[AstroPy] Coding in Python to Read/Write a fits file.

Andrew Ptak ptak at pha.jhu.edu
Wed Apr 8 11:30:23 EDT 2009

I wouldn't trust any science analysis done on an arbitrary image that 
isn't in FITS (or some other scientific format).   You need to be sure 
that flux, noise, etc. have been properly preserved and calibrated for 
any believable analysis.  First of all, jpeg is itself a lossy 
compression format, but in general you don't know what smoothing, 
filtering, etc. might have been done to the original data to make a 
press release image.  You might be able to get away an order-of-mag. 
estimate of some quantities if the amount of processing had been minimal 
and you are very familiar with the type of data that went into the image.

Andy Ptak

Megan Sosey wrote:
> Hi Wayne,
> One caveat I forgot to mention, the script
> I sent you will transform the JPG (or gif or whatever)
> image into FITS images, which you can then look
> at in ds9 if you like, however, the data numbers are
> not something appropriate to do further analysis with,
> they are just the RGB scalings. So if you want to take the
> FITS images and do more science analysis on them, you'll
> have to rescale the data to something appropriate.
> cheers
> meg
> Megan Sosey wrote:
>> Hi Wayne,
>> something like this work for me, perhaps it will help
>> get you closer to what you want. I attached the sample
>> image I used.
>> cheers,
>> Megan

More information about the AstroPy mailing list