#1  
Old 10-08-2017, 07:47 PM
Shiraz's Avatar
Shiraz (Ray)
Registered User

Shiraz is offline
 
Join Date: Apr 2010
Location: ardrossan south australia
Posts: 4,918
dark scaling with CMOS asi1600

Hi
while recently laid up, I tidied up some hard disks and found about 3 hours of quite good Helix neb subs that had not been processed. However the data was taken with a combination of gain/offset/temp/USB/time that I did not have darks for, so I could not dark cal it without a trip to my imaging site (where the camera is). CMOS chips can benefit from dark cal to tidy up slight fixed pattern irregularities in narrowband backgrounds.

Rather than wait until I had real darks, I tried dark scaling and got some pretty good results. The subs in question were 5 minute Ha and the closest darks (with the same gain/offset/temp/USB) were 10 minutes. The 10 minute master dark was manually scaled to 5 minutes in PI using pixelmath, to make sure that there was no unwanted optimisation. [masterdark5 = masterbias + 0.5*(masterdark10 - masterbias)].

The images in http://astrob.in/306733/0/ show the effect of the scaled dark - for illustration, the subs were stacked without alignment (so the stars show the dither pattern and the object is smeared out - but the fixed pattern noise is clear). Both uncalibrated and dark calibrated stacks (STF stretched) are shown for comparison. The two left hand panels show the stretched stack outcome with no dark cal - the faint glows typical of the 1600 are clearly visible, particularly in the lower crop panel which shows part of the bottom right corner of the full frame at 1:1 scale. The two right hand panels show the results after dark calibration with the scaled dark. It cleaned up most of the slightly warm pixels and removed the pattern structure well enough to turn almost unusable data into good data...

Not saying that dark scaling is the best way to do business with a CMOS chip - just pointing out that, if all else fails, manual scaling actually works OK and darks can usefully be scaled to different exposure times. Provided of course that the darks to be scaled are taken with the same gain/offset/temp/USB as the lights. Not a big deal, but hopefully of some use to CMOS owners. Thanks for looking. regards Ray

Last edited by Shiraz; 10-08-2017 at 08:52 PM.
Reply With Quote
  #2  
Old 10-08-2017, 08:36 PM
Atmos's Avatar
Atmos (Colin)
Ultimate Noob

Atmos is offline
 
Join Date: Aug 2011
Location: Melbourne
Posts: 6,980
It's turned a good considering all of the chatter of the end of the world that goes with Dark Scaling a CMOS chip.

Now you've just got to process the data
Reply With Quote
  #3  
Old 10-08-2017, 09:49 PM
glend (Glen)
Registered User

glend is offline
 
Join Date: Jun 2013
Location: Lake Macquarie
Posts: 7,033
Thanks Ray, good to know. Have not heard from you in awhile, hope your 'laid up' was not serious.
Reply With Quote
  #4  
Old 10-08-2017, 10:49 PM
Shiraz's Avatar
Shiraz (Ray)
Registered User

Shiraz is offline
 
Join Date: Apr 2010
Location: ardrossan south australia
Posts: 4,918
thanks Colin and Glen.
thanks for the concern Glen - am 70 - down this end of the bowling alley, everything is serious


FWIW, also found that dark scaling to account to correct for gain can work. Had a 10min master dark taken at gain 200 and used it to dark cal a sub taken at 10 min gain 100.

the gain 200 dark was scaled by 0.3, which is about the difference in gain from the ZWO website - the scaling function used is shown in the image http://astrob.in/306743/0/, which shows the "glow" region from the bottom right of the Helix sub, before and after calibration. The data had offsets sufficient to ensure that there was no excursion below zero.

As with time scaling, manual dark scaling to account for gain cleaned up the warm pixels and the fixed pattern and worked well enough to be usable - again, not recommending it as a standard procedure, but pointing out that it can work OK if all else fails. Regards Ray

Last edited by Shiraz; 11-08-2017 at 06:24 AM.
Reply With Quote
  #5  
Old 11-08-2017, 02:24 PM
Shiraz's Avatar
Shiraz (Ray)
Registered User

Shiraz is offline
 
Join Date: Apr 2010
Location: ardrossan south australia
Posts: 4,918
Quote:
Originally Posted by Atmos View Post
It's turned a good considering all of the chatter of the end of the world that goes with Dark Scaling a CMOS chip.

Now you've just got to process the data
here is - bit of overlap noise, but it still worked OK - http://astrob.in/306798/0/

regards Ray
Reply With Quote
  #6  
Old 11-08-2017, 04:07 PM
RickS's Avatar
RickS (Rick)
PI cult recruiter

RickS is offline
 
Join Date: Apr 2010
Location: Brisbane
Posts: 10,584
Looks good, Ray. Have you ever tried PI dark optimisation on CMOS data? Just wondering it if it works. I know it doesn't work when you do RBI preflash on a CCD. You get swirly patterns left behind.

Hope you're on the mend

Cheers,
Rick.
Reply With Quote
  #7  
Old 11-08-2017, 06:17 PM
Shiraz's Avatar
Shiraz (Ray)
Registered User

Shiraz is offline
 
Join Date: Apr 2010
Location: ardrossan south australia
Posts: 4,918
thank you Rick .

I tried using PI optimisation to calibrate with a non-matching dark, but it didn't work properly (see attached) - maybe the noise minimisation is upset by the fixed pattern noise?? Manual adjustment works fine.

regards Ray
Attached Thumbnails
Click for full-size image (automatic.jpg)
70.1 KB46 views
Reply With Quote
  #8  
Old 11-08-2017, 06:23 PM
RickS's Avatar
RickS (Rick)
PI cult recruiter

RickS is offline
 
Join Date: Apr 2010
Location: Brisbane
Posts: 10,584
Oh well
Reply With Quote
  #9  
Old 11-08-2017, 06:32 PM
Atmos's Avatar
Atmos (Colin)
Ultimate Noob

Atmos is offline
 
Join Date: Aug 2011
Location: Melbourne
Posts: 6,980
OR just heavily crop
Reply With Quote
Reply

Bookmarks

Thread Tools
Rate This Thread
Rate This Thread:

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off

Forum Jump


All times are GMT +10. The time is now 02:43 AM.

Powered by vBulletin Version 3.8.7 | Copyright ©2000 - 2024, Jelsoft Enterprises Ltd.
Advertisement
Testar
Advertisement
Bintel
Advertisement