On Fri, 12 Nov 2004 01:37:32 +0200, Lasse Karlsson wrote: > > http://www.mts.net/~cbrodgen/s1.jpg > > I have assumed that information simply is irrepairably lost (maybe some > clusters got tangled up with some other file which stole the bits that > now are missing).
That's too small a corruption to be an entire cluster. That file is about 131KB, so it's only a few dozen clusters. That corruption looks like one, maybe two, 8x8 "blocks" of pixels got mashed somehow, then the "progressive" characteristics of the JPEG algorithm were off, and it never really recovers when decoding. > and the only way to deal with it would be to open it in an image editor, > select the corrupted part and see what can be done to it. That is probably about it, but I haven't ever written a JPEG decoder for real, just studied the code and algorithms a little, so I'm not sure how bad it would be for a decoder to try to accommodate small discrepancies like that. The problem is that the decoder probably can't even detect that something has gone wrong, other than doing the decode and then comparing the decoded "block" with the ones around it, and that's surely not foolproof. TTYL, DougF KG4LMZ
