AWARE [SYSTEMS] Imaging expertise for the Delphi developer
AWare Systems, Imaging expertise for the Delphi developer, Home TIFF and LibTiff Mailing List Archive

LibTiff Mailing List

TIFF and LibTiff Mailing List Archive
June 2010

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



Valid HTML 4.01!



Thread

2010.06.23 16:16 "LZMA2 codec for libtiff", by Andrey Kiselev
2010.06.23 16:43 "Re: LZMA2 codec for libtiff", by Scott Ribe
2010.06.23 17:14 "Re: LZMA2 codec for libtiff", by Bob Friesenhahn
2010.06.24 15:01 "Re: LZMA2 codec for libtiff", by Andrey Kiselev
2010.06.24 15:54 "Re: LZMA2 codec for libtiff", by Bob Friesenhahn
2010.06.25 10:09 "Re: LZMA2 codec for libtiff", by Andrey Kiselev
2010.06.25 02:26 "Re: LZMA2 codec for libtiff", by Scott Ribe

2010.06.24 15:54 "Re: LZMA2 codec for libtiff", by Bob Friesenhahn

On Thu, 24 Jun 2010, Andrey Kiselev wrote:
>
> On my typical data it is much worse than LZMA. Actually I do not see a reason
> to use it because of worse compression ratios. I am talking about libtiff usage
> here, of course bzip2 utility is widely adopted as a general purpose
> compressor and LZMA stuff is not on par here.

When you compared LZMA with bzip2, did you do it by compressing whole 
image files?  What type of images are you testing with?

ImageMagick and GraphicsMagick support bzip2 compression in the "MIFF" 
file format.  Compression is done on a per-row basis without any 
modification via a 'predictor'.  This makes it a good test case for 
bzip2 compression.

> If we want to bring some new codec in libtiff world we need something that has
> clear advantage over the existing ones, either compression ratio or something
> other (decomression speed, like Bob mentioned).

What I am looking for is well-balanced compression/decompression so 
that TIFF can be used as a working format while still using 
compression.  Testing with certain filesystems which implement fast 
compression (my testing is done with zfs) shows that I/O can be faster 
with the compression than without.  It is not always desireable to use 
a compressing filesystem so offering similar compression in libtiff 
could offer the I/O and storage benefits of compression without 
dependence on an OS feature.  Compressors exist which are capable of 
compressing over 100MB/second on current CPUs.

Bob
-- 
Bob Friesenhahn
bfriesen@simple.dallas.tx.us, http://www.simplesystems.org/users/bfriesen/
GraphicsMagick Maintainer,    http://www.GraphicsMagick.org/