Dear Han,

The limit to the size of the region you retrieve is a characteristic of the size of projection used. The default Tangent projection will show a maximum of one hemisphere of the sky, and only reaches that as the image size approaches infinity. Other projections can be used to see more of the sky in a single image. When you specify the size of the image in degrees (with SFACTR above), you are specifying the size in the projected coordinates, not on the sphere. For small maps where distortions are minimal these will be the same, but as you try to get very large images substantial distortions are inevitable. Other projections,
Car, Zea, Ait may be more appropriate for very large images.

With regard to use of the Mellinger data: I am not a lawyer, but I’m dubious that the license we have from Dr Mellinger permits the kind of use you envisage. The Copyright notice indicates:

Copyright: Axel Mellinger. Permission is granted for use in research and personal, non-commercial use. Please contact Axel Mellinger for permission for other use.

This copyright still applies even if you get the data through us.

Regards,
Tom McGlynn

The advantage of 8 bit or 16 bit FITS will be the smaller file size and therefore increased download speed.

I have noticed that the field size of the downloaded FITS files is limited to about 142 degrees. Not a problem, just an observation. This is the web command line:
http://skyview.gsfc.nasa.gov/cgi-bin/pskcall?return=fits&sampler=Lanczos&Survey=MELL-R&VCOORD=0,0&PIXELX=1000&PIXELY=1000&SFACTR=300

Thanks again,
Han

Thanks for the comments…

While the input data may have byte values, resampling can cause the computed pixels to be non-integers, so even with
survey that start with integers, floating point outputs0 are possible. You are correct that with the default nearest neighbor sampling the user will only get integer values, but SkyView does all it’s computations in double precision. We could think about an option where the user could scaling to integers.

Your idea that we put the actual range of the data in the FITS file makes a lot of sense. We already provide that in the web page output, so we’re already computing it. We’ll put that in the next revision thanks.

You may wish to check with Alex Mellinger regarding use of his survey in your program. His data is not in the public domain.

Tom

SkyView’s hardware was not directly affected by the recent problems. Some related systems had issues, including the server which holds the database used by our blog.
Regards,
Tom

Thanks in advance, also looking forward to more medium resolution samples