[iaufwg] FITS conventions

[Doug Tody]

Like Steve I have always felt that color support was unimportant for FITS
and probably better left to the established graphics formats.  Thinking
about it again now however, I am not so sure.  There are many obscure
graphics formats out there already; numerous non-astronomical viewers
and packages can already handle FITS as just another graphics format.
If color support were included this software would probably support it.

Even with professional astronomical image viewers it would occasional be
useful to be able to be able to display RGB rendered astronomical images,
without having to resort to external rendering tools.  It is useful to
be able to manipulate RGB-rendered images with an associated WCS for
visualization purposes.  Multiband survey projects these days often put
considerable effort into producing such images and making them look good.
Such carefully rendered images have some value for scientific interpretation
of data, not merely as pretty images.

Currently within VO we routinely deal with both FITS and graphics images
of the same field.  To overlay the two we have to restort to the SIA query
response to pass WCS information for the graphics formats (in principle this
information can be included in some graphics formats but in practice this
is too difficult to deal with reliably).  We will always want to be able
to generate standard nonastronomical graphics formats for various purposes,
but it could be useful in some cases to have color support directly in FITS.

I think this is a relatively low priority issue, but it is not an
unreasonable feature to consider.  Just RGB encoding pixels is trivial,
but might not be sufficient - compression is an important issue which
would probably have to be considered as well.  A simple solution here
however, might be to adopt a standard graphics format (one is probably
enough) and incorporate it into FITS wholesale, so that existing standard
software could be used.    - Doug