Spacial beat frequency in integration

[gottsch]

Hello All,

Looking at WBPP and ImageIntegration I checked the reject high maps and find a spacial (beat frequency??) pattern. Not sure if it comes out in the screen copies.

Not sure if you can see it. Squint the eye a bit. Do you see the square pattern? What is that? Seeing it in the reject high maps, does it indicate of similar pattern in the image (just not visible at that stage)??

Here is a 100% zoom (for this one you have to lean back a bit and look at full screen):

First ran WBPP then SubFrameSelector to throw out some bad ones, then plain ImageIntegration. No drizzle.

This is from a 72 x 2min stack from asi2600mc 10" f4 Newtonian dither every frame from lightpolluted location.

Any ideas?

Thanks & Clear Skies,
Gert

 

[ngc1535]

Use a pedestal when calibrating your images.
You asked for ideas...I am permitted to give them Gert.
My idea is that you are using a filter in front of your sensor and that your data has zeros (oversubtraction during calibration) and that increasing the pedestal will solve the issue.
Please confirm these are ideas in the form you asked for them.
Thanks and Clear Skies,
adam

 

[gottsch]

Hello Adam,

My idea is that you are using a filter in front of your sensor and that your data has zeros

Thank you for sharing your thoughts. I mentioned this above "...from light polluted location..." My apartment place has Bortle 8 and SQM 18.1mag (As per frame analysis from Astap), so taking non-narrowband deepsky images is a futile undertaking. I don't know why I even tried. :-( Looking at the stats of raw, dark, and calibrated frames I don't see zeros.

Now of course a possible thought is to just stop imaging under these abhorrent conditions. Kind of along garbage-in-garbage-out. Well, I still like the challenge and maybe I can work hard to prevent some of the ingoing garbage from coming out.

Just want to share the image (crop) below. Notice the 'mottling'? seems to coincide with the pattern seen in the reject maps. Probably just pulling too much data out of from too close to the noise floor.

Cheers,
Gert

PS watched your comet video. Awesome stuff. I barely saw the comet from the apartment parking lot with 10x50 binoculars. Not sense taking a picture.

 

[ngc1535]

Ok.

1. Make certain you dark is fresh.
2. Dither your data as much as possible... every frame but a significant number of pixels (10+).

Basically I think you are seeing the "warm" pixels that are not being perfectly subtracted.Then, due to interpolation effects, you get a kind of correlated fix pattern of the residuals.

-adam