Couple of thoughts here: > I read through the UPDIG recommendations and found it really interesting > and helpful. I thought their recommendation for RAW format was > relatively unconvincing, though. Almost like they were saying "we want > to recommend RAW format, but we realize you're going to convert them > anyway, at least until the DNG format is widely-supported." Their best > arguments for RAW applied to oddball cameras--which to me is an argument > not to buy an oddball camera. Is anyone behaving differently, and > storing files in RAW (but not also storing in TIFF)? I think, although > I'm not sure, that the UPDIG Working Group has more faith in RAW than > the museum and library worlds do.
We generally store tiffs as our archival masters but we don't throw out device-specific files either (RAW being one of them). From a photographer's perspective (putting on a different hat for a moment), the RAWs hold so much valuable information that gets lost in the conversion that at long as manufacturers are making RAWs available, they're keepers. > The other question I've been asking myself a lot lately, but haven't > seen addressed much, is why not store files with some form of reversible > compression like zip (or gzip or bzip2)? UPDIG doesn't address this > (although it allows that compression is valuable and acceptable for > delivery). ZIP (and bzip2 and gzip) is perfectly reversible, and it's > tried and true. Why store 100Mb TIFF files when we could be storing 10Mb > tiff.zip files? Has anyone out there opted to use reversible compression > in digital repositories? If not, why not? There are a couple of reasons why compression can be a bad thing. The first is the issue of intermediary levels of complexity which add to the preservation problem -- something that Howard Besser put forward in discussions of preservation. The second is that most compression schemes are proprietary and patented; the result being that they cannot be easily implemented without cost. Zip is a good example of this -- it's based on the LZW algorithm which until very recently (2004 I think) was held by Unisys. It's only been in the last year that people could start thinking of using the LZW algorithm freely. Tim --- You are currently subscribed to mcn_mcn-l as: [email protected] To unsubscribe send a blank email to [email protected]
