Stacking out interference pattern? Pixel rejection?

[footbag]

I had an issue with my STF-8300 the other night where there was an interference pattern on one side of the image.  It began after the meridian flip on my first night, so it's in about 1/3 of my frames.  Since it's not in a majority of the frames, is there a specific type of combination method or pixel rejection that I can use to eliminate it from the final stack?  Since the pattern isn't in my darks or bias, it doesn't calibrate out on it's own. 

Is it possible to save the data?  Or do I need to re-shoot or crop it out?


Link to individual sub to see the issue...
https://www.dropbox.com/s/3bxi5e308pabhgi/pelican-1_1800sec_1x1_HA_frame7.jpg

[Carlos Milovic]

Most interference patterns may be reduced with noise reduction techniques in the Fourier (frequency) domain. In this case, the pattern seems to have a small frequency, so it may be difficult to follow that procedure without leaving some artifacts in the image. I would try the CanonBandingScript with your image rotated 90º.

If the pattern vary across the frames, a normal integration with mild or aggressive rejection parameters could be enough.

[MikeOates]

This could also be a flat field frame either not being used, or the camera was also rotated during the meridian flip and you did not take more flats with the camera rotated.

Normally the camera is not rotated at a meridian flip so just one set of flats can be used.

Mike

[footbag]

Most of it stacked out when I added the second nights data.  It's still there slightly, but barely noticeable.  I think that it'll disappear with one more night.

The camera didn't rotate, but the object did during the flip.  I think that helped calibrate some of it out. 

 

[MikeOates]

As the camera did not rotate with respect to the telescope, then one set of flats will work both before and after the flip. But, are you taking flats on each night? this is so imporant for the flats to work properly as conditions change from night to night, different focusing, dust on the sensor, sensor window, filters etc can move.  If your using filters, you need a set of flats from each filter.

This means you need to calibrate each set of nights subs seperately and then during integration all the subs from all the nights are integrated together.

Mike

[footbag]

Hi Mike,

I left the telescope setup.  So one set of flats was all I needed. 

I think this pattern was caused by a bad USB cable.  It didn't occur in my darks, bias or flats.  Just half of my subs from one night.  Since it's not in all of the subs, only 1/3, I think it's going to completely disappear with my next nights worth of data.  As it is, it's only visible at the very end of processing.  Even then it's minimal.