›

Use of the Optimize Dark Frame Option

Stickfarm8
in PixInsight
In lieu of what I think I have learned regarding calibration frames, I am evaluating my imaging workflow. In hindsight, I was pretty careless and not consistent in taking the necessary calibration frames because I did not appreciate the importance of them. Fundamentals has been an eye-opening experience to many of my poor habits.

My primary imaging camera currently is a Canon EOS Ra. Of course it does not have a temperature controlled sensor. The weather here in Kentucky is highly "variable" as far as night temperatures are concerned. If I have understood what Adam explained in the Fundamentals video 04 on calibration, the dark current could vary - even if the darks are taken at the same exposure time as the light frames - due to variances in the sensor temperature. By using the Optimize Dark Frame option when calibrating, this would be handled properly.

In the past, I would take the dark flats after imaging. So the sensor would certainly be warmer than it was at the start of imaging.

So, my question is this. As long as I use the Optimize Dark Frame option when processing the light frames, is there any downside to preparing, in advance of imaging, master darks and master bias' for different isos that I am likely going to use in the future and avoid taking the time to capture dark flats after image capture ?

Taking the light flats does not take much time at all. But not taking the time to capture the dark frames after imaging, allows me more imaging time before I have to throw the towel in and get into bed during the short summer shooting times - lol.

Any thoughts are appreciated.

Rich

Comments

  • BlockHeadBlockHead
    In general you are correct. Optimizing a dark to better match light frames that have variations in temperature is a good move. However, it comes with so many caveats. The main takeaway from my lessons is that the use of dark optimization comes with a host of pitfalls. Use and forget about it isn't part of this more advanced method since it requires more attention to the results... not less. More importantly, you have to look at the results and be able to determine (in various ways) that they really are optimal.

    The Canon Ra (I have one) is really not an astronomical camera. It simply doesn't have the electronic signatures that other cameras (dedicated to the pursuit) have. So it requires more experimentation on the user's part which for beginners in general is always a source of potential frustration. For example, for the Canon NOT using a Bias frame may be the right answer... which means optimization of the darks isn't even possible. See how complicated this becomes? The means matching darks is really the best way.

    "But wait", you say, "the temperature is changing..." Yes...but the question is what variation in temperature requires new darks (or optimization)? What variation do you observe if you given your camera time to equilibrate ? (The fact you need to do this is a "thing".) What kind of variation do you see when downloading images? If you are taking 20 second exposures (as I did) for a comet... you know that the sensor can heat up if the exposures are short enough?

    On and on it goes.... 

    Typically though the answer is 5 degrees F. If you are within 5F during your session- you can calibrate them with the same darks. Am I really correct? Well, you the user now need to do the experiment. You can calibrate with darks you take that match the temperature... matching these across all of your pictures. And then choose just one dark frame...calibrate with it. Do everything else the same and compare the masters. Can you tell the difference? Can you measure the difference? (Statistics Process) Is the measured difference meaningful?

    People think that DLSR imagers are a good entry level way to do astrophotography. I disagree in the sense that more understanding, not less, is required for dealing with the data. People confuse the ease of acquiring the data with processing it. 

    -the Blockhead
  • Stickfarm8
    edited July 2024
    Thanks for the reply......a lot to think about.

    I started with the DLSR simply because I had one - a 5D Mk3. I knew of some of the limitations using a DSLR. The Ra seemed liked a good next step in lieu of getting the glass filter in front of the sensor replaced on the Mk3.

    I was certainly not aware that NOT using a Bias frame might be appropriate. 

    I was mistaken in my WBPP post regarding the model of the ZWO camera I had acquired. I have the ASI120 - but its the guide camera replacement for the Orion guide camera that had stopped working. The imaging camera I had acquired is the ASI071MC Pro.  I had a lot of issues when I tried to use it with APT imaging software back in 2020/21.

    When I restarted into the hobby back in May, I started using the NINA imaging software and I have come like it. Everything I imaged in May/early June was done with the Ra. While the ZWO has an APS-C sensor - if I can master using the ZWO properly -  its probably a better choice long-term.  I am going to head in that direction based on what I think you are saying here. NINA is limited on its use of DSLRs - for example getting a histogram. 

    More to think about.....

    People confuse the ease of acquiring the data with processing it. Amen - how much I am becoming aware of that.

    Thanks !

    Rich

  • Stickfarm8
    edited July 2024
    Yesterday I got the ASI071 out of its case and fired it up in NINA to see if it still worked. After a little fumbling with getting the drivers right, it fired up and cooled down quite nicely to 0 C.

    Gave the ASI a workout over night. Using the NINA Sequencer I had it run thru, at what I think will be the best starting gain/offset (94 (unity) and 65 offset) based on a video by Cuiv The Lazy Geek (lol...), a who series of darks - almost 12 hours worth without any glitches at a progression of exposure times. May be not useful if I find a different gain is more appropriate - but I'm confident at least the unit appears to be working. A looked at a few of the darks - not much see though. 

    According to the documentation, the sensor in this camera did not seem to have the big drop in read noise with higher gains as the unit that Cuiv was using. 

    While I'm waiting for some earlier dark times to get setup again, I may pick another lower and higher gain, and run through the Sequencer some more just to have them handy later. Its only disk space - lol.

    I also started reading a copy of Bracken that the family had given me for Christmas a few years back to get a better understanding of the sensor at least that he seems to cover at the beginning of his book.

    Thanks for the subtle push to use the better tool Adam. 
Sign In or Register to comment.