Author Topic: Quality of Bias and Darks  (Read 3522 times)

Offline steve71291

  • Newcomer
  • Posts: 9
Quality of Bias and Darks
« on: 2014 September 17 15:52:28 »
I have been having some problems with calibration. I have read many, many posts. One thing I keep seeing is the recommendation that one "take a good look at their darks and bias frames" or, more to the point, make sure their bias and darks are of "good quality". I thus far have not been able to find just how you do that. I am sure it is very simple, but how do I determine the quality of my superbias or my master dark?

Thanks,

Steve

Offline mads0100

  • PixInsight Addict
  • ***
  • Posts: 116
Re: Quality of Bias and Darks
« Reply #1 on: 2014 September 18 10:53:21 »
Just do an STF and take a look at them.  Then, take a look at your processed images.  You should be able to correlate what you see in the darks/bias/flats to your processed images and see if it's related to your calibration frames or if it's something related to integration.


Offline steven_usa

  • Member
  • *
  • Posts: 74
Re: Quality of Bias and Darks
« Reply #2 on: 2014 September 18 19:49:45 »
I'm not sure if there is a "quality" metric for bias and darks.  I think you either did them correctly, or you didn't.   Both are done with the lens cap on.  Bias are taken at fastest exposure your camera supports (e.g. mine is 1/4000 for a Canon 450D).  Darks use the exposure duration of your lights (e.g. 10 seconds, 30 seconds, 5 minutes, 10 minutes, etc. whatever they are).  Both bias and darks, I'm told, are taken at the same ISO as your lights (ISO400, ISO800, ISO100, again whatever they are).

From there, then it is about quantity.  Seems you may want at least 20 of both of those, 50 being better.  The bias are easy to collect, the darks can be hard since in addition to all of the above, the darks should match the temperature of your lights.  That's where it can get a little subjective: you might be 1 degree off, 3 degrees, 5 degrees.  How far you can bend this rule might depend on the quantity and what software you end up using them with (I lack the experience here to really say).

Flats is where you do need some measure of quality, which basically comes down to getting the exposure right.  Basically, as I understand it, you want the histogram of the light to be somewhere in the middle (just NOT on the left side).  (1) The ISO matches your Lights, (2) the sensor should the same focus and orientation as your lights were taken.  Then (3) finally, the sensor is also looking at a "flat" field like a blue sky, or a diffused light, an ipad white screen, etc.

I think to "see" the bias, dark, flat you'll need to stretch that raw image (but don't save it and used it in the stretch form, just use the stretch to preview the content).  I don't have an example of a stretched dark... But here are examples of my stretched Flat and Bias, using a Canon 450D DSLR 12mp.  Remember this is not the what these images look like normally, this is a screenshot of them while doing Auto STF in PI.




Offline mads0100

  • PixInsight Addict
  • ***
  • Posts: 116
Re: Quality of Bias and Darks
« Reply #3 on: 2014 September 19 07:48:29 »
One thing with DSLRs is to look for light leak through the prism. I know it sounds crazy, but my D800 has a shutter that covers the eyepiece for that reason. 


Offline steven_usa

  • Member
  • *
  • Posts: 74
Re: Quality of Bias and Darks
« Reply #4 on: 2014 September 20 20:23:03 »
Are you meaning light reflected back thru the eye viewfinder?  I have my camera inside of a cooler, while attached to the scope.  When I open the cooler, I've noticed that the LCD screen is often on.  So it occurred to me, that light from the LCD is bouncing around inside the box and likely going inside the viewfinder.   Normally, when the shutter opens, there is a spring/hinge (whatever) and hence when using the viewfinder normally (daytime photos) you see it go black/dark at the time of taking a picture.   But that closure might not be perfect, and little specs of light still might scatter.    So I've gone to covering the viewfinder window (and cover the LCD just as an extra precaution) before I close up my cooler.     Without a cooler, then it would reflections of light either from stars themselves, flashlights, or even streetlights.  Either way, it's probably still overkill :)

Offline georg.viehoever

  • PTeam Member
  • PixInsight Jedi Master
  • ******
  • Posts: 2132
Re: Quality of Bias and Darks
« Reply #5 on: 2014 September 21 03:10:07 »
Light through the viewfinder is a real problem. You can proof it by shooting darks with and without covered viewfinder. A simple piece of black tape solves the problem.
Georg
Georg (6 inch Newton, unmodified Canon EOS40D+80D, unguided EQ5 mount)

Offline MikeOates

  • PixInsight Addict
  • ***
  • Posts: 278
Re: Quality of Bias and Darks
« Reply #6 on: 2014 September 25 05:02:13 »
Steven,

You really do not want the camera LCD to be lit at all, this adds heat and will effect your images just by the increased heat alone, i.e extra noise, which can show it'sself as patches of light in your dark frames. There should be setting to disable this there is with my Canon 500D.

Mike