[fitsbits] 16-bit floats {External}

[Richard J. Mathar]

The Linux octal dump (od(1)) supports in principle the half float
and the "quad float" types, 
see https://github.com/coreutils/coreutils/blob/master/src/od.c .
To check whether these binaries are actually compiled into the coreutils
of your distribution: one can test this with the syntax
   od -t fH <yourfile>
   od -t f2 <yourfile>
   od -t fL <yourfile>
   od -t f16 <yourfile>

Richard Mathar