserve where
otherwise it might not be able to
> afford it, is compression acceptable? Or is it so unacceptable that nobody
should implement a repository
> with compression?

Well -- since TIFF has entered into the books as an ISO standard (see
http://www.digitalpreservation.gov/formats/fdd/fdd000072.shtml), based on
the TIFF 6 spec, there isn't a strong reason to argue against it. I don't
have a copy of the standard in front of me and what it specifies as a part
of the standard in terms of compression so if someone else wants to comment
but I suspect that compression is allowed as part of the specification.

One thing to consider is the difference between ideal practice and
best-possible practice recommendations -- in a ideal situation, you
eliminate all complex encoding (including compression) so that the final
file is (as much as possible) almost human readable. That way, there is the
possibility of "deducing" the layout of the file without any sort of key or
guide. Realistically, this is a terribly inefficient way of storing data
(see XML for examples) that often conflicts with the on-the-ground
experience of practitioners. I wouldn't eliminate compression from the
toolbox of practitioners wanting to implement a repository as it may be
necessary (High definition video, for instance, is completely unrealistic to
store in an uncompressed form).

A final thing to note is that RAW files often already have compression, so
compressing again often yields no appreciable gains and only adds a layer of
complexity.

Tim





---
You are currently subscribed to mcn_mcn-l as: [email protected]
To unsubscribe send a blank email to 
[email protected]

Reply via email to