AWARE SYSTEMS
TIFF and LibTiff Mail List Archive

Thread

2002.09.13 15:14 "TIFF G4 decompression problems", by Michael O'Rourke
2002.09.13 19:27 "Re: TIFF G4 decompression problems", by Andrey Kiselev
2002.09.13 21:13 "Re: TIFF G4 decompression problems", by Joel Schumacher
2002.09.16 15:25 "Re: TIFF G4 decompression problems", by Michael O'Rourke

2002.09.16 15:25 "Re: TIFF G4 decompression problems", by Michael O'Rourke

Yep, I think this is exactly what is going on.

I don't think libtiff deals properly with EOFB. At least not when using the TIFFReadScanline calls. I would expect that I could read all the lines and libtiff would just fill in white for me.

Apparently this is not that case.

Michael

Hmmm.... All blank at the end? G4 has a keyword that says, "the rest of the page is blank white space". It's an End-Of-Facsimile-Block. (EOFB)

I no longer use the libtiff utilities since I got the project working, so this may be fixed, but at the time libtiff had a bug, as did our OCR software. Neither recognizes this keyword and blew up when it encountered the EOFB. So if two implementations had the same bug, maybe others do too.

Paint Shop Pro read the images just fine though.

Everyone at the OCR company insisted my images were corrupt, which didn't make sense if PSP could read them. After having the OCR company act authoritarian about it (we know TIFF, how dare you suggest we have a bug?), I discovered that all the "corrupt" images were using this EOFB codeword. Something they didn't know about. After presenting the pages from the TIFF documentation showing it was perfectly valid, they fixed the product.

I also brought it up to this list and I thought somebody was looking at it. But several months later, somebody was reporting a similar problem and at that time it was discovered it was still not fixed. But I thought somebody was going to take care of it.

Is it possible that libtiff is still broken? Is it possible that (being these were created with libtiff) that somebody coded the EOFB into the encode, but not into the decode?

Michael O'Rourke (morourke@adobe.com) wrote:

--------------------------------------------------------------------------------

The information transmitted is intended only for the person or entity to
which it is addressed and may contain confidential and/or privileged
material. If the reader of this message is not the intended recipient,
you are hereby notified that you have received this message in error and
that any review, dissemination, distribution or copying of this message
including any attachments is strictly prohibited. If you received this
in error, please contact the sender and delete the material from any
computer.