LibTiff Mailing List
TIFF and LibTiff Mailing List Archive
December 1994
Previous Thread
Next Thread
Previous by Thread
Next by Thread
Previous by Date
Next by Date
Contact
The TIFF Mailing List Homepage
This list is run by Frank Warmerdam
Archive maintained by AWare Systems
Thread
1994.12.14 23:51 "Re: JBIG compression", by Sam Leffler
To: tiff@sgi.sgi.com
Subject: JBIG compression
Date: Wed, 14 Dec 1994 17:00:31 EST
From: kg@mykonos.rc.rit.edu (Kyriakos Georgiou)
Is there anybody that is working on JBIG compression/decompression ?
(besides Image Magic, MA USA)
My understanding is that JBIG requires the licensing of several
patents. This makes it an unlikely candidate for a free software
package such as mine. I have, however, gotten together with the
JBIG folks so that the next distribution of the library will better
support "plug-in decoders", such as JBIG. That is, to add a new
codec you won't need to modify the library; all that you'll need is
an additional .o file and either recompilation or a substitute for
the codec configuration table.
Also, can the current libtiff G4 decompression implementation be
improved ? What sort of improvements can be expected ?
What's wrong with the current implementation?
I was surprised to find a reference to J.D.Murray, W.vanRyper,
"Encyclopedia of Graphics File Formats", O'Reilly & Associates, Inc.
1994, pp.148 which stated that some implementations of G4 encode as
fast as G3, and decode even faster.
G4 decoding can be faster than G3 depending on the algorithm used.
the above statement is, however, fairly specious.
The G3/G4 decoder in the TIFF library is built on a single set of FSM
tables and is quite fast; though the size of the state tables can cause
problems on small-memory machines. The v3.3beta021 code has some mods
to speed up the decoder. I've just received some other mods that
inline certain routines and speedup the decoder some more on certain
CISC-style architectures.
I don't want to get into algorithmic issues, my question is, is there
room for >15% improvement on G4 for the impelmentation of libtiff ?
Until you cite specific goals (and architecture for running the
software) this question is silly. Try measuring the performance of the
current algorithm before looking for improvements.
Sam