Hello,
This is my first post to AAVSO - hopefully its posted to the correct place and that these questions haven’t been raised before but apologies if that is not the case.
I am able to upload to VPhot “raw” FITS files that have come straight from my ASI294MM camera or where calibration has been performed using AstroImageJ.
However for various reasons, I am trying to upload to VPhot FITS files that has been calibrated using Astropy image reduction libraries and clipped using the cutout2D method to reduce the image size to manageable proportions (for example, I can crop a 4144x2822 down to 516x366 (in this case) which gives a reduction in size from 22Mb → 776kB.
When I attempt to upload the file into VPhot using the upload wizard, the FITS header inspection passes without issues but when the file is uploaded for ingestion, nothing appears in the processing queue nor ends up in my images folder.
I noticed that the Astropy data reduction routines do two things to the images: firstly, 16 bit image data is converted to 32bit floating point and secondly, two additional HDUs are injected into the FITS file to store mask and error data. In effect, a 22Mb file ends up being 196Mb. I can delete the additional HDUs and also recast the floating point data to signed INT16. However in doing so, the BZERO and BSCALE parameters are created with values other than 32768 and 1 (but not sure if this is an issue or not).
Nevertheless, the resulting file can be loaded and displayed in the Astropy code as well as in AstroImageJ, ASTAP and PixInsight without any warning or error which leads me to ask the following:
- does VPhot support uploads with 32bit floating point data?
- does VPhot support signed INT16 data where the BSCALE and BZERO values are anything other than 1 and 32768?
- does Vphot place any constraints on the image dimensions (NAXIS1, NAXIS2)?
- is there anything that stands out in the provided FITS header that seems wrong?
I would very much appreciate any help with this.
Regards to all
Tony
fitshdr.txt (4.1 KB)