#1  
Old 23-07-2011, 02:12 AM
netwolf's Avatar
netwolf
Registered User

netwolf is offline
 
Join Date: Jan 2005
Posts: 2,943
Bias, Darks, Flats and Scaling.

Hi All,

It just wont stop raining in Sydney, so plenty of time to dwell on processing and make sure I am geting this right.

Please check this and see if I got this right.

1. 10 Dark frames aquired at Temp -10 and 1x1 bin at X seconds.
2. 10 or more Bias frames aquired at Temp -10 and 1x1 bin, near to zero seconds. Create Master Bias.
3. Remove Master Bias and median stack 10 darks to make Scalable Master Dark.

Then we can
1. Acquire Light frames taken at -10 and 1x1 bin, with expousre time < X seconds.
2. Above Master Bias removed from Lights.
3. Remove Master scalable dark from Lights.

Is that right essentially you can use the Master Scalable Dark and Master Bias with any Lights with exposure time less than that of the Darks. So ideally you only need to keep Darks of the longest Exposure you are likley to do for a given Temp and Binning.

Do Bias frames need to be taken in the optical train?

Now I assume for Flat frames of Exposure < X with Temp -10 and 1x1 bin we can also use the same Scalable Master Dark and Master Bias as above to create the Master Flat Frame to remove from the above Lights.

From what I have read Flat frames exposure time needs to be set to get 1/3 to 2/3 saturation level of the chip. So then am I correct in assuming that Flat frames dont have to be the same Exposure lenght as the Lights?
Further then the Master Flat frame thus created as explained above, woudl be applicable to any Lights with temp and binning the same as the Flats.

So in the end you have in your library.
1. Scalable Master Dark for each Temp -T, and binning ZxZ with exposure equal to or greater thant the largest exposure you are going to take.
2. Master Bias for each Temp -T, and binning ZxZ, zero second exposure.
3. Master Flat for each Temp -T, and binning ZxZ, 1/3 to 2/3 saturation.

So then your librbary is independant of Exposure time, as long as Exposure time of Lights is less than the Scalable Master Dark.

Ouch my brain hurts.

Regards
Fahim
Reply With Quote
  #2  
Old 23-07-2011, 06:14 AM
gbeal
Registered User

gbeal is offline
 
Join Date: Oct 2004
Location: New Zealand
Posts: 4,345
Hi Fahim,
while I'm no pro on this, I couldn't think as much as you have about this.
Granted, I use a cooled CCDm, but even so, try to keep it simple.
Typically I will shoot a series of subs, and "normally" they are 300 seconds. Occasionally I will do 600 or 900 seconds, but of late 300 seconds is where I'm at. The total might be 6 - 10 or even 20 subs.
At the end of the night, before I dismantle anything I shoot my flats, I use a "Multiweb made me buy it" EL Panel, and shoot about 10 frames. These are with the scope exactly as it was when I was shooting the "light subs", except I pint the scope towards the zenith and sit the EL Panel on the front objective.
Then I shoot my bias frames, again another 10. These are done with the scope capped, so no light enters.
With flats and bias it is my understanding you don not necessarily need to have the cooling/temperature the same. I normally do though.
Darks I shoot at some convenient time, as I can cool at will. I normally shoot about 5 - 10 of them as well, SAME temperature and exposure as the light subs. So in my case, darks are done at -10º and 300 seconds.
Being the tidy kiwi I am I make a sub-folder for each, "Bias", "Flat", and "Dark", and direct the respective frames to the respective folders.
Once done, I process, normally the next day, and use Images Plus, which makes a Master (Bias/Flat/Dark) for me. Once it has processed I delete the single Bias/Flats/Darks and just retain the Masters.
In my case I have occasionally used a "Library" of Bias and Darks, but never Flats, I shoot them every session.
Possibly your processing sequence depends a lot on your Processing software too, mine will vary from yours if we are using different, but the fact remains, shoot flats each time, and look at the Bias and Darks as a library if you want.
Gary
Reply With Quote
  #3  
Old 23-07-2011, 11:51 AM
jase (Jason)
Registered User

jase is offline
 
Join Date: Sep 2006
Location: Melbourne, Victoria
Posts: 3,916
Quote:
Originally Posted by netwolf View Post
Is that right essentially you can use the Master Scalable Dark and Master Bias with any Lights with exposure time less than that of the Darks. So ideally you only need to keep Darks of the longest Exposure you are likley to do for a given Temp and Binning.
Yes, that is correct. Though the Master Scalable Dark frame is in essence just master dark frame. Its not scaled until the calibration is being performed as the scale factor will depend on the light frame. As you note, for scaling to be successful the exposure time dark frame must be at or longer than any of your light frame exposures. In other words with a 1800s (30min) dark, you can scale to a light frame sub exposure of anything below this. 120s, 300s, 900, 1200s whatever. This includes really short exposures such as .30s for flats if needed. What you can't do is scale up a dark frame. i.e. a 300s master dark can't accurate calibrate a 600s light frame.

Quote:
Originally Posted by netwolf View Post
Do Bias frames need to be taken in the optical train?
No. Bias frames are taken with the camera shutter closed. It is the signature of camera, not the optical path. Bias frame measure the bias level, bias structure and finally read out noise.

Note a dark frame also contains the above elements, but one important addition - the measurement of dark current which changes with time.

Quote:
Originally Posted by netwolf View Post
So then am I correct in assuming that Flat frames dont have to be the same Exposure lenght as the Lights?
Yes. Flat frames only need to be at an exposure duration that illuminates the field to capture pixel to pixel sensitivity. You mentioned 1/3 to 2/3 saturation which is correct. So for a camera with a well depth of 100,000 ADU, you want the hit around 33,000 ADU count as the saturation value. Need not be exact. I usually work on a 3,000 ADU tolerance on either side of the goal value. Reaching this value will vary depending on the light source. A bright EL panel may get you there in .3s whereas a sky flat on a darkening dusk sky could take 120s or more.

If you are taking sky flats, dark scaling is really the only way to achieve the task. A constantly varying sky brightness at dawn and dusk coupled to different light bandpass through each filter will generate flat exposures anywhere between .5s through to 240s. Taking darks for each possible flat sub exposure is impractical.

FWIW. I typically take 8 x 30min darks and use this for lights and flats with a bias to scale. Taking bias frames is quick and easy so there's no reason not to use them. This provides the most versatile calibration library as there is absolutely nothing I can't calibrate. In addition, its easy to refresh regularly as opposed to shooting darks for each light frame sub exposure you may use.

Hope this helps.
Reply With Quote
  #4  
Old 23-07-2011, 05:45 PM
Tandum's Avatar
Tandum (Robin)
Registered User

Tandum is offline
 
Join Date: Apr 2008
Location: Wynnum West, Brisbane.
Posts: 4,161
Quote:
Originally Posted by jase View Post
FWIW. I typically take 8 x 30min darks and use this for lights and flats with a bias to scale.
This would also make getting a new darks library trivial.
I'm guessing I'd need darks for each binning level. I tend to use binning a lot with the NB images.
Reply With Quote
  #5  
Old 24-07-2011, 12:58 AM
jase (Jason)
Registered User

jase is offline
 
Join Date: Sep 2006
Location: Melbourne, Victoria
Posts: 3,916
Quote:
Originally Posted by Tandum View Post
This would also make getting a new darks library trivial.
I'm guessing I'd need darks for each binning level. I tend to use binning a lot with the NB images.
Absolutely. If the camera is binned in 2x2, 3x3, etc. a full set of dark, flat and bias frames need to be acquired. The efficiencies of dark scaling still prevail however as the same rules apply.
Reply With Quote
  #6  
Old 24-07-2011, 01:18 AM
Tandum's Avatar
Tandum (Robin)
Registered User

Tandum is offline
 
Join Date: Apr 2008
Location: Wynnum West, Brisbane.
Posts: 4,161
Cheers Jase. I always knew maxim scaled cal files but now I know it does it to the extreeme ..... excellent ...
Reply With Quote
  #7  
Old 24-07-2011, 03:58 AM
netwolf's Avatar
netwolf
Registered User

netwolf is offline
 
Join Date: Jan 2005
Posts: 2,943
Thanks Gary and Jase, some great feedback there. I think i am starting to understand now.

I have the OSC M26c which I hope will not need Darks, but I will take them anyway for any hot pixels. But the H9 and my friends St-8300m will need them. So the above helps a lot in getting my thoughts in order.

So for the Bias and darks i can do them while i wait out the bad weather. The clay around my house has turned to mush and the lawn has just sprouted, hopefully the new moon brings better weather.

I will investigate the EL screens again and see what I can get.

Now this is probably a dumb question but how does one tell what Saturation point you have reached in a Flat frame? Do the EL screens have adjustable brightness controls?

Regards
Fahim
Reply With Quote
  #8  
Old 24-07-2011, 06:18 AM
troypiggo's Avatar
troypiggo (Troy)
Bust Duster

troypiggo is offline
 
Join Date: Nov 2008
Location: Brisbane, Australia
Posts: 4,846
How good/effective are the scaled darks to ones actually taken at the shorter exposure time? I can't seem to find documentation on how they're scaled. I mean, how does the software know which pixels to use at lower exposures? I must be missing something.
Reply With Quote
  #9  
Old 24-07-2011, 09:58 AM
marki's Avatar
marki
Waiting for next electron

marki is offline
 
Join Date: Mar 2008
Posts: 2,427
If you can get your hands on a copy of "The Handbook of Astronomical Image Processing" by Richard Berry and James Burnell it is all covered in great detail (including the math involved). They recommend your darks should be 5 x longer then the longest light sub you intend to take to get the best results using scaled darks. I have been using their method for some time now and it works very well plus it is convienient having only 1 master for each bin regardless of exposure time. Maxim, CCDstack and AIP4WIN let you add the bias to the master at the time of calibration which allows for scaling to take place and this can be done as a batch process which speeds things up a lot.

Mark
Reply With Quote
  #10  
Old 25-07-2011, 08:34 AM
netwolf's Avatar
netwolf
Registered User

netwolf is offline
 
Join Date: Jan 2005
Posts: 2,943
Mark, thanks for the book refrence, that book keeps coming up as a refrence in Image processign circles. I think I need to buy this one.
Reply With Quote
  #11  
Old 25-07-2011, 09:05 AM
troypiggo's Avatar
troypiggo (Troy)
Bust Duster

troypiggo is offline
 
Join Date: Nov 2008
Location: Brisbane, Australia
Posts: 4,846
Thanks Mark.
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 12:40 PM.

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