2005.01.29 15:42 "[Tiff] almost finished hacking libtiff into a convivial shape", by Antoine

2005.01.30 10:00 "Re: [Tiff] almost finished hacking libtiff into a convivial shape", by Antoine

This might be useful, "yet another tag viewer", http://www.apexintsoft.com/tiffprop_features.htm, perhaps, it will display problems in file structure but not image data..

I write in part because it's interesting to see how you are using the libtiff features to validate the TIFF file's conformance as input to another program that is not to be modified. I'm working on some software for that kind of purpose, validation and reformatting. The idea that you need the IFD at offset 8 and similar types of things gives me some ideas about that. Thanks for your input.

Thanks for that. It gives a little more detail than tiffdump. I looked at a number of files generated from the one original file - the original compressed file, an 8-offset uncompressed file, the original file copied (which recompresses it?), a compression of the uncompressed 8-offset (which is not forced to 8-offset), and a compressed copy forced to 8-offset. I come up with the following:

The files are identical except for two things - file size, and offsets/counts - the second being expected. The strip offsets for the valid and invalid compressed files match exactly, as do their stripbytecounts (i.e., apart from starting in different places strip offsets are no different.). All the file sizes are different though - with the invalid being the smallest of the compressed files. Is this a path to investigate? I am going to have a closer look at compression but this is getting out of my league.

Cheers
Antoine

ps. I might try putting rowsperstrip to a huge value and see what that does...