Bob Talbot <BobTalbot@xxxxxxxxxxxxxxxxxxx> writes: >> It varies. And the Fuji S2 "RAW" file is *after* interpolation to a >> rectangular 12MP image, so 50% compression really makes *perfect* >> sense to me. > > > Ah > > A) For some reason I'd understood that Fuji raw files were actually > raw after all: on Dave Coffin's DCRAW pages he refers to >>fujiturn.c -- rotate Fuji Super CCD images >>An alternative to dcraw's built-in Fuji rotation. > http://www.cybercom.net/%7Edcoffin/dcraw/ > > - Why would Fuji bloat their RAW files before writing them? > - Does that not defeet the whole object? I should say that it's my *impression* that they're after interpolation; I could be wrong, I don't have a source and I haven't taken apart the internals of the format. However, the files are nearly all just a little over 12MB. I suppose the other possibility is that they're storing 16-bit values from each sensor location. Nothing in the documentation even *tries* to pretend that they get real 16-bit data out of the sensors :-). As to why, the obvious reason is they were afraid that a real RAW format would be too weird and unlike others, and wouldn't be supported by third-party software because it would be a lot of extra effort. Whereas if they did the weird part first, the RAW format would be pretty much similar, except for size. However, this assumes that the software developers and their management recognized the importance of third-party support for RAW format, something that not all camera companies seem to recognize. > B) I loved the naïf quote from another site BTW: > "RAW files are the unprocessed output of digital image sensors." > http://www.imatest.com/docs/imatest_instructions.html > > - Well, that's how you would think it should be :o) It's probably as much as the vast majority of people need to know, at least to start with. > C) I've tried Google but failed to find any definitive clear statement > about Fuji's RAW format. > For the Super CCD you might expect a TRUE unprocessed raw file to > contain all the original sensor data (so for instance the images can > be reprocessed in their entirety with any new software upgrades from > Fuji :o). If FUJI really do rotate by 45degrees (aka bloat) their > sensor data before saving as "RAW" it's easy to understand how you can > compress upwards. As you are aware I'm sure David: you can't create > information from nothing. An upsampled image is indeed "hollow" and > easily susceptible to compression. Yes, that's why I suggested that neatly explained the compression numbers. > D) There simply are no (zero) technical reasons for encrypting image > files (unless the user chooses the option). The only real reason is > control - loss of freedom - and exploitation. It probably is the > future with the way software-patent law is going. What surprises me > is not that the manufacturers are trying it on but that the majority > of consumers don't seem to give a damn. I suppose it's how people like me buy cars even though they depend on computers and can't really be worked on by a back-yard mechanic any more. People just don't seem to care! -- David Dyer-Bennet, <mailto:dd-b@xxxxxxxx>, <http://www.dd-b.net/dd-b/> RKBA: <http://noguns-nomoney.com/> <http://www.dd-b.net/carry/> Pics: <http://dd-b.lighthunters.net/> <http://www.dd-b.net/dd-b/SnapshotAlbum/> Dragaera/Steven Brust: <http://dragaera.info/>
