Why am I getting such horrible walking noise?

[StuartT]

I dithered by 7px every 2 frames. 

What's going wrong?

M65_and_M66.fits

[wimvb]

The general recommendation is to dither at least 12 pixels. If you align the raw subs (without calibration) and stack them without pixel rejection, you should see a random pattern in the hot pixels. If ototh, hot pixels are in more or less a straight line, then your dither steps are too small.

[GalaxyGael]

That looks like a case for bigger dithering steps as Win said. If that image if the imaging focal length, and you are using an OAG, you can double the dither step in your software or change the scale option in PhD2 setting to 2.0, depending on whether your software has an inbuilt maximum for dithering (sharpcap wont go above 30).

It looks like the random dithering directions were biased toward top left to lower right (just eyeballing). The random selection usually allows a lot more variation in the dithering, but bigger steps might sort it out for you. Are you using OAG or guidescope option at that FL?

[StuartT]

2 hours ago, wimvb said:

The general recommendation is to dither at least 12 pixels. If you align the raw subs (without calibration) and stack them without pixel rejection, you should see a random pattern in the hot pixels. If ototh, hot pixels are in more or less a straight line, then your dither steps are too small.

 

2 hours ago, GalaxyGael said:

That looks like a case for bigger dithering steps as Win said. If that image if the imaging focal length, and you are using an OAG, you can double the dither step in your software or change the scale option in PhD2 setting to 2.0, depending on whether your software has an inbuilt maximum for dithering (sharpcap wont go above 30).

It looks like the random dithering directions were biased toward top left to lower right (just eyeballing). The random selection usually allows a lot more variation in the dithering, but bigger steps might sort it out for you. Are you using OAG or guidescope option at that FL?

Thanks both. In fact I am not using any guiding here (my OAG has not been delivered yet), so this was just using NINA's manual dither. Nevertheless, the direction should be random, not top left lower right, as you say. Blinking them seems to confirm that they mostly move along a top left / bottom right diagonal 

[Adam J]

9 hours ago, StuartT said:

I dithered by 7px every 2 frames. 

What's going wrong?

M65_and_M66.fits 34.63 MB · 19 downloads

Are you dithering in both axis? There is an option for RA only that could potentially cause issues. 

Adam 

[Adam J]

1 hour ago, StuartT said:

 

Thanks both. In fact I am not using any guiding here (my OAG has not been delivered yet), so this was just using NINA's manual dither. Nevertheless, the direction should be random, not top left lower right, as you say. Blinking them seems to confirm that they mostly move along a top left / bottom right diagonal 

I would say that this is likely your problem as there will be a consistent direction of drift on top of the dither. 

[wimvb]

If you have backlash, you need to add that to any dither pulse. Otherwise dithering has very little effect.

Edited April 26, 2022 by wimvb

[StuartT]

Thanks all. I shall look at increasing the dither then.

Here is what I managed to make of the above FITS with Pixinsight. Managed to remove the walking noise, but may have lost a little signal too in the process.

This is only half an hour of integration time though, so is probably about as good as can be expected.

 

Edited April 26, 2022 by StuartT