›

Image Cal inconsistency...

Khushrow Machhi
in PixInsight
When I do an image cal typically on more than 20 frames in some cases there will be 2-3 frames where the dark frame is incorrectly subtracted.  I am using a master dark with the exact same time as my light frames, and this happens only when I have "Optimize" checked.  When I look at the process console I notice for these incorrectly calibrated frames that the weighting factor is way off sometimes it is 0, other times it is more than 1.5, whereas on the correctly calibrated frames the weighting factor are all typically between 0.95 and 1.03.  I learned to look at this from Adam's tutorial, very helpful. If I uncheck "Optimize" then all frames calibrate fine.  I would like to understand why I get these incorrectly calibrated frames when "Optimize" is checked.

Thanks,
Khushrow

Comments

  • BlockHeadBlockHead
    Indeed, if you have matching dark frames it is best NOT to optimize. It can get things wrong in terms of noise evaluations- because it is the noise evaluation that determines the scaling factors!! So turn off optimize and no scaling factor errors to mess up the weighting of your images.

    (If I ever said differently than the above, please let me know...I do not want to lead anyone down the wrong path.)

    -the Blockhead
  • Khushrow Machhi
    The fundamental series is fantastic!  You did not mislead in any way.  I had optimize turned on and was surprised to see a few frames with an amp glow (I use a CMOS ASI183MM-pro camera and the amp glow even with 90s subs is substantial) and without your tutorial would never have have known why.  Looking at the process console as you pointed in the calibration section showed incorrect weighting factors.  Would you know if it is an artifact of PI where ~10% of the frames do not calibrate correctly when optimize is checke?

    Thanks, Khushrow 
  • BlockHeadBlockHead
    HI Khushrow,

    Without looking at the data, I couldn't say if there is a particular reason. CMOS cameras have very little dark current. It is likely that the math for scaling the dark ends up with very small values which will mess up the calibration. Basically... there is likely nothing to scale in terms of dark current. The amp glow and other things need to be matched in exposure time. The next section I am producing at WBPP I hope will clear all of this up.
    -the Blockhead
Sign In or Register to comment.