How does CUBISM render slit sizes?
Posted by IRSA HelpDesk on 11 March 2013 02:15 PM
The aperture shown in CUBISM is data-driven, and does not intend to represent the full slit. The width in the CUBISM visualization reflects that of the 1 pixel width wavesamp, which is set that way on purpose, reflecting the fact that in the cross-slit direction, you have a convolution of spectra and spatial information. It also uses the actual configured length of wavesamp pseudorectangles (not just some fiducial slit length number). So it's on the conservative side, whereas Spot's visualization is "generous" with the slit dimensions.

If you fail to step your maps by 1/2 physical slit width, you can have gaps open up as a result. This is fairly evident in the visualizations. So when someone sees a 2 pixel wide slit, and says about the CUBISM vizualization "but it's lying, I clearly got information from this area to the side", the answer is "yes and no". You got that spatial information, but because of the undersampling, it was hopelessly convolved with the spectral information. So did you really get it? CUBISM shows you the region where you *really* got spatial information. Another answer is: if you are worrying about your slit position at the half-pixel level, you've got bigger problems.
(0 vote(s))
This article was helpful
This article was not helpful

Comments (0)