PixInsight Forum

PixInsight => General => Topic started by: arunasmazeika on 2015 June 25 00:58:23

Title: Problem calibrating CR2 files
Post by: arunasmazeika on 2015 June 25 00:58:23
Hi everyone!,

I come to you since I'm trying to solve an issue I came across yesterday. I cannot figure it out.

Normally I use Nebulosity for capturing (FIT lights), but two days ago I went out to try an IOptron SkyTracker and made some quick shots using the DSLR directly (CR2 lights).

The thing is that when calibrating these CR2 files using my master BIAS and DARK (no flats here), the resulting calibrated images are very dark and the histogram shows a peak with an abrupt cutoff like at 1/3 of the total range. There's obviously a severe loss of signal.

I've tried converting these CR2 files to FITs (16 and 32 bits) with no avail, same results. I have no problems calibrating when using my FITs from Nebulosity with the same master BIAS and DARK. There must be something wrong with my CR2 lights.

My settings in the DSLR_RAW format section seem good ("Create raw Bayer CFA image" checked). I'm attaching one of the CR2 images as well as my master BIAS and DARK:


Any ideas?.

Thank you for your help! :).
Title: Re: Problem calibrating CR2 files
Post by: oldwexi on 2015 June 25 06:22:25
Hi arunasmazeika!
Load your 3 images into PI and watch the histograms and the statistics.
The histogramm will clearly show your CR2 (from June last year) is DARKER  than your MasterBias and
your MasterDark. So if you subtract bias and dark there is nothing left to show!! Except black!!

In addition if you look at the statistics of your images the median and mean value
of the bias is higher than the values of the dark!!!

So, These Masters are simply wrong!

Usually light is the brightest than comes the dark and than the bias as darkest.
In your Images it is exact the opposite. The shortest exposure Bias is MUCH brighter than the 3 Minute CR2.

So try to get the Bias and Darks correctly.

Title: Re: Problem calibrating CR2 files
Post by: arunasmazeika on 2015 June 25 07:38:31
Hi Gerald,

Thank you for your reply, really appreciate it :).

Good catch on the histogram ... just noticed it, I don't know why I didn't take the time to compare those ... probably because Nebulosity calibrates them alright, which is odd.

So it makes sense what you say about subtracting something that's lighter to something darker, the result I get is exactly consistent with your conclusion. I've used a UHC filter (also a first timer), which would explain why the lights are so dark in comparison with my previous shots. However, I have now clue whatsoever why these lights would be darker than the master bias and dark. This doesn't make sense. I'll dig a little bit more, but this is the main reason why I'm getting these results.

Regarding the masters, I never thought about this, but how could they be wrong?. I've just performed an average integration using the raw bias and darks and I still get the same results. What you say makes sense though, darks should be lighter than bias since we get more signal out of them due to thermal noise, and these also contain the bias. The mean should be higher in the master dark. I guess I'll try generating those again and see what it comes out of it, but I'm really surprised. Biases and darks aren't that hard to generate, so I really don't know what I could be doing wrong here.

Again, thank you for your input, this is all very helpful.

Title: Re: Problem calibrating CR2 files
Post by: msmythers on 2015 June 25 09:08:55

Are the master bias and dark files presented calibrated also in Nebulosity? If you have the cr2 files for the dark and the bias I would try to make the masters in PixInsight and see what happens. I don't understand how your bias and dark mean values are so high. My guess is there is a scaling issue between the fit files from Nebulosity and PI. That is why I ask about the raw cr2 dark and bias, to avoid the Nebulosity generated fit files.


Title: Re: Problem calibrating CR2 files
Post by: arunasmazeika on 2015 June 25 11:57:51
Hello Mike,

The darks and biases are FIT files generated by Nebulosity (I've captured them using it). Maybe those FITs are scaled somehow?. I'll dig in. This is all good info. I'm rather new to astrophotography so I've never had the chance to compare master biases and darks to actually realize that they are too bright :).

Based on all this info I'm gonna be doing some testing on both PI and Nebulosity. I'll also make a new set of biases and darks using the CR2 from the camera (Canon EOS 450D) and generate the masters using PI. Then I can compare with those I already have.

As you say, the problem may lie in the masters being scaled somehow and not necessarily in the CR2 files that I got from the camera. The other lights I've previously used are also FIT files generated by Nebulosity, so this all holds up.

Thank you Mike :).
Title: Re: Problem calibrating CR2 files
Post by: arunasmazeika on 2015 June 27 04:36:43
Hello guys,

I've generated a new set of darks and biases during the night using the camera directly. I've just created the master calibration files and bingo!. Their intensity are good now and my lights are perfectly calibrated.

It would seem that nebulosity FITs are indeed scaled or something.

Anyway, thank you so much for your help!, you were both spot on! :).