›

Simple question, (I think?) regarding role of Bias frames

Anthony Quintile
edited September 2021 in PixInsight
Please correct my statements below if they are not correct:

-Darks include Bias signal, therefore...
-Only Darks, (and calibrated Flats/Master Flat created with Flat Darks), not Bias, are required to create a Master Dark to calibrate Light frames unless...
---You are attempting to scale Darks taken at an exposure different from your Lights.
---You are calibrating Flats with Bias, which can be done with Flat Darks more reliably across different camera platforms anyway.

Basically I want to be sure that there is not some role in calibrating a Master Dark that is played by separate Bias/Master Bias, except as noted for scaling. In order to minimise potential issues, I am fine with keeping, and occasionally updating, Dark libraries for different exposure times, and shooting Flat Darks each time I shoot Flats. I am shooting with an ASI2600MC Pro CMOS camera if this information is helpful.

Is this correct?

Comments

  • BlockHeadBlockHead
    I think you have the right idea...but I would state it differently.

    -Only darks are required to make a Master Dark. (They are just combined in ImageIntegration)

    -If you are attempting to scale a MasterDark in order to calibrate (light) data taken at a different exposure time you need a Bias Frame to do the trick.

    -If your sensor does not have ampglow or other sources of electrical behaviors, you can use a MasterBias to calibrate all Flats (and create a Master Flat) of ANY reasonable exposure time. This is a wonderful thing. It saves time, saves on files... 

    "....Flat Darks more reliably across different platforms anyway."

    The problem is a single camera manufacturer to be honest. There are a few cameras, popular ones, that are the problem. The vast majority of sensors/cameras are well-behaved and do not require matching darks for calibrating Flats. A single MasterBias is usually good enough. BUT- these cameras has really contaminated the market in the sense we have to discuss this other need. 

    Matching Darks always is of course going to work- but it is entirely unnecessary for most cameras. I just had to put that language there so the people with the cameras for which it does not work do not have their head's explode in frustration. So I suggest you can always match darks and get the right answer- but once you know the kind of camera you have, you may not need to. That is... once you have left the nest... not a beginner and understand what is going on.

    Your particular camera? I can't keep it straight which ones are the funky ones. That doesn't ring a bell as a funky one...so I suspect you are in the normal group with just using a Master Bias to calibrate your flats. 

    -the Blockhead



  • Anthony Quintile
    But there is no issue with using Flat Darks instead of Bias to calibrate Flats, right?

    I takes ~5 minutes extra to shoot Flat Darks after shooting Flats, so not a huge time savings, and I have heard (so much hearsay, but nonetheless...), that Bias shot on some CMOS cameras can be suspect.
  • Anthony Quintile
    And as a side note, there seems like there may be a bias, (lowercase "b", not the calibration frame), amongst those who can afford fancy CCD cameras against the idiosyncrasies of some of the newer affordable CMOS cameras.

    When I go through the tutorials on all of the extra steps necessary to calibrate out bad columns, (create defect lists, save that somewhere where you can find it, remember to apply it every time...), I question which type of sensors are "well behaved", and at what extra cost.

    Maybe I am missing something, or maybe I just lack sophisticated taste... ;-P
  • Wade Hilmo
    Whether flat darks are easy or difficult depends on your workflow.

    If you use a flat panel, with identical exposures for each raw flat, then flat darks are trivial.  I do sky flats, with adjustments to the exposure time between each raw flat.  It would be completely impractical to take flat darks under those conditions.

    The good news is that I've always found that calibrating my raw flats with a master bias works great, even though I used to use one of the "funky" ones (an ASI1600MM-cool).

    It's really difficult to get past some of the misinformation out there around certain topics, like this one.

    The difference between a master flat made from flat darks, and one a master bias, is the dark current associated with the exposure time of the flat dark.  Unless you are taking really long exposures for your flats, that dark current is going to be trivial.  The other factor is that the signal in a flat is really high, which further reduces the impact of dark current.

    If you ever want to determine whether you *need* flat darks for your own system, it can be easily determined objectively.  To do this, make a master flat dark and a master bias.  Subtract the master bias from the master flat dark.  Whatever is left after the subtraction is the dark current.  If there is no significant dark current in that result, then you can just use a master bias to calibrate your flats.

    I think that some of the discussion around flat darks vs bias comes from an unfortunate default setting in PixInsight.  Specifically, PixInsight defaults to "optimize" checked in the ImageCalibration process.  This setting will wreak havoc with most CMOS cameras, including the ASI1600 that I had.  To avoid this, you need to match the dark frame exposure time with the light frames that you are calibrating.

    I suspect that this thinking extended to calibrating flat frames, even though calibrating flats with a master bias almost always works.  And I say "almost always" here because I've not tried every camera out there.  But really, you would need a *lot* of dark current to need a flat dark for an exposure as short as a typical flat frame.
Sign In or Register to comment.