Confused by differences in orientation between different software

[Alan Burgess]

Hi, just getting into comets and decided to process latest HST 3I/ATLAS data. Stacked in PI and applied Larson Sekanina filter with Astroart which seems to work well. The only issue that I encountered was that the orientation in Astroart seems flipped vertically compared to Toni Scarmato’s published version. Mine is the top and his is the bottom. Can anybody shed any light on which one is correct please?

Attachments:

1. 

   31comparison-1.png
   

[Nick James]

It is very difficult to interpret these images!

Part of the problem is that the original FITS image standard was not clear on the row order. Professional tools tend to assume that the origin is in the bottom left corner so rows are numbered upwards from the bottom. Many amateur tools assume that the origin is in the top left corner so that rows are numbered downwards from the top. If you open the file with a tool which assumes the opposite convention you get a mirror image. There is now a FITS keyword (ROWORDER) which signifies the order to use but many tools don’t recognise that. I would think that the HST FITS files have their origin at the bottom left. I haven’t used PI or Astroart and I don’t know which convention they follow.

At present the prominent tail of 3I is in a PA of 100 degrees, i.e. just south of east. You can see some recent images here:

https://britastro.org/cometobs/3i/thumbnails.html

Could you post the greyscale stack that you get prior to doing the LS filter please along with some scale information? We could then compare with other images to get some idea of the orientation. Also, let me know the parameters you have used for the LS filter. You need to be very careful with that since it can easily generate radial structures that are not real. As a rule of thumb you should keep the rotation to 15 degrees or less.

[Author]

Posts