Jump to content

  •  

CNers have asked about a donation box for Cloudy Nights over the years, so here you go. Donation is not required by any means, so please enjoy your stay.

Photo

DSS issues (?) first attempt with new camera

  • Please log in to reply
32 replies to this topic

#1 Muskoka

Muskoka

    Vostok 1

  • -----
  • topic starter
  • Posts: 103
  • Joined: 08 Apr 2020
  • Loc: Muskoka Ontario

Posted 15 May 2021 - 09:44 AM

First attempt at using my new ASI533MC Pro. Started astrophotography last fall, never used anything other than my Canon T3i. I've used DSS on numerous occasions, and never really had any issues. This first attempt shooting with the 533 doesn't appear to have worked too well.

 

As usual, used Nina (native zwo driver) for image capture. Used W/O Z61, Z61A flattener, guided, Eqm-35 Pro. Took 99 lights, 60 seconds, gain 101, offset 70, temp set to -5c, dither every other frame. Also took 20 darks, 60 seconds, gain 101, offset 70, temp set to -5c. There was a Meridian flip about half way through the session. My skies here at home are Bortle 3.

 

In the past I've never worried about removing images with comet streaks, and DSS has always taken care of them, not this time? There's also a lot of other issues going on with the stacked image, and I have no clue what's causing the issues.

 

Here's a image straight from DSS into Photoshop, with a light stretch applied. What is going on in this image? Never got results like this from the dslr, so I'm sure it must be some settings I don't have correct, but not sure what. Any help, insight, suggestions would be greatly appreciated.

 

M101 DSS1.jpg


Edited by Muskoka, 15 May 2021 - 12:58 PM.


#2 Tapio

Tapio

    Skylab

  • -----
  • Posts: 4,303
  • Joined: 24 Sep 2006
  • Loc: Tampere, Finland

Posted 15 May 2021 - 10:21 AM

Guess you mean satellite streaks.
Change stacking algorithm to kappa sigma.

#3 Muskoka

Muskoka

    Vostok 1

  • -----
  • topic starter
  • Posts: 103
  • Joined: 08 Apr 2020
  • Loc: Muskoka Ontario

Posted 15 May 2021 - 11:36 AM

Thanks for the response. That is using Kappa Sigma, I tried Kappa-Sigma and Median Kappa-Sigma, neither of them removed the sat/comet streaks. It always did with the dslr, but not after this first attempt with the 533. 

 

I'm also curious as to what the red, blue, and green blobs/blotches are scattered about in the image. I don't see them in the light subs, and never saw them with the dslr. I'm wondering if it's all hot pixels, which I do see a lot of, which I wouldn't expect to see at 101 gain for 60 seconds.

 

blotch.jpg

 

 


Edited by Muskoka, 15 May 2021 - 11:42 AM.


#4 Muskoka

Muskoka

    Vostok 1

  • -----
  • topic starter
  • Posts: 103
  • Joined: 08 Apr 2020
  • Loc: Muskoka Ontario

Posted 15 May 2021 - 12:06 PM

Here's what I see in Startools with the autodev stretch, straight after stacking in DSS.

 

STARTOOLS2.jpg

 

 



#5 RedLionNJ

RedLionNJ

    Skylab

  • *****
  • Posts: 4,321
  • Joined: 29 Dec 2009
  • Loc: Red Lion, NJ, USA

Posted 15 May 2021 - 12:41 PM

Can you also show a single (random) sub, processed the same way?  And how about a representative dark?



#6 Muskoka

Muskoka

    Vostok 1

  • -----
  • topic starter
  • Posts: 103
  • Joined: 08 Apr 2020
  • Loc: Muskoka Ontario

Posted 15 May 2021 - 12:53 PM

Hi, here's the first light sub autodev stretched in Startools, straight from the camera.

 

single sub2.jpg

 

Here's a link to the first light sub from above, untouched.

 

https://drive.google...iew?usp=sharing

 

Here's a link to the first dark sub.

 

https://drive.google...iew?usp=sharing


Edited by Muskoka, 15 May 2021 - 01:02 PM.


#7 Mike in Rancho

Mike in Rancho

    Viking 1

  • -----
  • Posts: 882
  • Joined: 15 Oct 2020
  • Loc: Alta Loma, CA

Posted 15 May 2021 - 01:53 PM

Well for one, you have a debayering problem.  The oddball colors were the first clue, and the grid-like pattern you see on your last ST image was the second.

 

In DSS, you'll need to change your FITS settings for your camera.  It's the same window, one side is RAW (which you probably had set correctly for your DSLR), and the other tab is FITS.  At minimum, you will need to alter the channel pattern to match what your camera is putting out.  It's probably on RGGB.

 

The satellite trails should disappear by tightening the kappa down.  Do you have it at something like 3.0 to let more in?  Try 2.5 or 2.0.  Or, if you don't want to screen out potential detail, since you have Startools, healing satellite trails is a fairly straightforward healing process at the end.  It just gets a little tricky if the trial bisects stars - but that's just some mask alteration and flipping back and forth to see if it worked.

 

The other stuff - I don't know astrocams so others will have to chime in.  I would guess hot pixels too, based on the color and the pattern shape they all have.  Maybe something with doing your darks the right way, or can otherwise clean up in DSS.  There are hot/cold pixel settings for some frames, as well as a global hot/cold pixel remover under the cosmetic tab.

 

Let us know what you find!


  • Muskoka likes this

#8 Muskoka

Muskoka

    Vostok 1

  • -----
  • topic starter
  • Posts: 103
  • Joined: 08 Apr 2020
  • Loc: Muskoka Ontario

Posted 15 May 2021 - 02:34 PM

Well for one, you have a debayering problem.  The oddball colors were the first clue, and the grid-like pattern you see on your last ST image was the second.

 

In DSS, you'll need to change your FITS settings for your camera.  It's the same window, one side is RAW (which you probably had set correctly for your DSLR), and the other tab is FITS.  At minimum, you will need to alter the channel pattern to match what your camera is putting out.  It's probably on RGGB.

 

The satellite trails should disappear by tightening the kappa down.  Do you have it at something like 3.0 to let more in?  Try 2.5 or 2.0.  Or, if you don't want to screen out potential detail, since you have Startools, healing satellite trails is a fairly straightforward healing process at the end.  It just gets a little tricky if the trial bisects stars - but that's just some mask alteration and flipping back and forth to see if it worked.

 

The other stuff - I don't know astrocams so others will have to chime in.  I would guess hot pixels too, based on the color and the pattern shape they all have.  Maybe something with doing your darks the right way, or can otherwise clean up in DSS.  There are hot/cold pixel settings for some frames, as well as a global hot/cold pixel remover under the cosmetic tab.

 

Let us know what you find!

Thank you. Here's some shots of my DDS settings. Alignment tab, Automatic selected. Intermediate Files, Fits selected. Cosmetic, nothing selected. 

 

dss settings.jpg dss2.jpg

dss3.jpg dss4.jpg



#9 Mike in Rancho

Mike in Rancho

    Viking 1

  • -----
  • Posts: 882
  • Joined: 15 Oct 2020
  • Loc: Alta Loma, CA

Posted 15 May 2021 - 03:00 PM

On further thought - I'm probably wrong!

 

I should have read closer - your last image into Startools was straight from the camera you say, so it had no debayering done?  Well oops, that explains the grid pattern.  Startools doesn't debayer the files.

 

So it is probably RGGB, with the only other reasonable choice being BGGR if the red and blue are swapped.  There's nothing obvious in the image (like a horsehead) for me to get a clue on it.  The blue satellite trial still has me a bit confused though.

 

Your settings look pretty normal to me, except I only align RGB when necessary (say, atmospheric dispersion imaging at the horizon).  If 2.0 oddly isn't getting rid of the trail, you might just uncheck those subs from being used.  Should only be a few.

 

For the colored smears, again I don't know astrocam dark procedure, but you could try the cosmetic hot pixel removal options in the second to last tab.  I gave it a quick run on your sub (guassian instead of median because there is only one frame) and a subtraction did show some stuff being removed.  You might need to play with the strength of the settings.



#10 acrh2

acrh2

    Vostok 1

  • -----
  • Posts: 123
  • Joined: 16 Mar 2021

Posted 15 May 2021 - 03:08 PM

My ASI533MC Pro needs BGGR pattern, or the red and blue colors get reversed with RGGB.

(EDIT: I use Sharpcap with the native ZWO driver.)

The red, blue and green blobs of pixels look like darks were never applied at all.

They should be gone when darks are properly applied. There might still be individual hot pixels here and there after the darks have been applied but those can be removed by the Cosmetic settings.

My light and dark Stacking settings are Average and Median, respectively. I think those were defaults after installation.

 

Perhaps what is truly happening, and this is a guess on my part, is that the hot pixels haven't been properly burned in because the camera is brand new. So the hot pixels may need some time to properly settle after you use the camera for maybe, I don't know, a hundred hours of use?

 

EDIT: I recently learned that you want to set your color balance options for the camera hardware to 50/50 R/B, which means that raw data will be used without any alteration by the driver. And have the exact same settings for darks as well.


Edited by acrh2, 15 May 2021 - 03:50 PM.


#11 rj144

rj144

    Viking 1

  • -----
  • Posts: 529
  • Joined: 31 Oct 2020

Posted 15 May 2021 - 03:40 PM

Good advice here.

 

Just want to point out that if you have comet streaks in your picture, it's a very bad sign for all of us probably.  ;)



#12 Muskoka

Muskoka

    Vostok 1

  • -----
  • topic starter
  • Posts: 103
  • Joined: 08 Apr 2020
  • Loc: Muskoka Ontario

Posted 15 May 2021 - 05:04 PM

Thanks everyone for the responses.

 

Still not getting anywhere with this. I removed all the images that had streaks, they were not being removed at all in DSS. Again, never had that issue when using the dslr.

 

So the stack is clear of the streaks now. I tried BGGR, not sure if it's any better. There's a huge glow now all the way down the left side after a stretch, no idea what that's from. Not getting anything reasonable, or worth looking at from Startools yet. 

 

I'm wondering if it's a Nina thing. These were taken with the native driver, wondering if I should try the Ascom driver instead. It was suggested yesterday before I started the nights shoot to use the native driver, but the results don't look very good at all.



#13 acrh2

acrh2

    Vostok 1

  • -----
  • Posts: 123
  • Joined: 16 Mar 2021

Posted 15 May 2021 - 08:05 PM

Thanks everyone for the responses.

 

Still not getting anywhere with this. I removed all the images that had streaks, they were not being removed at all in DSS. Again, never had that issue when using the dslr.

 

So the stack is clear of the streaks now. I tried BGGR, not sure if it's any better. There's a huge glow now all the way down the left side after a stretch, no idea what that's from. Not getting anything reasonable, or worth looking at from Startools yet. 

 

I'm wondering if it's a Nina thing. These were taken with the native driver, wondering if I should try the Ascom driver instead. It was suggested yesterday before I started the nights shoot to use the native driver, but the results don't look very good at all.

I tried to wrap my head around it and I didn't come up with any kind of breakthroughs. 

However, I did process your fits files, and I had the same results as you.

Which tells me that, the problem must be with the fits files themselves.

 

Perhaps it is some setting in Nina that does this?

I use Sharpcap, so I am not familiar with those.\

 

Good luck.



#14 cybermayberry

cybermayberry

    Vostok 1

  • -----
  • Posts: 199
  • Joined: 26 Aug 2010

Posted 15 May 2021 - 08:28 PM

Are you using a UV/IR cut filter?Those blotches also all look like they have the same pattern.

Try to re-stack again with median stacking mode or median kappa sigma clipping and see if they are still present.

I download your single fit light frame. Debayered, did auto background extraction, and applied STF and there is nothing there that would explain the colored "blobs".

Based on what I am seeing in the single fit file you supplied, it's a DSS issue.

M_101_2021.jpg


Edited by cybermayberry, 15 May 2021 - 08:56 PM.


#15 acrh2

acrh2

    Vostok 1

  • -----
  • Posts: 123
  • Joined: 16 Mar 2021

Posted 15 May 2021 - 09:11 PM

Are you using a UV/IR cut filter?Those blotches also all look like they have the same pattern.

Try to re-stack again with median stacking mode or median kappa sigma clipping and see if they are still present.

I download your single fit light frame. Debayered, did auto background extraction, and applied STF and there is nothing there that would explain the colored "blobs".

Based on what I am seeing in the single fit file you supplied, it's a DSS issue.

attachicon.gifM_101_2021.jpg

Very interesting. If I may ask, how did you Debayer and auto-background extract those fits?



#16 cybermayberry

cybermayberry

    Vostok 1

  • -----
  • Posts: 199
  • Joined: 26 Aug 2010

Posted 15 May 2021 - 09:38 PM

Very interesting. If I may ask, how did you Debayer and auto-background extract those fits?

I used Pixinsight. 



#17 acrh2

acrh2

    Vostok 1

  • -----
  • Posts: 123
  • Joined: 16 Mar 2021

Posted 15 May 2021 - 10:10 PM

I used Pixinsight. 

I tried using saturation on your picture, and it's no different from what the OP posted - unnatural yellow tint.

Seems to me like all Pixinsight did was reduce color saturation.

 

BTW, the "blobs" are simply hot pixel artifacts from stacking, a stack that was dithered. 



#18 cybermayberry

cybermayberry

    Vostok 1

  • -----
  • Posts: 199
  • Joined: 26 Aug 2010

Posted 15 May 2021 - 11:12 PM

I tried using saturation on your picture, and it's no different from what the OP posted - unnatural yellow tint.

Seems to me like all Pixinsight did was reduce color saturation.

 

BTW, the "blobs" are simply hot pixel artifacts from stacking, a stack that was dithered. 

I did nothing to effect color other then auto background extraction. I did not do any color correction or calibration.

Any hot pixels should be removed via the stacking process (dithering should help in this not make it worse). There is nothing apparently wrong with the cameras sensor that would produce such results.

I did look at the dark frame supplied, here the result with a STF applied. This isn't not helping the situation at all. This would help explain the issue with the odd gradient

I am still willing to bet changing the stacking parameters in DSS will more then likely remove the color artifacts. If median doesn't work try average forget the kappa clipping if that ineffective.

You should also redo your calibration frames.

Attached Thumbnails

  • Target_2021-05-14_19-01-58_101_70_60.00s_0001.jpg

Edited by cybermayberry, 15 May 2021 - 11:29 PM.


#19 acrh2

acrh2

    Vostok 1

  • -----
  • Posts: 123
  • Joined: 16 Mar 2021

Posted 15 May 2021 - 11:25 PM

Any hot pixels should be removed via the stacking process (dithering should help in this not make it worse).

I am new here, but the artifacts of red/green/blue pixel "blobs" look exactly like stacking raw frames, that were dithered, without using darks. 

Have you tried doing that? I have. It's the same result.

I am sorry. Do you actually use DSS or just Pixinsight?



#20 cybermayberry

cybermayberry

    Vostok 1

  • -----
  • Posts: 199
  • Joined: 26 Aug 2010

Posted 15 May 2021 - 11:54 PM

I am new here, but the artifacts of red/green/blue pixel "blobs" look exactly like stacking raw frames, that were dithered, without using darks. 

Have you tried doing that? I have. It's the same result.

I am sorry. Do you actually use DSS or just Pixinsight?

I used DSS and dealt with this exact same issue with a 533MCPro

The issue arose when I installed a DSS update. I was able to fix it the first time by changing stacking parameters. I left DSS for PI soon after, and have never had an issue since with this problem since.

Dither itself removes hot pixels, cold pixels, along with cosmic rays, satellites, and fix pattern noise.



#21 acrh2

acrh2

    Vostok 1

  • -----
  • Posts: 123
  • Joined: 16 Mar 2021

Posted 16 May 2021 - 12:37 AM

I used DSS and dealt with this exact same issue with a 533MCPro

The issue arose when I installed a DSS update. I was able to fix it the first time by changing stacking parameters. I left DSS for PI soon after, and have never had an issue since with this problem since.

Dither itself removes hot pixels, cold pixels, along with cosmic rays, satellites, and fix pattern noise.

I am still learning, so could you please explain how stacking/dither can remove hot pixels? I thought that stacking+dithering would simply "dilute" hot pixels - their intensity would be diminished by the number of stacked frames. For example, stacking of 100 frames would reduce hot pixel intensity by a factor of 100.



#22 Mike in Rancho

Mike in Rancho

    Viking 1

  • -----
  • Posts: 882
  • Joined: 15 Oct 2020
  • Loc: Alta Loma, CA

Posted 16 May 2021 - 12:38 AM

Dither only removes those things (satellites?) depending on your stacking and rejection parameters.  That will be the case regardless of the stacking software.

 

I downloaded the dark too just now to check it out, and found the same odd rippling effect.

 

I also played around with them and the settings.  Hey, when there's only one light and one dark, stacking is fast!

 

The results - after multiple perturbations, it seems to me the dark is not scaled right.  Why, I don't know, but then again I don't use astrocams.

 

However, if I set the darks to "dark optimization," then I got a decent result.  Any other way was just a hot mess and will not properly stretch in Gimp.  In fact, I even subtracted versions of the dark-optimized "stack" with and without cosmetic hot/cold pixel removal, and didn't turn anything up.  Contrast that with the single light sub, which when I did that same subtraction showed red, blue, and green hot pixels all over the place.

 

So, you may not need the cosmetic hot pixels after all.  Try checking the optimization box and see what happens.  Probably turn off RGB align too, for now.  I bet both PS and Startools will be much happier with your output.  Then we can go from there and see what else might need doing.  I'm still not sure about the R and B being correct or swapped over.  Because of that I'm not entirely sure of the proper coloration here, but...here's a 1 light 1 dark post-process (no star shrinking obviously lol).

 

Mukosa 1 light 1 dark.jpg


  • Muskoka likes this

#23 acrh2

acrh2

    Vostok 1

  • -----
  • Posts: 123
  • Joined: 16 Mar 2021

Posted 16 May 2021 - 12:55 AM

Dither only removes those things (satellites?) depending on your stacking and rejection parameters.  That will be the case regardless of the stacking software.

 

I downloaded the dark too just now to check it out, and found the same odd rippling effect.

 

I also played around with them and the settings.  Hey, when there's only one light and one dark, stacking is fast!

 

The results - after multiple perturbations, it seems to me the dark is not scaled right.  Why, I don't know, but then again I don't use astrocams.

 

However, if I set the darks to "dark optimization," then I got a decent result.  Any other way was just a hot mess and will not properly stretch in Gimp.  In fact, I even subtracted versions of the dark-optimized "stack" with and without cosmetic hot/cold pixel removal, and didn't turn anything up.  Contrast that with the single light sub, which when I did that same subtraction showed red, blue, and green hot pixels all over the place.

 

So, you may not need the cosmetic hot pixels after all.  Try checking the optimization box and see what happens.  Probably turn off RGB align too, for now.  I bet both PS and Startools will be much happier with your output.  Then we can go from there and see what else might need doing.  I'm still not sure about the R and B being correct or swapped over.  Because of that I'm not entirely sure of the proper coloration here, but...here's a 1 light 1 dark post-process (no star shrinking obviously lol).

 

attachicon.gifMukosa 1 light 1 dark.jpg

Wait. You got that from a single light and a single dark that the OP posted?

 

Teach me, sensei!


  • Baron von Smoogle likes this

#24 Mike in Rancho

Mike in Rancho

    Viking 1

  • -----
  • Posts: 882
  • Joined: 15 Oct 2020
  • Loc: Alta Loma, CA

Posted 16 May 2021 - 01:59 AM

Wait. You got that from a single light and a single dark that the OP posted?

 

Teach me, sensei!

Ha, I don't know about that.  But oddly it does harken me back to judo class at the Y in the 70's.  lol.gif

 

It was still bugging me so I tried to make another one.  This was with RGGB, actually, and for whatever reason it balanced out better.  ?

 

Mukosa 1-light 1-optimizeddark and RGGB.jpg

 

Basic DSS and Startools.  I registered and then stacked the one light and one dark frame.  Obviously there will be no rejection algorithms possible.  In FITS I set it to RGGB.  Dark optimization on, and Dark hot pixel removal.  No cosmetic.  No RGB align.  Intermediate and final files in FITS.

 

Basic Startools processing.  Immediately superpixel binned it down to like the 750x750 final size here, because with 1 frame we're not going to do deconvolution or much else really.  Wipe.  With the darks optimized or scaled or whatever it is, wipe was easy peasy.  Autodev - again, there's only so far we can stretch this frame out.  I didn't try to stretch out what wasn't there, just treated the background like I would any other processing.  Contrast - with some compression of the shadows into a smaller dynamic range to help clear out the sky noise (similar to an S when doing curves).  HDR - didn't help a lot because there's not too much here.  Sharpening.  Color balance and saturation.  SS and then denoise, which I was a little more aggressive with because, again, one light frame lol.  Save file!  Then for this one instead of direct to jpg, I took the tiff out of Startools and used gimp for the jpg export, because it has very good compression percentage options.

 

Hopefully Mukosa can get a good result, and maybe even toss those satellite subs back in to see if the stacking with optimized darks removes things in the way he is used to.

 

EDIT:  Yes I just checked Stellarium for star color indices, and this seems more correct now.  For the bottom two big stars, the left is blue-ish and the right is red-ish.


Edited by Mike in Rancho, 16 May 2021 - 02:08 AM.

  • Muskoka likes this

#25 Muskoka

Muskoka

    Vostok 1

  • -----
  • topic starter
  • Posts: 103
  • Joined: 08 Apr 2020
  • Loc: Muskoka Ontario

Posted 16 May 2021 - 08:42 AM

Wow, you guys have been busy.smile.gif

 

Just getting the day started, didn't shut down until 3am this morning so just having coffee. Tried a bunch of different things last night shooting wise. Changed from Native driver in Nina to Ascom, took a bunch of shots, and they were very close in comparison, but there was a difference, albeit slight. Tried some shots from the AsiStudio program, and Sharpcap. All results seemed very close. 

 

To answer one question from above, no I don't have a uv/ir cut filter yet, ordered 2 days ago, and it's still not shipped.

 

Question about shooting darks with a astro cam. Is the procedure any different than with a dslr? Put the cap on the scope, and shoot at the same settings as the lights? One thing I'm not sure of, can light leak in past the grills on the sides of the 533? I took the 20 darks during the day with the cap on, at the same temperature, gain, offset, and exposure, as the lights.

 

I'm just about to give DSS another go with the suggested settings above, will report back.


  • Mike in Rancho likes this


CNers have asked about a donation box for Cloudy Nights over the years, so here you go. Donation is not required by any means, so please enjoy your stay.


Recent Topics






Cloudy Nights LLC
Cloudy Nights Sponsor: Astronomics