[iaufwg] FITS conventions

[Doug Tody]

On Tue, 14 Feb 2006, Lucio Chiappetti wrote:

> DT> Currently within VO we routinely deal with both FITS and graphics
> DT> images of the same field.  To overlay the two we have to restort to
> DT> the SIA query response to pass WCS information for the graphics
> DT> formats
>
> Would you mind to elaborate on this ?
>
> One of the reason I like to use FITS is that it contains information which
> is useful to science (WCS and data) instead of aesthetic one (colour).

I am just suggesting this as something to think about. I don't think this
is a terribly important issue, but it would not be all that hard either.

The type of graphics rendering I refer to is only for visualization;
however survey projects often generate these images for multiband data,
and put quite a bit of work into making them look good.  They are useful as
a preview and and as a tool for visual understanding of a field.  When we
render a FITS image into an image display this is all we are doing anyway,
and it could be useful to have a carefully pre-rendered view of the field
generated by the data provider.  The idea is that this would have the
same footprint and scale as the science data so that it would inherit
the same WCS.  One could overlay it with the science data, use it to
mark regions for analysis, draw graphics to represent the footprints of
subfields or objects for which additional data is available, and so forth.
Non-astronomical image viewers (xv etc.) would probably use such a preview
as the default view of the image, but unlike a JPEG one would still have
science-grade data with a full header.

The idea is to take your FITS image and generate this pre-rendered preview
which would be included somewhere in the file.  It would just be a binary
blob with some externally defined encoding such as JPEG, but with the
appropriate graphics library would expand to a 2D pixel array of RGB
(or whatever) encoded pixel values.  Including such a preview would be
optional, of course.  Presumably since it would be a lossy compressed
image, the cost in terms of space would be much less than the science data.

What we do currently in VO is, when the client queries for data, the
query response lists several renderings of the same field which the client
can retrieve.  One might be FITS, another one of the graphics formats.
The query response includes standard metadata describing the WCS etc.,
hence provides the information required to be able to use the graphics
rendering to determine positions, so that it can be used for more than
just a view of the field.  However since this doesn't travel along with
the rendered image it is somewhat harder to work with.

     - Doug