LibTiff Mailing List
TIFF and LibTiff Mailing List Archive
May 2004
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
2004.05.18 13:53 "Re: Further doubt on bits per sample", by Bob Friesenhahn
On Tue, 18 May 2004 mayankg@iiitb.ac.in wrote: > I further have a doubt, > 1. How does the bits per sample cases of 1,4 or 5 are handled by the libtiff > library? Does the readstrip/readtile functions return the data, such that, the > pixel values are always packed in byte boundaries? Nope. For strips, pixel values are packed without any padding. > i.e. for bits per sample case of 4, are two pixel values clubbed into one byte > or one byte always contains one pixel value only? Yes, both will be in one byte. For bits per sample of three (a very odd number), two pixels will be entirely represented in a byte, along with two bits of the next pixel sample. My comments apply to strips. I am unsure about tiles. See GraphicsMagick (http://www.graphicsmagick.org/) for some free source code which makes unpacking/packing samples of any size (up to 32 bits) from a byte stream a trivial operation. There is some similar code in libogg, but the Ogg version seems to be much more complex/lengthy. Bob ====================================== Bob Friesenhahn bfriesen@simple.dallas.tx.us http://www.simplesystems.org/users/bfriesen