›

Making a Defect Map - Hot Pixel Count

Stickfarm8
in PixInsight
Adam,

Happy Holidays sir!

After rewatching the three videos in Fundamentals, Cosmetic Correction, I wanted to build a defect map for the ASI2600 MC Pro camera I bought in late summer as you did in the last third of the second video (the longest one - 56 minutes). 

I have not seen any evidence in my calibrated images, of columns or rows similar to your example in the video, but I wanted to check the camera, and map if any were present, the way you demonstrated.  I assumed that a combined image of all the bias files for taken at a given gain and offset, rather than just a single bias image would be best.

Here are the steps I took.
1. Placed the Bias images (125 gain/45 offset) in an image container.
2. Opened up Image Integration, selected "no normalization", weighting images all equal to 1, generate an output image, and did not select any pixel rejection.
The process completed successfully.
In examining the grey image after applying the STF, there were no signs of light or dark columns or rows. Just an image of varying extremely low values when using the readout mode.

I then used the same process to combine the dark images I had taken for the same gain/offset (125/45, 0 C) taken at 4 minute exposures back in August. Same result  - no visible bars. But using cosmetic correction on the combined dark file image, at a 4 sigma, there were a lot of hot pixels -  more than 480,000 hot pixels  - 1.8%. That is a lot more than what you showed in your video, even accounting for the camera sensor being 26 MP instead of 16 MP. 

Repeating this with 2 minute dark exposures the hot pixel count dropped to 219,000 hot pixels (0.8%).

I did some searching in various forums, without coming up with concrete numbers. From your experience is this about typical or do I need to have a conversation with ZWO ?

I would appreciate your thoughts Adam.

Rich


Comments

  • BlockHeadBlockHead
    CMOS and CCD cameras are different. I would expect the CMOS sensor to have more variation from pixel to pixel since the pixels each have electronics controlling things. With CCD..it was more the substrate itself. 
    So I would not compare the sensor I demonstrated to yours. 

    The defect map is really means for bad columns/rows, cluster defects, and other maladies. If you do not see these items... I do not think you need to worry about a defect map.

    -the Blockhead
  • Stickfarm8
    Adam,

    Thanks for the reply. 

    I was mainly concerned about the number of hot pixels. I did not realize (or missed it if you mentioned it) that the camera sensor you were referring to was a CCD. I have really come to enjoy imaging with the ASI2600 - which I bought after our discussion on the Canon Ra back in the summer. 

    I've gotten into narrowband imaging - still climbing the learning curve on those. I'm going to sign up for your narrow band series next - as soon as I finish re-watching some of the fundamentals. I've got a lot of images to process and now that imaging nights are few and far between here in KY with winter - I hope to catch up.

    Your workflow video on processing the Orion nebula was fantastic. I've been waiting to have the rising of Orion early enough to start capturing it. I was able to get some broadband images last weekend - but that is all so far.

    Thanks,

    Rich
  • BlockHeadBlockHead
    At the time I made the video... most of the market WAS just CCD!
    I think you will like the NB course. Lots of good stuff in there. 
    -the Blockhead
Sign In or Register to comment.