#1  
Old 18-10-2013, 06:18 AM
dcalleja's Avatar
dcalleja
Registered User

dcalleja is offline
 
Join Date: Oct 2005
Location: Sydney
Posts: 513
A Question on Darks

All
I use an SBIG STF8300 and have always taken my darks and subtracted them 'on the fly' in maxim during imaging

Does anyone have a view on this technique vs building specific dark masters and using them for calibration later.

Which technique gives a better result.?

Thanks
Dan
Reply With Quote
  #2  
Old 18-10-2013, 06:46 AM
naskies's Avatar
naskies (Dave)
Registered User

naskies is offline
 
Join Date: Jul 2011
Location: Brisbane
Posts: 1,865
I have the same camera - master dark frames will always give you better results.

Just like light frames, dark frames also have noise (random fluctuations in the dark noise, cosmic rays, radiation hits, etc). If you only subtract a single dark frame, then you effectively add the noise of the dark frame to your light frames. Worse still - I believe that Maxim re-uses dark frames for consecutive shots, so you won't be cancelling out that noise with multiple light frames.

I find that stacking at least 10 darks gives me a noticeable improvement over a single dark frame, though I usually try to aim for 30+. You'll get better results from shooting master dark frames for each sub duration that you use (unless you go to the hassle of using overscan calibration in PixInSight).

That said, the 8300 chips are very low noise in 1x1 binning so if you're shooting bright targets such as Orion Nebula and not stretching your subs particularly hard you may not even notice much/any difference between a single dark vs a master dark.
Reply With Quote
  #3  
Old 18-10-2013, 07:23 PM
White Rabbit's Avatar
White Rabbit
Space Cadet

White Rabbit is offline
 
Join Date: Feb 2007
Location: Sydney
Posts: 1,409
Ditto the above, but I stack 50 darks, 50 flats (for each filter)and 50 bias. I've heard some people taking as as many as 500 bias frames. I have a master library for every bin mode as well.
You want an average of the noise floor of your camera, so the more darks and bias frames you take the better your average, to a point. There comes a point adding more frames is not that benificial.

I'm no expert and I'm on the learning curve as well but this is what I've read from various places and it seems to work.
Read this it explains a lot http://deepskystacker.free.fr/english/index.html

Its almost full moon so get cracking in your dark library
Reply With Quote
  #4  
Old 18-10-2013, 07:37 PM
dcalleja's Avatar
dcalleja
Registered User

dcalleja is offline
 
Join Date: Oct 2005
Location: Sydney
Posts: 513
Thanks folks
This seems like good advice. The moon is indeed rising (and very red too) so I will make a start

Dan
Reply With Quote
  #5  
Old 21-10-2013, 09:50 AM
White Rabbit's Avatar
White Rabbit
Space Cadet

White Rabbit is offline
 
Join Date: Feb 2007
Location: Sydney
Posts: 1,409
Sorry just realised that the link I posted does not point to the article. Click on the "How to make better images" link on the page my link points to.
Reply With Quote
  #6  
Old 21-10-2013, 09:52 AM
multiweb's Avatar
multiweb (Marc)
ze frogginator

multiweb is offline
 
Join Date: Oct 2007
Location: Hinchinbrook
Posts: 18,140
Library of 20min darks. I always re-shoot flats per session and corresponding dark flats same duration same temperature.
Reply With Quote
  #7  
Old 24-10-2013, 09:20 PM
2stroke's Avatar
2stroke (Jay)
The devil's advocate

2stroke is offline
 
Join Date: Jul 2011
Posts: 816
As above but dark flats? I've never used them, its always bias darks and flats for my cals. Will read up on this one lol. My darks are for given temps and duration to work with subs of the same, i also go for 20mins-30mins worth of dark data, 6x5mins 10x2mins ect
Reply With Quote
  #8  
Old 25-10-2013, 08:33 AM
Merlin66's Avatar
Merlin66 (Ken)
Spectroscopy Wizard

Merlin66 is offline
 
Join Date: Oct 2005
Location: St Leonards, Vic
Posts: 6,968
Every CCD frame taken includes the Bias....
I've been following the recommendations given in the AstroArt guide (p135) since moving into CCD work.
-Subtract a (master) dark frame a 5-10 image stack ( average or median combine) at the same temp and exposure as the lights.
- Divide by a (master) flat field (also corrected with it's own dark frame), a 5-10 image stack ( average or median combine)

This does away with the need for separate Bias frames (unless you're trying to scale your darks ie use a stack of ten minute darks to replicate a 40min exposure......)
Reply With Quote
  #9  
Old 25-10-2013, 11:13 AM
multiweb's Avatar
multiweb (Marc)
ze frogginator

multiweb is offline
 
Join Date: Oct 2007
Location: Hinchinbrook
Posts: 18,140
Quote:
Originally Posted by 2stroke View Post
As above but dark flats? I've never used them, its always bias darks and flats for my cals. Will read up on this one lol. My darks are for given temps and duration to work with subs of the same, i also go for 20mins-30mins worth of dark data, 6x5mins 10x2mins ect
What I call dark flat is probably the same. If I shoot a 5s flat I do a corresponding 5s bias. I chose to do a libray of 20min darks because I never go over 20min for each sub and I scale them down when calibrating.
Reply With Quote
  #10  
Old 25-10-2013, 12:50 PM
Paul Haese's Avatar
Paul Haese
Registered User

Paul Haese is offline
 
Join Date: Jan 2009
Location: Adelaide
Posts: 9,393
Every three months I take a battery of darks of the length required. Usually two lots one at 30 minutes and one at 20 minutes.

You don't need as many darks if you dither. I am using about 10 darks per 30 frames of lights. I master the darks up and keep those until I need redo them in three months.

I don't use dark flats either.
Reply With Quote
  #11  
Old 25-10-2013, 12:59 PM
2stroke's Avatar
2stroke (Jay)
The devil's advocate

2stroke is offline
 
Join Date: Jul 2011
Posts: 816
Quote:
Originally Posted by Merlin66 View Post
Every CCD frame taken includes the Bias....
I've been following the recommendations given in the AstroArt guide (p135) since moving into CCD work.
-Subtract a (master) dark frame a 5-10 image stack ( average or median combine) at the same temp and exposure as the lights.
- Divide by a (master) flat field (also corrected with it's own dark frame), a 5-10 image stack ( average or median combine)

This does away with the need for separate Bias frames (unless you're trying to scale your darks ie use a stack of ten minute darks to replicate a 40min exposure......)
Yer i scale darks in maxim when needed, though arn't subtracting the bias twice or are you dividing the whole image data by the master light value after subtracting you your dark. That seems weird as i thought it was

cal'd data = Sub - (Dark with bias data included) - ((Flat with bias data included) + (Bais frame as you have subtracted it twice))

Can you explain how your doing it with maths is it

cal'd data = (Sub -(Dark with bias data included)) / (Flat with bias data included)

so like if you had a pixel that was warm and its value was 102 for your dark inc bias value, your bias value was 2, your flat was 204 and your sub was 2000

(2000 - 102)/204
1898/204 = 9.3~

hmm that dosent seem right lolz, have a cold atm and doped up so help a brother out lol, i'am now expert either and i know you sure as hell have a vast experience please show me your maths.

Last edited by 2stroke; 25-10-2013 at 01:29 PM.
Reply With Quote
  #12  
Old 25-10-2013, 01:27 PM
Merlin66's Avatar
Merlin66 (Ken)
Spectroscopy Wizard

Merlin66 is offline
 
Join Date: Oct 2005
Location: St Leonards, Vic
Posts: 6,968
cal'd data = Sub (Bias included) -(Dark with bias data included) / (Flat (Bias included) corrected with it's own dark which includes bias data)

which gives:
cal'd data = sub(bias) - Dark(bias)/ flat
Reply With Quote
  #13  
Old 25-10-2013, 02:13 PM
2stroke's Avatar
2stroke (Jay)
The devil's advocate

2stroke is offline
 
Join Date: Jul 2011
Posts: 816
just seems weird because if you put values in for a pixel and say

pixel=20 dark inc bias=4 flat inc bias=2 bias =2

your way
20- (4/2) = 18

my way
(((20 - 4) -2) +2) = 16
Reply With Quote
  #14  
Old 25-10-2013, 02:19 PM
Merlin66's Avatar
Merlin66 (Ken)
Spectroscopy Wizard

Merlin66 is offline
 
Join Date: Oct 2005
Location: St Leonards, Vic
Posts: 6,968
Jay,
pixel=20
dark inc bias=4
flat inc bias=2
bias =2

your way
20- (4/2) = 18

No not correct...
The flat dark removes the Bias
20 - (4/(2-2))=16
Reply With Quote
  #15  
Old 25-10-2013, 02:24 PM
2stroke's Avatar
2stroke (Jay)
The devil's advocate

2stroke is offline
 
Join Date: Jul 2011
Posts: 816
Quote:
Originally Posted by Merlin66 View Post
Jay,
pixel=20
dark inc bias=4
flat inc bias=2
bias =2

your way
20- (4/2) = 18

No not correct...
The flat dark removes the Bias
20 - (4/(2-2))=16
Sweet that makes alot more sense now lol
Reply With Quote
  #16  
Old 25-10-2013, 02:25 PM
naskies's Avatar
naskies (Dave)
Registered User

naskies is offline
 
Join Date: Jul 2011
Location: Brisbane
Posts: 1,865
With darks and dark flat frames, each pixel consists of the following:

LightPixel = Bias + Dark + LightSignal
FlatPixel = Bias + DarkFlat + FlatSignal
DarkPixel = Bias + Dark
DarkFlatPixel = Bias + DarkFlat

Calibration involves the following:

Calibrated Light Pixel = (LightPixel - DarkPixel) / (FlatPixel - DarkFlatPixel)
= (Bias + Dark + LightSignal - Bias - Dark) / (Bias + DarkFlat + FlatSignal - Bias - DarkFlat)
= LightSignal / FlatSignal
Reply With Quote
  #17  
Old 25-10-2013, 02:32 PM
2stroke's Avatar
2stroke (Jay)
The devil's advocate

2stroke is offline
 
Join Date: Jul 2011
Posts: 816
Quote:
Originally Posted by naskies View Post
With darks and dark flat frames, each pixel consists of the following:

LightPixel = Bias + Dark + LightSignal
FlatPixel = Bias + DarkFlat + FlatSignal
DarkPixel = Bias + Dark
DarkFlatPixel = Bias + DarkFlat

Calibration involves the following:

Calibrated Light Pixel = (LightPixel - DarkPixel) / (FlatPixel - DarkFlatPixel)
= (Bias + Dark + LightSignal - Bias - Dark) / (Bias + DarkFlat + FlatSignal - Bias - DarkFlat)
= LightSignal / FlatSignal
Thanks for that that sums it up even better. This looks like a very nice method so what are darkflats? I get that your using them to do away with bais but how are they taken? Are these a reversed values of flats? or flats - bais


Dark Flat
This is taken using the same settings (Exposure and ISO) as your "Flats (as above)" but with the telescope capped off so not to allow any light to enter the camera sensor. Take at least 25 exposures. googled lol to much codiene hahah
Reply With Quote
  #18  
Old 25-10-2013, 02:42 PM
naskies's Avatar
naskies (Dave)
Registered User

naskies is offline
 
Join Date: Jul 2011
Location: Brisbane
Posts: 1,865
Dark Flats are just dark frames taken with the same settings as your flat frames. For example, if your flats are 3 sec @ -10 deg C, then your dark flats need to be 3 sec @ -10 deg C as well.

This way, you don't need to explicitly take bias frames at all (and better results to boot), but all your darks/dark flats must be perfectly matched to the lights/flats.
Reply With Quote
  #19  
Old 25-10-2013, 02:47 PM
2stroke's Avatar
2stroke (Jay)
The devil's advocate

2stroke is offline
 
Join Date: Jul 2011
Posts: 816
Sweet will give a run on the next session, how long has this method being about? I always run flats at the start of the session once the camera cooled down anyhow so shouldn't be an issue though its no regulated which is a pain. Only on the learning path here though this method looks great
Reply With Quote
  #20  
Old 25-10-2013, 02:52 PM
multiweb's Avatar
multiweb (Marc)
ze frogginator

multiweb is offline
 
Join Date: Oct 2007
Location: Hinchinbrook
Posts: 18,140
Quote:
Originally Posted by 2stroke View Post
Sweet will give a run on the next session, how long has this method being about? I always run flats at the start of the session once the camera cooled down anyhow so shouldn't be an issue though its no regulated which is a pain. Only on the learning path here though this method looks great
Shoot your dark flats right after your flats and you'll get the same temperature. I've done it with my OSC for the past 5yrs and it's not regulated. Never had an issue. If you shoot your dark flats another night though it's likely it won't work. But you can use your flats/dark flats on any sub at any temperature as long as focus and camera orientation haven't changed.
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 03:45 AM.

Powered by vBulletin Version 3.8.7 | Copyright ©2000 - 2019, Jelsoft Enterprises Ltd.
Advertisement
Celestron Australia
Advertisement
Bintel
Advertisement
OzScopes Authorised Dealer
Advertisement
Lunatico Astronomical
Advertisement
Meade Australia
Advertisement
SkyWatcher Australia
Advertisement
NexDome Observatories
Advertisement
Astronomy and Electronics Centre
Advertisement