Pixinsight Image Calibration

[Buzzard75]

I've recently been having some issues with image calibration in Pixinsight. A few issues actually.

Darks Issues:

1. I'm using darks that are exactly the same exposure length, gain, offset, and temperature as my lights. I have never had to optimize my darks before and my understanding is I shouldn't have to as long as I'm using darks with exactly the same properties as my lights. However, if I don't select optimize when calibrating my lights, then my images come out completely black with maybe a couple bright stars even after a stretch. When trying to stack in DSS, it throws up errors during registration that it can't find enough star pairs after the calibrations have been completed. This suggests there has to be something wrong with the FITS files, but I don't know what it could possibly be.

2. Some of my light frames will come out black after calibration with my dark master regardless of whether I select optimize or not. I can't see anything in the header information that would suggest why it works with some lights and not others even with optimize selected as stated above.

3. My darks are not calibrating out the hot pixels in my lights at all. I am having to do Cosmetic Correction to remove them. I have never bothered with CC before because the calibration with darks usually did a sufficient job for my level of laziness. 😁

The darks I was using are only a month old and worked for three or four of my most recent images without issue. I don't know why they have not worked for the last two images I have taken. I am currently retaking dark frames to see if new darks will correct the issue or if it persists. Any other insight would be appreciated.

Flats Issues:

1. My flats don't calibrate out my dust donuts at all. They are present in my light images and I can see them in my flats, but they are still present in my lights after calibration and in some cases more exaggerated. My flats are 2-seconds long so should be sufficient length. I've used flats that were much shorter and some that were longer with no issues before. I'm wondering if the dark optimization of my dark flat frames is playing a role here as well, but my flats don't seem to suffer the same issue as my lights do when calibrating.

2. EDIT Forgot to mention the gradient. When calibrating with my flats, my images come out with a terrible gradient. The images end up darker in the center and significantly brighter towards the edges. This normally wouldn't be a problem because there is typically some vignetting in images, but mine are fairly flat across the field to begin with. Again, it's like Pixinsight is overcompensating in this regard.

And for what it's worth, I'm currently using my QHY247C camera. Thanks in advance!

 

Edited November 22, 2020 by Buzzard75

[dodgerroger]

I can’t really comment on suggestions to the above as I am a mere beginner with pixinsight but have you tried stacking your images in a different program, APP or DSS. See if you get the same results. This will narrow it down to your issue 

[Laurin Dave]

Maybe experiment with adding a Pedestal in Pixinsight when you calibrate to avoid the black clipping.. and see if that helps ..   

[Buzzard75]

1 hour ago, dodgerroger said:

I can’t really comment on suggestions to the above as I am a mere beginner with pixinsight but have you tried stacking your images in a different program, APP or DSS. See if you get the same results. This will narrow it down to your issue 

See #1 under Darks. Yes, I've tried DSS and the result is the same or at least similar. When I try and stack in DSS it can't integrate the light frames because there aren't enough star pairs because the images are basically all black. That to me suggests it's a file issue and not a program issue.

1 hour ago, Laurin Dave said:

Maybe experiment with adding a Pedestal in Pixinsight when you calibrate to avoid the black clipping.. and see if that helps ..   

I'm not familiar with Pedestal, but I'll look into it. Thanks!

[Laurin Dave]

Another possibility would be some kind of light leak affecting your darks ..  maybe compare current ones with historic 

[wimvb]

And don’t forget to just look at the darks. At what level are the mean and median? What about the mean and median of your lights?

[Buzzard75]

2 hours ago, Laurin Dave said:

Another possibility would be some kind of light leak affecting your darks ..  maybe compare current ones with historic 

Darks look fairly similar to the previous ones taken about a month ago. See the notes below about the mean and median values. It may explain why they are slightly different.

1 hour ago, wimvb said:

And don’t forget to just look at the darks. At what level are the mean and median? What about the mean and median of your lights?

Mean and median of my darks is around 437 for my last set. Mean and median of my lights is around 334. Ruroh!!😖

For fun (and comparison) I looked at the mean/median of my darks from a session a month ago and they are around 873. The lights were around 985.

Based on those two sets of values, it would seem something is amiss with my last set of data. Actually, my last two sets of data as my lights for the previous set had similar values. Of note would be that I used a mini PC to capture my last two sets of data whereas all my previous sets of data used my laptop. I am running different versions of APT on both as I had issues with the most recent version on my mini PC. Short of exposing longer I'm not sure how to increase my mean and median. Unfortunately, with my poor guiding at the moment, I don't know that that is possible. Obviously increasing gain isn't an option as that would just increase my gain required for my darks.

However, it make sense now why optimizing darks is the only thing that seems to work.

 

 

[wimvb]

For cmos astro cameras, you not only set gain, but also offset. Offset is to make sure that you won't get any negative/clipped pixel values. You should probably check if offset has changed. You can add offset before image calibration, at least in Pixinsight. A change in offset can also be the reason why your flats aren't working.

[Laurin Dave]

I’d look at your files and check the fits header to see if gain and offset have changed 

[Laurin Dave]

Just now, wimvb said:

For cmos astro cameras, you not only set gain, but also offset. Offset is to make sure that you won't get any negative/clipped pixel values. You should probably check if offset has changed. You can add offset before image calibration, at least in Pixinsight. A change in offset can also be the reason why your flats aren't working.

Snap! 

[Buzzard75]

Take a look. As best I can tell everything is identical as far as the imaging properties go. Gain, temp, exposure. I don't see anything for offset specifically beside BZERO.

 

Lights Header:

Image #1
    SIMPLE  =                    T / file does conform to FITS standard
    BITPIX  =                   16 / number of bits per data pixel
    NAXIS   =                    2 / number of data axes
    NAXIS1  =                 6088 / length of data axis 1
    NAXIS2  =                 4052 / length of data axis 2
    EXTEND  =                    T / FITS dataset may contain extensions
    COMMENT   FITS (Flexible Image Transport System) format is defined in 'Astronomy
    COMMENT   and Astrophysics', volume 376, page 359; bibcode: 2001A&A...376..359H
    BZERO   =                32768 / offset data range to that of unsigned short
    BSCALE  =                    1 / default scaling factor
    OBJECT  = 'NGC7380 '           / The name of Object Imaged
    TELESCOP= 'iOptron ASCOM Driver for Mount' / The Telescope used
    INSTRUME= 'QHY247C-e5af9d4d873698c45' / The model Camera used
    DATE-OBS= '2020-11-22T01:22:42' / The UTC date and time at the start of the expo
    HIERARCH CAMERA-DATE-OBS = '2020-11-22T01:22:42' / The UTC date and time at the
    EXPTIME =                 180. / The total exposure time in seconds
    CCD-TEMP=                 -10. / Temperature of CCD when exposure taken
    XPIXSZ  =                 3.91 / Pixel width in microns (after binning)
    YPIXSZ  =                 3.91 / Pixel height in microns (after binning)
    XBINNING=                    1 / Binning factor in width
    YBINNING=                    1 / Binning factor in height
    XORGSUBF=                    0 / Sub frame X position
    YORGSUBF=                    0 / Sub frame Y position
    EGAIN   =                   1. / Electronic gain in e-/ADU
    FOCALLEN=                 1624 / Focal Length of the Telescope in mm
    JD      =     2459175.55743056 / Julian Date
    SWCREATE= 'Astro Photography Tool - APT v.3.84' / Imaging software
    SBSTDVER= 'SBFITSEXT Version 1.0' / Standard version
    SNAPSHOT=                    1 / Number of images combined
    SET-TEMP=                 -10. / The setpoint of the cooling in C
    IMAGETYP= 'Light Frame'        / The type of image
    OBJCTRA = '22 46 51'           / The Right Ascension of the center of the image
    OBJCTDEC= '+58 03 10'          / The Declination of the center of the image
    SITELAT =       / The site Latitude
    SITELONG=       / The site Longitude
    AMB-TEMP=                 10.7 / The Ambient Temperature
    FOCUSPOS=                 8250 /  Focuser position in steps
    GAIN    =                 2200 / The gain set (if supported)
    BAYERPAT= 'RGGB    '           / The Bayer color pattern

Darks Header:

Image #1
    SIMPLE  =                    T / file does conform to FITS standard
    BITPIX  =                   16 / number of bits per data pixel
    NAXIS   =                    2 / number of data axes
    NAXIS1  =                 6088 / length of data axis 1
    NAXIS2  =                 4052 / length of data axis 2
    EXTEND  =                    T / FITS dataset may contain extensions
    COMMENT   FITS (Flexible Image Transport System) format is defined in 'Astronomy
    COMMENT   and Astrophysics', volume 376, page 359; bibcode: 2001A&A...376..359H
    BZERO   =                32768 / offset data range to that of unsigned short
    BSCALE  =                    1 / default scaling factor
    INSTRUME= 'QHY247C-e5af9d4d873698c45' / The model Camera used
    DATE-OBS= '2020-11-22T16:28:11' / The UTC date and time at the start of the expo
    HIERARCH CAMERA-DATE-OBS = '2020-11-22T16:28:11' / The UTC date and time at the
    EXPTIME =                 180. / The total exposure time in seconds
    CCD-TEMP=                 -10. / Temperature of CCD when exposure taken
    XPIXSZ  =                 3.91 / Pixel width in microns (after binning)
    YPIXSZ  =                 3.91 / Pixel height in microns (after binning)
    XBINNING=                    1 / Binning factor in width
    YBINNING=                    1 / Binning factor in height
    XORGSUBF=                    0 / Sub frame X position
    YORGSUBF=                    0 / Sub frame Y position
    EGAIN   =                   1. / Electronic gain in e-/ADU
    FOCALLEN=                 1624 / Focal Length of the Telescope in mm
    JD      =     2459176.18623843 / Julian Date
    SWCREATE= 'Astro Photography Tool - APT v.3.84' / Imaging software
    SBSTDVER= 'SBFITSEXT Version 1.0' / Standard version
    SNAPSHOT=                    1 / Number of images combined
    SET-TEMP=                 -10. / The setpoint of the cooling in C
    IMAGETYP= 'Dark Frame'         / The type of image
    SITELAT =       / The site Latitude
    SITELONG=       / The site Longitude
    GAIN    =                 2200 / The gain set (if supported)
    BAYERPAT= 'RGGB    '           / The Bayer color pattern

Edited November 23, 2020 by Buzzard75

[wimvb]

You may need to experiment, it seems. Similar to this

 

Edited November 23, 2020 by wimvb

[Buzzard75]

16 minutes ago, wimvb said:

You may need to experiment, it seems. Similar to this

 

Thanks! Will definitely check it out.

[Buzzard75]

@wimvb, @Laurin Dave, @dodgerroger, Appreciate all the input. I haven't figured it out yet. I took new darks, new flats, and new dark flats the day after I imaged so them being old wasn't really an issue. I have used Pixel Math to fix the symptom rather than the cause per that video series, or so I thought, but the end result didn't really change all that much. I can get a decent individual sub calibrated, but when they're all stacked it just isn't right. I still think there's something wrong with my dark frames which is making it worse. I'll just have to keep at it.