Help!!! : bias reduction from overscan region line by line
Mayuresh Arvind Kulkarni wrote on Mar 13, 2008
Hello all iraf masters.
I am quite new to this forum and a student who is learning iraf for my data which is standard .fits format.
I am badly in the need of help. I have looking for following task to do.
In my data there is overscan region in left and right side for 50 Pix. I want to calculate median value for each line and reduce that from the whole line in the image.
e.g. for 1st line left side, I want to find median from left most 50 pix.(overscan pix) and reduce that value from all the pixels in the first row of the left half. same for right and for all rows.
Could someone please help me for this? I tried "Colbias" in the bias task but it seems it is doing something else which I couldn't understand.
Can I use "ccdproc"?
Please help.
Thanks in advance
-a student
mak
I am quite new to this forum and a student who is learning iraf for my data which is standard .fits format.
I am badly in the need of help. I have looking for following task to do.
In my data there is overscan region in left and right side for 50 Pix. I want to calculate median value for each line and reduce that from the whole line in the image.
e.g. for 1st line left side, I want to find median from left most 50 pix.(overscan pix) and reduce that value from all the pixels in the first row of the left half. same for right and for all rows.
Could someone please help me for this? I tried "Colbias" in the bias task but it seems it is doing something else which I couldn't understand.
Can I use "ccdproc"?
Please help.
Thanks in advance
-a student
mak
Cold wrote on Mar 13, 2008
hi kulmak,
there are some ccd reduction exercises on iraf's home page. find them (i think in the exercises link) and download the exercise 2103.tar.Z and its pdf file. It's about what you need. Install it into your hard drive with the command
zcat exer2103.tar.Z | tar -xf-
install it to the directory where login.cl exits. (for easy access) .
Nowadays i'm doing some reduction exercises too. But my frames don't need to be trimmed so i didn't use any kind of task that you need. And ccdproc task is for subracting dark,bias and dividing flats as i know...
your sincerly
there are some ccd reduction exercises on iraf's home page. find them (i think in the exercises link) and download the exercise 2103.tar.Z and its pdf file. It's about what you need. Install it into your hard drive with the command
zcat exer2103.tar.Z | tar -xf-
install it to the directory where login.cl exits. (for easy access) .
Nowadays i'm doing some reduction exercises too. But my frames don't need to be trimmed so i didn't use any kind of task that you need. And ccdproc task is for subracting dark,bias and dividing flats as i know...
your sincerly
Francisco Valdes wrote on Mar 13, 2008
Hello,
Generally one uses only the overscan. You can subtract the median of the overscan for each line using the value "median" for the ccdproc parameter "function". You also need to specify the overscan region by the biassec keyword in IRAF image section notation.
Yours,
Frank Valdes
Generally one uses only the overscan. You can subtract the median of the overscan for each line using the value "median" for the ccdproc parameter "function". You also need to specify the overscan region by the biassec keyword in IRAF image section notation.
Yours,
Frank Valdes
Jason Quinn wrote on Mar 13, 2008
valdes
Generally one uses only the overscan. You can subtract the median of the overscan for each line using the value "median" for the ccdproc parameter "function". You also need to specify the overscan region by the biassec keyword in IRAF image section notation.
It seems that ccdproc's overscan correction code is completely separate from colbias and linebias code although does the same thing (plus some more). What are the purposes of the colbias and linebias tasks? It doesn't seem possible to get the non-fitted overscan subtraction using those two tasks as the user wanted but certainly these standalone tasks should be able to do that, I think. It seems strange to me that the embedded overscan correction in ccdproc is more powerful than the standalone tasks.
I took a look at the code for colbias and linebias and there is a lot of duplication there. Seems like an excellent spot for code generalization by having each task call a generalized routine. Moreover, since ccdproc has extra functionality, why not combine (or replace) col- and linebias' overscan correcting code with the ccdproc overscan code to reduce redundancy, ease maintainability, and minimize user complexity? It seems the logically proper thing for ccdproc to call the more specialized task rather than dancing to its own tune.
Hopefully, I haven't mis-understood the design here. I'm a decent IRAF scripter but not an IRAF programmer. I'm tempted to try to tackling merging them myself as it seems not all that difficult and would be a good place to learn IRAF programming; but, unfortunately, I'm too busy with my new postdoc right now (at some point I'll get around to spellchecking the NOAO half of IRAF's help files, fitz).
I'm guessing that the reason line- and colbias exist is because they were coded first, and then ccdproc consumed them, leaving them around as only historical legacy. But my main point is that, ideally, ccdproc, linebias, and colbias should all operate via a unified code base. Am I wrong here?
Jason
Francisco Valdes wrote on Mar 13, 2008
Hi Jason,
You are guessing correctly. And you are correct about the ideal but lack of time and priority prevents this. In general we are conservative about removing or changing tasks which people may have used or be using in reductions. So one could also argue that leaving them unchanged is a proper thing to do. Thanks for the comments.
Yours,
Frank
I'm guessing that the reason line- and colbias exist is because they were coded first, and then ccdproc consumed them, leaving them around as only historical legacy. But my main point is that, ideally, ccdproc, linebias, and colbias should all operate via a unified code base. Am I wrong here?
You are guessing correctly. And you are correct about the ideal but lack of time and priority prevents this. In general we are conservative about removing or changing tasks which people may have used or be using in reductions. So one could also argue that leaving them unchanged is a proper thing to do. Thanks for the comments.
Yours,
Frank
Last post on Mar 13, 2008