Username:

Password:

Altair AstroDIO DehumidifiersAtik CamerasModern AstronomyDavid HindsNe3 Filters
UKAI is moving to Facebook

Author Topic: Image Calibration Frames (Bias, Darks and Flats)  (Read 5724 times)

0 Members and 1 Guest are viewing this topic.

Offline chris.bailey

  • Administrator
  • Galactic Poster
  • *****
  • Posts: 10009
  • If at last you do succeed, don't try again
    • Rosemount Observatory
Image Calibration Frames (Bias, Darks and Flats)
« on: 10:41:21, 27 February, 2013 »
In line with many aspects of this hobby, it is possible to find great images where no image calibration has been used but in general correct image calibration is a key component of good astro images.

How the calibration frames are used will largely depend on the processing software used. Deep Sky Stacker takes a rather chuck it all in and press go approach. This is fine as long as the contribution of each type of calibration frame is understood and in particular the impact if some are missed out. Software such as Maxim, Astro Art and Nebulosity take a more manual interactive approach and though PixInsight has a fully automated script, it is also possible to go fully manual.

Bias Frames (subtracted from light frames)

A bias frame is essentially a zero-length exposure (or as close as possible to zero length) with the shutter closed or with all light excluded. Each pixel will have a slightly different value, but except for a small amount of noise, the value for any one pixel will be consistent from image to image and may be subtracted from the light frames.

The bias frame itself contains a small amount of readout noise. This readout noise is produced inside the electronics that read the pixels. Combining a number of bias frames reduces this noise.

The bias for a particular camera is generally constant over a period of time. This means that you can take bias frames just once, and use them on all your images for many months. Bias frames are cheap to produce in terms of time expended so there is little reason not to take several of them.

Though bias signal is only weakly temperature dependent, it is worth taking bias frames at similar temperatures to that at which the light frames are acquired. With a regulated CCD there is no reason not to.

Dark Frames (subtracted from light frames)

Every camera sensor produces a certain amount of dark current, which accumulates in the pixels during an exposure. This dark current is time and temperature dependent and it accumulates at a different rate in every pixel with some being especially “hot” or “cold”.

A dark frame is an exposure taken under the same conditions as the light exposure, but with no light striking the sensor. Since each pixel is consistent in its dark current at any one temperature, the dark frame can be subtracted from the light frame to remove this fixed pattern from the image.

Dark frames also contain a noise component and though they will deal effectively with the dark current, they will also impart an element of noise. Simply subtracting a single dark frame from an image increases noise levels by 41%. Hence a number of dark frames should be taken and combined so as to reduce the noise level imparted. Dark frames are costly in terms of time expended, especially when sub-exposure times are lengthy, and should be carried out for each sub-exposure length utilised.

Dark current is highly temperature dependent and a small variation in temperature can have a significant effect on dark current levels. As DSLR’s are not temperature stabilised, use of dark frames with DSLR’s is of somewhat questionable value.

Dark frames themselves contain the bias signal and hence care needs to be taken that the software used for processing does not double subtract the bias signal, which would have the effect of adding noise back.

Some CCD sensors have very low dark current characteristics and for these it is perhaps better to take flats and bias and deal with any hot pixels by some form of hot pixel filter rather than use darks.

Flat Frames (divided into light frames)

CCD sensors have significant pixel-to-pixel variation in sensitivity. Vignetting in the optical system can reduce the light levels at the corners of the sensor. Dust on optical surfaces near the sensor can cast shadows (often called ”dust donuts” ). Good flat frames correct all of these defects.

Flats are by far the most troublesome calibration frame to acquire. The full aperture of the optical system must be evenly illuminated with light and if this is not done very carefully, then the flat frame can do more harm than good.

Electroluminescent (EL) panels are by far the easiest means of acquiring flat frames. http://www.earlsmann.co.uk/sections/product/id/10 are of reasonable quality. With the scope turned to a vertical position the EL panel can simply be laid on the end of the scope. Adjust exposure such that the histogram is no further than ½ way up the scale (this may require the addition of a few sheets of paper between the scope and the EL panel). If filters are being used, flat frames should be taken for each filter used.

If nothing significant in the imaging train changes from object to object, flat frames can be re-used though of all the calibration frames, they are the ones to re-do most often.
Flat frames themselves contain bias noise and if the flat frames are of more than a couple of seconds, also a dark signal component. As such they themselves require calibration. Subtracting the bias from flat frames is a vital step to correct calibration. Most processing software such as Deep Sky Stacker will have a means of subtracting bias and/or dark frames from the flat frames but it is worth checking with the software itself exactly how this is done. Non-calibrated flat frames will add noise!

As with bias frames, it is good practice to obtain flat frames at a similar temperature to that at which light frames are acquired.

How Many?

This is largely down to you but the more the merrier, to a point. The old adage was to obtain as many of each calibration frame as you do lights. This is largely unfounded.

For portable setups it is quite easy to have a library of dark and bias frames (though see above for DSLR darks) but flat frames really do need to be acquired each session. Even though it eats in to imaging time there is a value in obtaining a set of at least a dozen or so flats for each filter used. The slightest light leak can ruin a dark frame so it is safest to acquire dark frames after dark.

With permanent setups there is merit in leaving the imaging train unchanged from session to session such that flats can be re-used. This being the case flats may be obtained at times when imaging is not possible and 30 or so flats for each filter used would not be unreasonable.
LX200|ZS70|FSQ85|FLT110|Altair DF250RC|EQ6 Pro(Rowan Belt Mod)|ParamountMX
ATIK383L+/EFW2/OAG|Lodestar|Baader 36mm LRGBHaSIIOIII
Starlight Express SXVRH16/ONAG/FW|Lodestar X2|Baader 2" Filters
Starlight Express SXVRH814/ONAG/FW|LodestarX2|Baader 2" Filters
Lunt LS60PTBF1200|DMK41|Quark Chromo
Samyang 135mm f1.8

Offline Trow

  • New Member
  • *
  • Posts: 36
Re: Image Calibration Frames (Bias, Darks and Flats)
« Reply #1 on: 14:52:06, 23 April, 2014 »
I'm still a way off Deep Sky Imaging but I'd like to understand it all before trying to choose equipment.

So, I think I get Bias frames (capturing the way the pixels behave without the effects of time and temerature), Dark frames (capturing the way the pixels behave during the same length of exposure, and temperature, that the actual image is going to be captured), but, don't really understand form the description what flat frames are? As it is stated for both Bias and Dark frames that light is excluded, it implies that the shutter is open to light for the flat frames, and that an artificial light source is needed? I read somewhere that flats could be taken from the twilight sky - is this a good idea? It was also suggested that new flats are required for any change to the set up including camera orientation, change of focus, filters, etc.


Thanks,

Trow
Bresser 8" f4.3 Newt, CGEM, unbranded Chinese digital eyepiece.

Expanding Universe denier.

Offline chris.bailey

  • Administrator
  • Galactic Poster
  • *****
  • Posts: 10009
  • If at last you do succeed, don't try again
    • Rosemount Observatory
Re: Image Calibration Frames (Bias, Darks and Flats)
« Reply #2 on: 15:22:19, 23 April, 2014 »
Trow

Flats calibrate how the imaging system reacts to light so yes the shutter is open. The aim is to replicate any pixel to pixel variations in sensitivity that will be present in the light frames as well as to compensate for any light fall off owrards the corners due to vignetting and for any particles of dust that can cause largish darker halos due to their distance from the sensor. If you take flats you MUST take bias frames as the Flats need to be bias subtracted prior to being divided into the lights. Bias and Darks are additive so are subtracted but Flats are multiplicative so need to be divided. If you divide a flat into the lights that is not bias subtracted the maths does not work and the flat can tend to overcorrec.

Ideally flats would be taken afresh each time but if you are permanently mounted you can get away without doing flats all that often (I redo mine every 6 weeks or so) wheras Bias and Darks should last a lot longer (6 months). With my small refractors I use an EL panel for flats but with the bigger RC do twilight flats myself. The problem with twilight flats is you need to keep changing the duration as it gets darker so I tend to do 5 x 2 sec, 5 x 3 sec, 5 x 4 sec or similar so each are similarly exposed. For twilight flats you need to turn the tracking on the mount off so that any stars that do appear can be removed as outliers in stacking. I do find twilight flats a pain to do as my CCD has a physical shutter which leaves a shadow on exposures less than about 2 seconds. As its pretty light sensitive it needs to be quite dark to get a controlled flat signal but very quickly gets too dark.

Hope that all helps

Chris
LX200|ZS70|FSQ85|FLT110|Altair DF250RC|EQ6 Pro(Rowan Belt Mod)|ParamountMX
ATIK383L+/EFW2/OAG|Lodestar|Baader 36mm LRGBHaSIIOIII
Starlight Express SXVRH16/ONAG/FW|Lodestar X2|Baader 2" Filters
Starlight Express SXVRH814/ONAG/FW|LodestarX2|Baader 2" Filters
Lunt LS60PTBF1200|DMK41|Quark Chromo
Samyang 135mm f1.8

Offline Trow

  • New Member
  • *
  • Posts: 36
Re: Image Calibration Frames (Bias, Darks and Flats)
« Reply #3 on: 10:46:05, 30 May, 2014 »
Thanks for the comprehensive reply Chris.

Yes, I understand the principal now. Though, I guess there's no substitute for actually getting on and trying it for myself. I hope to take the plunge soon.

Thanks again.

Trow.
Bresser 8" f4.3 Newt, CGEM, unbranded Chinese digital eyepiece.

Expanding Universe denier.

Offline fwm891

  • Full Member
  • ***
  • Posts: 201
  • Retired Photographer
Re: Image Calibration Frames (Bias, Darks and Flats)
« Reply #4 on: 12:32:21, 20 September, 2014 »
Hi Chris,
Can you elaborate on your statement " Adjust exposure such that the histogram is no further than ½ way up the scale"....

Francis
Altair Astro 115 EDT/APO and 8 inch CF/RC, Avalon Linear mount, RoR and lots of stuff...
flickr: https://www.flickr.com/photos/97440028@N04/with/14450746185/

Offline chris.bailey

  • Administrator
  • Galactic Poster
  • *****
  • Posts: 10009
  • If at last you do succeed, don't try again
    • Rosemount Observatory
Re: Image Calibration Frames (Bias, Darks and Flats)
« Reply #5 on: 13:55:45, 20 September, 2014 »
Francis

Very general advice aimed more at DSLR imagers, the histogram on the rear of the DSLR should be no more that half way across. This can be a bit of a judgement call but it makes sure you stay in the linear portion of the device. Go too far to the right and the flat will tend to overcorrect. With a CCD I tend to keep the brightest part of the histogram below about 20,000 ADU on a 16 bit image, more often in the 16,000 area. Flats do take a bit of experimenting with particularly with systems where vignetting leads to very dark corners but erring on the side of underexposed seems to lead to less issues that making them too hot.

Chris

LX200|ZS70|FSQ85|FLT110|Altair DF250RC|EQ6 Pro(Rowan Belt Mod)|ParamountMX
ATIK383L+/EFW2/OAG|Lodestar|Baader 36mm LRGBHaSIIOIII
Starlight Express SXVRH16/ONAG/FW|Lodestar X2|Baader 2" Filters
Starlight Express SXVRH814/ONAG/FW|LodestarX2|Baader 2" Filters
Lunt LS60PTBF1200|DMK41|Quark Chromo
Samyang 135mm f1.8

Offline fwm891

  • Full Member
  • ***
  • Posts: 201
  • Retired Photographer
Re: Image Calibration Frames (Bias, Darks and Flats)
« Reply #6 on: 15:08:42, 20 September, 2014 »
Thanks Chris, Probably more than half my problem in that I've been working on around 40-48K ADU. Wait for the new EL panel to arrive and try with much lower ADU values...

Francis
Altair Astro 115 EDT/APO and 8 inch CF/RC, Avalon Linear mount, RoR and lots of stuff...
flickr: https://www.flickr.com/photos/97440028@N04/with/14450746185/

Offline chris.bailey

  • Administrator
  • Galactic Poster
  • *****
  • Posts: 10009
  • If at last you do succeed, don't try again
    • Rosemount Observatory
Re: Image Calibration Frames (Bias, Darks and Flats)
« Reply #7 on: 15:47:43, 20 September, 2014 »
Francis

Yes, 48k in my experience is too hot but every sensor is different. Much easier to experiment and fine tune them with an EL panel. Once you have a master flat it is worth trying to calibrate individual flats with the master. The end result should be a clean frame with no residual artefacts or shading.

Chris
LX200|ZS70|FSQ85|FLT110|Altair DF250RC|EQ6 Pro(Rowan Belt Mod)|ParamountMX
ATIK383L+/EFW2/OAG|Lodestar|Baader 36mm LRGBHaSIIOIII
Starlight Express SXVRH16/ONAG/FW|Lodestar X2|Baader 2" Filters
Starlight Express SXVRH814/ONAG/FW|LodestarX2|Baader 2" Filters
Lunt LS60PTBF1200|DMK41|Quark Chromo
Samyang 135mm f1.8

Offline Timothy Brown

  • Posts: 4
Re: Image Calibration Frames (Bias, Darks and Flats)
« Reply #8 on: 15:23:14, 08 January, 2015 »
Hello,

Apologies if this is a dumb question but having got Deep Sky Stacker running on my Linux PC under Wine, I came to process my first set of images and was somewhat confused by DSS wanting Dark Flat images for full calibration of the images.  I had the Bias, Darks and Flats available and am trying with just those however as my PC is old and slow so it's only managed to register a few light frames so far, I'm working with Canon EOS7D Raw data.   I have had a look through the DSS documentation but I can't find how one produces Dark Flats or whether they are really necessary.  Does anyone know about this?


Offline chris.bailey

  • Administrator
  • Galactic Poster
  • *****
  • Posts: 10009
  • If at last you do succeed, don't try again
    • Rosemount Observatory
Re: Image Calibration Frames (Bias, Darks and Flats)
« Reply #9 on: 16:38:14, 08 January, 2015 »
Timothy

Not a dumb question at all.

For any non-zero exposure (i.e. Bias), dark signal starts to accumulate. Flats need to be calibrated in the same way as light frames are so IDEALLY would have dark frames subtracted from them that are the same duration as the flats. In practice, flat frame exposures are very short (a few seconds) so the dark signal accumulation is very low. As such the flat dark approximates to a bias. Calibrating the flat frames with a bias frame has therefore become pretty normal. Note that flats MUST be bias subtracted or they will not properly calibrate the bias subtracted light frames.

Dark signal has a good linear correlation with elapsed time but little correlation with temperature. A few degrees difference between the temperature at which the dark and light are taken means there is no match between the dark signal in them. As such the use of dark frames for non set point cooled cameras is questionable. This mismatch can have significant impacts on the signal to noise ratio in the light frames and you may well find you end up with cleaner results if the dark frames are dropped altogether.

Lastly, don't skimp on the number of bias and flat frames taken. Both are quick and easy to acquire so there are few excuses not to take plenty. The old adage used to be to take as many calibration frames as you do lights but I would advise on doing more than this.

Chris

ps DSS does struggle a little with big DSLR files. There is a lot of processing going on under the hood so don't be too surprised if it takes a while.
LX200|ZS70|FSQ85|FLT110|Altair DF250RC|EQ6 Pro(Rowan Belt Mod)|ParamountMX
ATIK383L+/EFW2/OAG|Lodestar|Baader 36mm LRGBHaSIIOIII
Starlight Express SXVRH16/ONAG/FW|Lodestar X2|Baader 2" Filters
Starlight Express SXVRH814/ONAG/FW|LodestarX2|Baader 2" Filters
Lunt LS60PTBF1200|DMK41|Quark Chromo
Samyang 135mm f1.8

Offline Timothy Brown

  • Posts: 4
Re: Image Calibration Frames (Bias, Darks and Flats)
« Reply #10 on: 13:26:39, 09 January, 2015 »
Thanks for the explanation Chris, I think I'm starting to get this now.

Regards

Tinothy

 

ukbuysellRemote Imaging from AustraliaSharpSkyblank APTUKAI on Facebook
Powered by SMF 2.0.15 | SMF © 2006, Simple Machines LLC
DarkBreak by DzinerStudio. Theme modified by The UKAI Team

Page created in 0.325 seconds with 38 queries.
TinyPortal © 2005-2012