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

Made an Oops - NB Imaging at 0 Gain?

  • Please log in to reply
12 replies to this topic

#1 ESzczesniak

ESzczesniak

    Vostok 1

  • -----
  • topic starter
  • Posts: 129
  • Joined: 30 Apr 2020

Posted 19 September 2020 - 12:34 PM

I thought I was getting a good start on my Heart Nebula last night with almost 6.5 hours of Halpha.  Focus is still a bit of an issue through the night, as I'm using ASIAIR to collect images, which still doesn't have autofocus capability.  So it's a Bahtinov mask at the start of the session, and then let it run.  But weather, time, etc seemed to cooperate. 

 

This morning, I see at some point the gain for my ASI1600MM got switched from unity at 139, to 0 for this image run.  I'm pretty new to this whole AP thing.  I have seen many people are doing LRGB at 0 gain, but NB is recommended at 139.  Probably for a number of reasons, but the most common I've seen is to reduce posterization. 

 

So, two questions:

 

1. Should I dump all the data at 0 gain and start over?  As a quick stretch and stack, it doesn't look terrible to my eye.  This is a link to the final stack if anyone is so inclined to take a look and tell me if it looks terrible or not ( https://drive.google...iew?usp=sharing ).  This is a stack of lights only.  I had all the calibration files for unity gain.  But I don't have any at 0 gain.  So if it shows some hope, I'll have to acquire those and restack.

 

2. If I keep this Halpha, there's no reason I couldn't go back to unity gain for OIII and SII, correct? 

 

I appreciate any input.  Unfortunately I don't know that my eye is good enough yet to say "this is bad data".  And while I'd rather not lose a whole night of imaging, I'd rather know now than after 2 more nights. 



#2 DRK73

DRK73

    Viking 1

  • *****
  • Posts: 703
  • Joined: 08 Mar 2013
  • Loc: Maryland, United States

Posted 19 September 2020 - 12:58 PM

I don't understand - if it looks good to you then what's the problem? 

 

The file you linked to is too large to view without downloading - can you put up a smaller scale image?



#3 ESzczesniak

ESzczesniak

    Vostok 1

  • -----
  • topic starter
  • Posts: 129
  • Joined: 30 Apr 2020

Posted 19 September 2020 - 01:23 PM

I don't understand - if it looks good to you then what's the problem? 

 

The file you linked to is too large to view without downloading - can you put up a smaller scale image?

I understand your point, but I am so new that things that looked good to me proved to have some pretty severe problems.  About a month ago on my first NB/SHO image, I thought the OIII was fine.  However, I was having problems stacking and someone pointed out that is was pretty badly out of focus.  In retrospect, I see it.  I didn't at the time though.  So I am hoping to learn from those with more experience and identify issues early where they are easier to fix.

 

The file linked above was the raw stack, as many have seemed to suggest that to be the most helpful really identifying.  Attached below is the stack with a simple autostretch in PI.  To me, it looks like there's plenty of data and I don't see any posterization.  But again, my eye lacks experience.  As noted, it is not calibrated, as I don't have and darks, bias, etc at a gain of 0 at the moment. 

 

More directly in this case, what I'm trying to make sure I understand (and am not sure I fully do yet):

 

1. What is bad about using a gain of 0?  I understand superficially that the data is less "amplified" and may get lost in noise.  But practically and at a deeper level, do not have a much deeper understanding.

 

2. What, if any, problems might I create for myself in processing using a gain of 0?

 

3. Would this prevent me from imaging OIII and SII at unity gain. 

 

4. Any obvious problems with the stacked image?

 

EDIT: If I'm critical of the image myself, if anything the signal is seeming stronger than I was expecting in the interior of the heart.  Is it possible this is actually read noise coming through and saturating true data/signal?

Attached Thumbnails

  • IC1805.jpg

Edited by ESzczesniak, 19 September 2020 - 01:31 PM.

  • dmdouglass and DRK73 like this

#4 DRK73

DRK73

    Viking 1

  • *****
  • Posts: 703
  • Joined: 08 Mar 2013
  • Loc: Maryland, United States

Posted 19 September 2020 - 01:52 PM

I can't really speak to your other questions regarding gain) but I think a lot of people would hope to get a finished image as good as your unfinished one. 


  • AAbby likes this

#5 RJF-Astro

RJF-Astro

    Viking 1

  • -----
  • Posts: 665
  • Joined: 13 Aug 2018
  • Loc: Zeist, Netherlands

Posted 19 September 2020 - 04:02 PM

I think the data is still very usable. So that is on of the more forgiving user errors. One thing I do notice is amp glow. The corners look brighter in the typical asi1600 pattern. Did you use separate darks at corresponding gains during calibration? If not, I would advise to do so.

I do not know the technical part of gain, maybe someone else does.

#6 ESzczesniak

ESzczesniak

    Vostok 1

  • -----
  • topic starter
  • Posts: 129
  • Joined: 30 Apr 2020

Posted 19 September 2020 - 04:43 PM

I think the data is still very usable. So that is on of the more forgiving user errors. One thing I do notice is amp glow. The corners look brighter in the typical asi1600 pattern. Did you use separate darks at corresponding gains during calibration? If not, I would advise to do so.

I do not know the technical part of gain, maybe someone else does.

Yes, definitely! It has characteristic ASI1600 amp glow.  I haven't applied any calibration filed yet.  I don't have any calibration frames at a gain of 0.  I wanted to mess with this a little before spending 3 hours taking dark frames (I'm using 300 second exposures). 



#7 Stelios

Stelios

    Voyager 1

  • *****
  • Moderators
  • Posts: 10,246
  • Joined: 04 Oct 2003
  • Loc: West Hills, CA

Posted 19 September 2020 - 05:30 PM

I understand your point, but I am so new that things that looked good to me proved to have some pretty severe problems.  About a month ago on my first NB/SHO image, I thought the OIII was fine.  However, I was having problems stacking and someone pointed out that is was pretty badly out of focus.  In retrospect, I see it.  I didn't at the time though.  So I am hoping to learn from those with more experience and identify issues early where they are easier to fix.

 

The file linked above was the raw stack, as many have seemed to suggest that to be the most helpful really identifying.  Attached below is the stack with a simple autostretch in PI.  To me, it looks like there's plenty of data and I don't see any posterization.  But again, my eye lacks experience.  As noted, it is not calibrated, as I don't have and darks, bias, etc at a gain of 0 at the moment. 

 

More directly in this case, what I'm trying to make sure I understand (and am not sure I fully do yet):

 

1. What is bad about using a gain of 0?  I understand superficially that the data is less "amplified" and may get lost in noise.  But practically and at a deeper level, do not have a much deeper understanding.

 

2. What, if any, problems might I create for myself in processing using a gain of 0?

 

3. Would this prevent me from imaging OIII and SII at unity gain. 

 

4. Any obvious problems with the stacked image?

 

EDIT: If I'm critical of the image myself, if anything the signal is seeming stronger than I was expecting in the interior of the heart.  Is it possible this is actually read noise coming through and saturating true data/signal?

1. You need longer total integration time to get the same results. You need longer individual exposures to swamp the (higher at gain 0) read noise. 

 

2. No problems as long as you have enough integration time and properly matched calibration frames.

3. Absolutely not. Or gain 200 for that matter. 

4. I like that image quite a bit. The interior of the heart is a famous object "Melotte 15" which is extremely beautiful and very bright. 

BTW, darks should be taken at home in a fridge (to allow cooling to low temperatures) while you sleep :)



#8 dmdouglass

dmdouglass

    Surveyor 1

  • *****
  • Posts: 1,967
  • Joined: 23 Dec 2007
  • Loc: Tempe, AZ

Posted 19 September 2020 - 05:43 PM

I think the data is still very usable. So that is on of the more forgiving user errors. One thing I do notice is amp glow. The corners look brighter in the typical asi1600 pattern. Did you use separate darks at corresponding gains during calibration? If not, I would advise to do so.

I do not know the technical part of gain, maybe someone else does.

Not sure where RJF-Astro is coming from. Here is a 300s dark (bin 2x2) with this camera, and the "Amp Glow" is VERY APPARENT on the right edge...  And, of course, with the calibration images, that disappears. Now, i don't ever do "0" gain, but the amp glow should be virtually non existent, i would think.  I have looked your image over and CANNOT see any amp glow.

 

By the way, beautiful image.

 

Edit: Add the image...

ZWO 2x2 Dark Master 300s PS.jpg


Edited by dmdouglass, 19 September 2020 - 05:45 PM.

  • elmiko likes this

#9 RJF-Astro

RJF-Astro

    Viking 1

  • -----
  • Posts: 665
  • Joined: 13 Aug 2018
  • Loc: Zeist, Netherlands

Posted 20 September 2020 - 01:11 AM

That looks like the 183s amp glow. Different pattern. The 1600 has two blotches on the right side and one top left. Could be happening in this image, but if the calibration was ok then it might be flats or just nebulosity. In any case, not a big deal. If it is not nebulosity, proper processing can get rid of it.
  • dmdouglass likes this

#10 ESzczesniak

ESzczesniak

    Vostok 1

  • -----
  • topic starter
  • Posts: 129
  • Joined: 30 Apr 2020

Posted 20 September 2020 - 04:20 PM

Thank you all for the input.  It seems sometimes it's better lucky than good!  Seems I'll keep the Halpha data at the moment.  Perhaps in the future I might revisit it, but will move on for now.  I got OIII last night and will get SII next time the weather cooperates.  It's not due to be cloudy here, but forecast transparency terrible again.  It'll give me some time to catch up on the calibration frames I need, and process a good stack for each of these while waiting for SII.



#11 dmdouglass

dmdouglass

    Surveyor 1

  • *****
  • Posts: 1,967
  • Joined: 23 Dec 2007
  • Loc: Tempe, AZ

Posted 20 September 2020 - 06:36 PM

That looks like the 183s amp glow. Different pattern. The 1600 has two blotches on the right side and one top left. Could be happening in this image, but if the calibration was ok then it might be flats or just nebulosity. In any case, not a big deal. If it is not nebulosity, proper processing can get rid of it.

Wow !!  Good call.  YES.. i have both cameras, and that is, in fact, the 183. I checked the fits header to verify.

And your description of the 1600 pattern is "spot on"...   And i had to really "stretch" it out to see it.

 

If i were the original poster (OP)... i would be VERY HAPPY with his results....


  • RJF-Astro likes this

#12 ESzczesniak

ESzczesniak

    Vostok 1

  • -----
  • topic starter
  • Posts: 129
  • Joined: 30 Apr 2020

Posted 27 September 2020 - 02:38 PM

Just as a follow-up to this topic.  Here's what I'm calling the final result at this point.  This is only the second NB image I've processed, and the first I processed past simple denoise/gradient extraction and then SHO combination.  For my lack of experience, I'm quite pleased with it.  However, I'm sure in a year or two I'll look back and want to make it better.  Each channel had it's own unique acquisition issues.  Halpha had the 0 gain mistake I made.  OIII was a good night, but still pre-auto focus on the ASIAIR Pro and was not refocused during acquisition while I slept.  And SII was captured on the best night we've had in a couple weeks, but was still pretty marginal transparency.  This image represents 16.42 hours of total integration in 300 s subs split evenly between each channel.  I welcome any comments and criticism. 

Attached Thumbnails

  • CN_IC1805_Final.jpg

  • dmdouglass, Andy Brown, Deesk06 and 2 others like this

#13 limeyx

limeyx

    Ranger 4

  • -----
  • Posts: 358
  • Joined: 23 May 2020
  • Loc: Seattle, WA

Posted 27 September 2020 - 04:05 PM

Just as a follow-up to this topic.  Here's what I'm calling the final result at this point.  This is only the second NB image I've processed, and the first I processed past simple denoise/gradient extraction and then SHO combination.  For my lack of experience, I'm quite pleased with it.  However, I'm sure in a year or two I'll look back and want to make it better.  Each channel had it's own unique acquisition issues.  Halpha had the 0 gain mistake I made.  OIII was a good night, but still pre-auto focus on the ASIAIR Pro and was not refocused during acquisition while I slept.  And SII was captured on the best night we've had in a couple weeks, but was still pretty marginal transparency.  This image represents 16.42 hours of total integration in 300 s subs split evenly between each channel.  I welcome any comments and criticism. 

Fantastic




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