Walking noise even when dithering

[matt_baker]

I have a bit of a strange issue.

I've given up using darks with a DSLR since the temperature varies by 3-4 degrees throughout the night, so there's no use in darks. Instead I've just been dithering by the max setting of 5 in APT and using my superbias as a replacement for my dark.

In WBPP in pixinsight, if I check the calibrate dark frames, it skews my flat and makes the image look awful but there's no walking noise. However if I uncheck it, my flat frame gets calibrated just fine but it introduces walking noise to the image.

Is there something I'm doing wrong?

Here's an example of the noise stretched out

[alacant]

36 minutes ago, matt_baker said:

superbias as a replacement for my dark.

Hi

Keep it simple? Why not use the bias as a bias? Lose reference to dark frames altogether. Subtract the bias from both the light and flat frames then make a master flat frame to calibrate the light frames. Stack using an algorithm such as sigma clip. Experiment with the sigma high and low values which best match your camera. E.g. my 700d produces the least noise with low 5, high 2 whilst the same algorithm with 4-3 is awful. This is with Siril; I think the implementation of the rejection varies between apps, so YMMV.

HTH

Edited May 16, 2020 by alacant

[matt_baker]

I've ended up trying a plethora of different rejection algorithms and sigma values, none of which have fixed the issue. I also put my bias as bias. I did notice something odd however.

The noise seems to be like it's orbiting a central point, like star trails.

Here's an overstretched example:

Note how the streaks aren't in a constant direction like typical walking noise

[wimvb]

Put all the calibrated subs in blink and look how the stars move, from first sub to last. Is there a rotational movement? If so, then your dithering isn't enough. Star movement should be random if you dither.

What is the result of star alignment? If you see an increased rotation of subs, you have field rotation due to polar misalignment. Again, dithering should cause random movement (or a spiraling movement, depending on your dither settings).

To remove bad pixels, if you can't use darks, apply cosmetic correction during the calibration process. Set a fairly strong rejection for bright pixels. 

[matt_baker]

I've blinked the calibrated and registered frames and the stars are completely fine, no rotation whatsoever, although it was slightly rotated from a separate night of imaging.

Surely that would have been dealt with when they got registered. 

I guess I'm just not dithering enough. I have got a beefy ST102 as a guidescope and it's only 150mm out of my 130PDS at 500mm, and the scale must be messing with the settings, which makes it not move enough?

Here's the rejection high and you can clearly see the spiral

[matt_baker]

I've now taken all the frames and divided them by 2 to get the mean and safe to say, there is a spiral

Can be seen here: