Comet Alignment & Comet Subtraction Help

[brent1123]

Good Evening,

I am attempting to use the Comet Alignment Process to churn through my capture of Comet 46P from this past week. I have calibrated my frames (Darks, Bias, Flats), Star Aligned them, successfully Comet Aligned them and have produced a good Integration image of the comet. With large-scale Pixel Rejection used, the background stars are all but removed during Integration.

My difficulty is when trying to use this same process to subtract the Integrated Comet image from each of the individual subs for later integration into a starfield.

 - I have tried subtracting from the Star Aligned subs, the Comet Aligned subs, and even the unaligned (calibrated) subs

 - In each instance I have also selected the first and last sub to ensure the comet location is marked. The best result I have received is a slowly disappearing comet (see video link below), which still makes no sense to me.

 - I have attempted Comet Subtraction using both LinearFit and Normalization options turned on and off. At best, this results in a dark background with the comet still slowly subtracting as seen in the video below

If the Integrated comet was not correctly aligning to the subs, I would see dark patches as the comet subtracted from background space, but I do not see that here, instead the comet seems to be incrementally subtracted from each sub in successive amounts until on the last few its properly subtracted. For something which should be straightforward this is maddening and I have no idea how to correct this. Can anyone offer processing advice?


https://www.youtube.com/watch?v=a8bG2ayLLVM - short Blink Process video of the comet slowly disappearing



Equipment is a ZWO ASI-1600MM-P, Luminance Filter, and AT72ED OTA at 60" exposure, Gain: 200, Offset: 50

[johnpane]

One possibility is that your subframes are progressively darker as time goes on (due to fading twilight or decreasing moon interference). I agree with avoiding normalization during the process of creating star-only and comet-only integrations, but you may need to perform a normalization step prior to all of this.

I have accomplished this by putting the subframes in an image container and applying this PixelMath: $T - median( $T ) + k, where k is a small constant that prevents the background from clipping.

John

[brent1123]

John,

Thank you for the suggestion. I tried running your PixelMath string using 0.18 as k (a value chosen to minimize any brightening or darkening of the subs), but I do not think this solved the issue. Subtracting the resulting Comet Integration (the Integration itself looked fine as always) from the Star Aligned frames (the ones originally loaded into Comet Align which were normalized in PixelMath) still showed this slowly disappearing comet. Attempting this again with normalization turned off (in the process that is, PixelMath was still applied) resulted in the same as well as significantly darker frames.

In both attempts I made sure to verify the comet was still marked in the subframes. I feel as though I may be missing a step somewhere but this seems like it should be a simple process

Brent

One quick note is that i have been following the 2 tutorials below and as far as I can tell, my processing steps are identical (though the Comet Alignment process has updated since these were posted):

 - http://pixinsight.com.ar/en/info/processing-examples/38/lovejoy-comet.html

 - https://pixinsight.com/forum/index.php?topic=8161

[brent1123]

I believe I have mostly solved this issue - I found that when using the Comet Alignment process, I had to keep the LFit option turned off while keeping Normalize on. This still cause some minor fluctuation in subtraction with some subs, such as some having good subtraction and others being slightly under or over-corrected, but this could easily be explained by changes in Moonlight, LP, and differences in altitude (no electronic focuser here, have to do all my subs for each filter all at once)