I'm curious as to why I've not encountered this before as I use lzw
compression all the time. Anyway, as always thanks for the education. :)
I've discovered that for my present project that jpeg-in-geotiff
compression is good enough, which makes me happy because even with
predictor=2 the space savings was only 10% or so.
matt wilkie
--------------------------------------------
Geomatics Analyst
Information Management and Technology
Yukon Department of Environment
10 Burns Road * Whitehorse, Yukon * Y1A 4Y9
867-667-8133 Tel * 867-393-7003 Fax
http://environmentyukon.gov.yk.ca/geomatics/
--------------------------------------------
LZW compression packs a little dictionary at the front of each strip/tile
with the table of patterns. The size of an LZW compressed image can be
noticably larger than an uncompressed image if the chunk size (strip/tile)
is small so that this overhead becomes significant, and of course if the
imagery is essentially uncompressable with LZW.
Using larger strip/tile sizes can improve efficiency with LZW.
Also, using -co PREDICTOR=2 can help with imagery that is smoothly
varying as it compresses the differences from pixel to pixel instead of
the absolute values, and these will tend to be small and have more
patterns.
Best regards,
_______________________________________________
gdal-dev mailing list
gdal-dev@lists.osgeo.org
http://lists.osgeo.org/mailman/listinfo/gdal-dev
