Tiled Image Compression Convention
-
Description:
A convention for compressing FITS images and storing the compressed
byte stream in a variable-length array column in a
FITS binary table. Optionally, the image may be divided
into rectangular tiles, and each tile is compressed separately. Any
number of different compression algorithms can be supported. Current
implementations support GZIP, RICE, H-Compress, and the IRAF pixel list compression
algorithms. The
fpack
FITS file compression utility supports this table compression convention.
The
fpack
Users Guide contains useful general information and tips regarding image compression.
-
Submitted:
Version 2.0 in November-2006 by W. Pence, NASA/GSFC (on behalf of the authors).
-
Registered: March 2007
-
Date and Place of First Usage:
circa 2001: support for this image compression convention was
implementd in the CFITSIO library that is maintained by the HEASARC.
Soon after, support for pixel lists in this compressed format
was added to IRAF.
-
Extent of Usage:
This convention has been used by various astronomical data archives, including, the
Canadian Astronomy Data Center (CADC), the Optical Gravitational Lensing
Experiment (OGLE), and the NOAO Science Archive.
-
Full Documentation:
-
Latest 2.3 version (July 2013):
PDF format,
postscript,
Latex source
-
Updates include
- added support for a new variant of the GZIP algorithm, called GZIP_2, in
which the bytes in the pixel array are shuffled into decreasing order of
significant before being compressed with gzip. This is especially
effective with floating point arrays.
- added new variant to the algorithm for dithering floating-point pixels in which the pixels that are equal to 0.0
are not dithered.
- added support for the '1Q' variable-length array FITS format, to support compressed files that are too large to
be represented with the 32-bit pointer addresses in the '1P' variable-length array format.
-
Latest 2.2 version (May 2011):
PDF format,
postscript,
Latex source
-
Modified Section 3 to document the optional "GZIP_COMPRESSED_DATA" column, which provides an alternative
compression method for floating-point image tiles that cannot be compressed with the usually quantization
method.
-
Previous 2.1 version (August 2009):
PDF format,
postscript,
Latex source
-
Modified Section 4 and Appendix A to describe a new "subtractive dithering" technique which preserves more information when
quantizing floating-point FITS images.
-
Previous 2.0 version (November 2006):
PDF format,
postscript,
Latex source
-
Comments:
FITSBITS
discussion,
-
Sample FITS Files (all files contain the same image of Messier 13)
-
Sample Header Listing:
rice compressed image header listing
Last revised: Monday, 05-Jan-2015 17:11:04 EST
Contact us: fits @ fits.gsfc.nasa.gov
Hosted by: The HEASARC
(High Energy Astrophysics Science Archive Research Center)
Responsible NASA representative: Dr. Tess Jaffe
Privacy, Security & Accessibility Statements.
Download the PDF reader.