Apall extracts only 26 pixels
sirmarcos wrote on Aug 08, 2008
Well, this is a new one. I'm befuddled. apall is acting weird. I'm trying to use it to extract some echelle scectra. The apertures have already been found and traced, but I let apall edit them and review the tracings, everything looks fine.
But then when I'm reviewing the extract spectra, instead of extracting in X all the way from the first column to the 2048, it's showing me just columns 1-26. And, that's all it's extracting. The final image is [26,40], when it should be [2048,40] or something along those lines.
What's weird is that the spectra seem to become "infected" with this problem only after I try to run a certain automated script on them. I just tried apall on one spectra and it extracts fine, all 2048 pixels. But on another spectra, just 26 pixels, after I tried to use my script. But when I run apall directly on them, clearly, I'm using the same parameters and such.
The script/procedure just calls apall with specific parameters on the stellar and also some associated thorium spectra.
The 2D spectra display and imexam normally.
All I can think is that somehow the apertures are getting corrupted? Anyone ever seen anything like this?
But then when I'm reviewing the extract spectra, instead of extracting in X all the way from the first column to the 2048, it's showing me just columns 1-26. And, that's all it's extracting. The final image is [26,40], when it should be [2048,40] or something along those lines.
What's weird is that the spectra seem to become "infected" with this problem only after I try to run a certain automated script on them. I just tried apall on one spectra and it extracts fine, all 2048 pixels. But on another spectra, just 26 pixels, after I tried to use my script. But when I run apall directly on them, clearly, I'm using the same parameters and such.
The script/procedure just calls apall with specific parameters on the stellar and also some associated thorium spectra.
The 2D spectra display and imexam normally.
All I can think is that somehow the apertures are getting corrupted? Anyone ever seen anything like this?
Francisco Valdes wrote on Aug 08, 2008
This will take a little detective and guess work. Could it be that dispaxis is getting change to the wrong axis? I can't think of anything in the current apall that would limit the dispersion length during extraction. [Someday I 'd like to have a feature that would actually define an aperture into 2 dimensions to handle objective prism and slitlet data.]
So things to try (and send output to me if in doubt) is to look at the aperture defintion that results. Does anything show in the output header? You could use d_trace to turn on the debugging output. Since this is quite verbose and you are using a script the first step would be to but d_trace before and after the apall command. This will then output the actually apall command, with all parameters, actually being executed.
Good luck,
Frank Valdes
So things to try (and send output to me if in doubt) is to look at the aperture defintion that results. Does anything show in the output header? You could use d_trace to turn on the debugging output. Since this is quite verbose and you are using a script the first step would be to but d_trace before and after the apall command. This will then output the actually apall command, with all parameters, actually being executed.
Good luck,
Frank Valdes
sirmarcos wrote on Aug 08, 2008
Indeed, I examined the aperture files and there was something fishy going on. I deleted them, unlearned aptrace, tried them afresh using a working aperture reference image and they came back. but then my script screwed things up again.
I finally unlearned my script/procedure, deleted the bad apertures, and tried it again... that worked. I'm not sure if I was passing it a bad parameter that I didn't notice until I had to re-enter everything or what.
Anyway, peace in our time, thanks!
I finally unlearned my script/procedure, deleted the bad apertures, and tried it again... that worked. I'm not sure if I was passing it a bad parameter that I didn't notice until I had to re-enter everything or what.
Anyway, peace in our time, thanks!
Last post on Aug 08, 2008