Question on Dark Frame Library

[Pogo]

Hi all

I am rethinking my Dark library set, as I have to reshoot a large part of it.
So far I had a master dark created of 20 frames of e.g. 15minutes, 10 minutes etc. So I had the exact exposure time in the darks as in the light frames.

Now I was thinking of just creating one master dark of e.g. 30*30minutes and then simply use scaling.
However I can't find anywhere whether this leads to the same results as having a master dark that matches the exposure time.

Having one master dark and use scaling will take a lot less time to shoot then having individual length ones.

What do you guys use, or recommend?

Thanks
Martin

[jkmorse]

Martin,

Like you, I was always building a whole library because I was paranoid about scaling not working.  However recently I just said that's crazy and have been using one long iteration Master Dark along with a Master Bias frame.  You need both since otherwise you end up scaling the bias noise in the Master Dark and you definitely don't want that.  So I load up both the scaled Master Dark and my Master Bias, then check the box under the Bias Frame and the two boxes under the Dark frame and it works like a charm.

One comment on your plan though is that I would shoot more dark and bias subframes.  No pain since you are doing that on off nights anyway so you are not missing anything.  I shoot 100 dark subs and 400 bias subs.  That creates really clean images for calibration purposes.

Best,

Jim   

[Pogo]

Hi Jim

Thanks for your reply and the information.
I am just starting to take 30min darks because that's the longest sub I use anyways.

Thanks for the hint on the Bias. I already have a Master Frame there (forgot to mention that), but "only" made up from 50 subs and then "boosted" with the Superbias process. Do you think it would be beneficial to take more Bias Frames for the master?

What do others do?

Cheers
Martin

[LarryC]

That is more or less what I do too.  However, I've never been able to resolve whether PI is able to accurately scale dark thermal noise based on temperature as well as it scales based on exposure. As I understand it, whereas Dark noise scales in linear manner on exposure, thermal noise increases in a non-linear manner with temperature.  Does anyone know if PI's algorithm accounts for this or are darks simply scaled on exposure and assumes thermal noise is linear and ignores temperature? Or do I have it all wrong?

Larry

[Warhen]

Hi gents, to the best of my knowledge, no software can fully address temperature differences and it's best practice to use matching temp calibration/light frames. With my part-time work as a customer service rep for QSI cameras, we see the importance of using contemporaneous frames. In some cases even a matching sensor temperature set point can yield imperfect calibration if ambient conditions varied a lot between the acquisition of light frames and dark frames. All that being said, ImageCalibration's Optimize feature does a great job of evening out any disparity in time (and temp). 

On the positive side Martin, a Jim says, PI does accurately scale a long dark to shorter lights, as well as flats, making flat dark acquisition unnecessary. One caveat, and I'm not certain of this, but I think that the scaling ability of the BPP script is more limited than ImageCalibration. Juan can tell us for sure, but with flats I don't think it's an issue as dark subtraction is often skipped when the action will add noise.

Martin I think 50 bias are great, and your idea of Superbias is great too. Jim, that's a lot of calibration frames! :>) I generally do 50 bias and 25 darks, but more is good too!

[Pogo]

Thanks for the replies everyone. I'll go ahead and try using just one MasterDark then. The Darks and bias Frames will of course be temperature matched to the light frames, so I don't expect much surprises there.

Cheers
Martin

[rdryfoos]

Hello all--I am in the process of shooting some darks for my library when I came across this thread.  If I make a master dark consisting of 33 40min dark frames, will that be good for calibrating all exposure times from 1 sec to 20 min (about my limit typically).  Is it really as good as using a dark composed of same exposure times as lights?  certainly 1 master dark will be easier than dozens.  I have always wondered which boxes to check in Image Calibration. Should I check calibrate under master Flat as well as Master Dark?  Should I only check calibrate under master dark when calibrating the flats and  darks and biases?  How about the lights"

Thanks,

Rodd

[Geoff]

This thread http://pixinsight.com/forum/index.php?topic=8839.msg57073#msg57073 by Juan is a must read.

[rdryfoos]

I have read it--and again.  For the life of me I can't say whether he implies that scaling works as well as same exposure times of darks and lights.   I understand teh theory, but applying it to that simple question is not so simple. 

[jkmorse]

Key is the following comment from Juan:

"This is not the case in practice, especially for high intensity values. As a result of this simplistic optimization model, the algorithm tends to undercorrect hot pixels. This is a very minor issue, however, since hot pixels are easily rejected during integration (because you dither your images, don't you?) and can be removed also with cosmetic correction techniques."

Bottm line, so long as you dither, you will fix any under-correction in the optimization model.

Best,

Jim

[rdryfoos]

One thing about dithering that is strange for me--my dithered images tend to have giga pixel sizes rather than mega pixel sizes.  Is this normal?  It extends processing time-I'm not concern with hard disk space.

[Geoff]

Dithering does not change the image size. Are you confusing dithering and drizzling?

[rdryfoos]

Well--doesn't drizzle come from dithering?  I thought one dithers so drizzle files can be created during alignment and used during drizzle integration. 

[jkmorse]

That is only one purpose of dithering.  But even when you are adequately or even over sampled, you still want to dither though you would get nothing from drizzle in those cases.  Dithering is critical to get the cleanest images, especially as a part of the image integration process.  Dithering moves the bad pixels around so when you align the images the bad pixels from each individual image end up in different places on the master stacked frame while the signal is now stacked all in the same place, making it much easier for the image integration rejection routines to give you the best possible image quality.  Dithering is one of the first skills every imager should master and every frame needs to be dithered to make it all work as intended.

Best,

Jim