Next: 7 Random Groups Structure
Up: 6 Data Representation
Previous: 6.2.4 Unsigned Integers
Transmission of 32- and 64-bit floating
point data within the FITS
format shall use the ANSI/IEEE-754 standard .
BITPIX = -32 and
BITPIX = -64 signify 32- and 64-bit IEEE floating
numbers, respectively; the absolute value of BITPIX
used for computing the sizes of data structures. The full
IEEE set of number forms is allowed for FITS
interchange, including all special
The BLANK keyword
should not be used when BITPIX = -32 or
-64; rather, the IEEE NaN should be used to represent an
undefined value. Use of the BSCALE and BZERO
keywords is not recommended.
Appendix H has additional details on the IEEE format.