PixInsight Forum

PixInsight => General => Off-topic => Topic started by: marekc on 2011 February 08 14:09:31

Title: S & T Interview that mentions Pixinsight (+ IRAF !)
Post by: marekc on 2011 February 08 14:09:31
Hi Everyone,

There's an article (on the Sky and Telescope website) about the recent `Hidden Treasures' competition held by ESO:


I couldn't help noticing that Joe DePasquale (the 4th-place finisher), who is mentioned in the article, says that he uses Pixinsight! Nice bit of publicity for Juan and Co.  :D   As a perpetual PI Newbie, this was one more bit of encouragement to keep `fighting the good fight' as I learn PI.

I found one of Mr. DePasquale's comments very interesting... he said that he calibrated the data with IRAF, and then used PI.

Yes, Pixinsight is very challenging, but IRAF seems to be a much scratchier `hair shirt' to put on! I once attempted to learn how to use it, but I haven't progressed very far at all. It's a UNIX-based software package that the professional astronomers use for their processing and analysis, and there's not much reason for amateurs to use it. I doubt that it does anything that `our' software (PI, PS, Maxim, AIP4WIN, etc) can't do, at least in terms of making asthetically pleasing images. My only reason for wanting to `know IRAF' was purely for the satisfaction of having mastered such a difficult beast... but I haven't gotten very far  :-[  I really don't know UNIX, and so it would be a great deal of work for me to learn IRAF, and it probably wouldn't `buy' me anything.

But... I would be very curious to know why Joe DePasquale used IRAF for his calibration. I wonder why he didn't use PI? It seems likely that the files he used were FITS files (though I don't know this), so it would seem to me that PI could `eat' pretty much anything that ESO would have in their archives.

I don't know if he reads the PI Forum, but if so, I'd be curious to learn more about his IRAF/Pixinsight workflow someday.

- Marek Cichanski
Title: Re: S & T Interview that mentions Pixinsight (+ IRAF !)
Post by: Carlos Milovic on 2011 February 08 16:07:57
It may be because there are some already done pipelines that calibrates raw data from the ESO telescopes. Also, it may be a bit challenging to do in PI if he is using multisensors.
Title: Re: S & T Interview that mentions Pixinsight (+ IRAF !)
Post by: Carlos Milovic on 2011 February 08 16:08:59
BTW, you may run IRAF from almost every Linux distro. ;)
Title: Re: S & T Interview that mentions Pixinsight (+ IRAF !)
Post by: Josh Lake on 2011 February 09 17:37:59
Wow, what a cool bit of synchronicity for me. I know Joe well, did undergrad astronomy work with him. He now works for the Harvard Smithsonian Center for Astrophysics, processing Chandra X-ray data, as seen here: http://chandra.harvard.edu/photo/chronological10.html

I believe most of those images are his work, and he's been transitioning from Photoshop to PI. He's the one who told me about PI, and now I don't use anything else after collecting data. It looks like PI does great work on instrument data from some of the most famous telescopes...

So I imagine that IRAF is their standard tool for making actual scientific measurements -- a command line tool that allows the scientists to very methodically and uniformly process instrument data that can be verified and replicated.

I emailed Joe and told him about this thread, so I hope he weighs in. If so, hey Joe!

Title: Re: S & T Interview that mentions Pixinsight (+ IRAF !)
Post by: Joe DePasquale on 2011 February 10 08:39:32
Hi all, and hey Josh!

I'm glad to see that article is getting noticed.  I'm a huge fan of PI!  I attended Vicent's workshop at Adler last year and found it hugely beneficial.  If any of you are attending the NEAIC/NEAF meeting this year, I'll be giving a talk about my workflow as an image processor with an emphasis on the importance of image metadata for getting your image out there.  And if you are going, be sure to attend Vicent's talk and his PI mini-workshop!

Regarding the S&T article and my use of IRAF: I actually tried to do the calibration work with PI but I ran into a problem with the specific dataset I was working with.  The data came from the ESO's Wide Field Imager (WFI) which is an array of 8 CCDs.  Each FITS file generated from that detector is actually a muti-extension FITS file (MEF) containing data from all 8 CCDs in one file.  When I tried to generate the master flat and bias files in PI, I found that PI cannot process MEF files (I did some forum searches to confirm this as well).  I was faced with either breaking down the extensions and working with individual files for each CCD (which would've been a huge mess) or learning IRAF, which I had been meaning to do anyway. 

So, I did a lot of research on using IRAF with WFI data and actually found some helpful documentation and pipeline tools and even an IRAF "plugin" specifically developed for WFI data and did the reduction in IRAF.  I would hardly say that I mastered IRAF (or PI for that matter), but I managed to beat it into submission long enough for me to generate some pretty pictures.  This definitely slowed down my progress significantly and I actually missed out on the M78 data (the winning entry). I had worked on that dataset a lot and then found it had already been posted!  However, it was totally worth it in the end to have spent the time dealing with IRAF.  I learned a lot about image calibration doing it this way, and even if I never have to use it again, those skills are transferable. 

Title: Re: S & T Interview that mentions Pixinsight (+ IRAF !)
Post by: mmirot on 2011 February 10 12:08:46
Hi Joe,

Good to hear from you again.

I think Juan is working on the update of the Calibration module now.
 Right now calibration works great but is not as friendly as some programs

I know eventually we will be able to to do multiple CCDs with overscan and everything in a pipeline mode.

I have not had a chance to do much imaging since we met at Adler.

Weather and equipment problems...

I tried to download a few hubble data sets recently.
It it hard to figure out what you need.
The couple of objects I downloaded had only one exposure per filter.
I did not know how to fix the cosmic rays without a stack.
Finding out how to use a professional data base takes quite a bit of self education. It is not as fun as learning PI.

Max Mirot