Persistent walking noise

[TheMan]

I own a zwo 294 mc pro, a 72 ed f/5.6 refractor and a skywatcher az gti on eq mode. For the past 3 nights some of my integrated images have walking noise. I have tried different stacking methods in pixinsight and in deep sky stacker yet the same result comes back to haunt me. Does anyone know how I can fix this? I use extreme dithering in sgp for anyone wondering.

No darks in the integrated image

[alan potts]

I had something a bit like this from my Canon 40D, and found that it was me having ticked an incorrect box in DSS, I changed the darks to Median setting and it appeared to have solved it. I was told dither was the way to stop it without darks but have not been able to explore this yet due to weather.

Alan

[wimvb]

Have you tried cosmetic correction in PixInsight? Then while stacking, you can set an aggressive clipping factor. What exposure time did you use? Your image looks under exposed.

[TheMan]

9 minutes ago, wimvb said:

What exposure time did you use? Your image looks under exposed.

I used 120sec exposures. For some reason though that sequence was at gain 10 offset 10. That is a possible issue

10 minutes ago, wimvb said:

Have you tried cosmetic correction in PixInsight?

I'll give that a go.

[Whirlwind]

Fixed pattern noise is difficult to post process out because it isn't random per se so it is difficult for software to identify it (it appears more like 'real' data).

Dithering is meant to resolve this because it adds a random shift in to each image so then the fixed pattern noise then becomes more apparent to the software as the noise it is.

The noise in the fixed pattern arises in the read noise so can be more prevalent in CMOS cameras because of the low read noise and the tendency to use very short exposures.  Increasing exposure durations so that thermal noise overwhelms the read noise would be one way to overcome this.  You could also image with the camera slightly warmer to increase the thermal noise a bit.

For this image it may be worth checking that your darks don't need updating if you are using them.

Also try blinking you images to check that dithering is actually working as expected.  If you have a lot of backlash in the mount then you may think you are dithering but in one direction it isn't and using up the backlash in the mount.  The 'rain pattern' might suggest dithering isn't working in one direction.

[TheMan]

3 hours ago, Whirlwind said:

you may think you are dithering but in one direction it isn't and using up the backlash in the mount.

If that is the case, how would I cancel out the backlash during dithering?

 

[wimvb]

38 minutes ago, TheMan said:

If that is the case, how would I cancel out the backlash during dithering?

 

Make sure that all the gears are optimised (trade off between smoothness and backlash). Then use a large enough dither step so that the mount moves past the backlash. You can also dither in a spiraling path: ra+ dec+ ra- ra- dec- dec- ra+ ra+ ra+ dec+ dec+ dec+ etc. This sequence traces an outward spiraling dither pattern.

[Whirlwind]

I think it is better to confirm this first.  Have you blinked the sequence of images without a star correction to see whether there is any movement?

I am unsure whether SGP can manage the dither to the above degree of requirements, but increasing the dithering pixel step size/duration enough to wind in the backlash should help.  Are you guiding at all or is it very short subs stacked?

Sometimes placing a slight misalignment in the balance that works against the backlash can also help - but is dependent on where you are imaging.  Effectively the slight moment on one side acts to 'press against' the backlash.

[TheMan]

8 minutes ago, Whirlwind said:

I think it is better to confirm this first.  Have you blinked the sequence of images without a star correction to see whether there is any movement?

Are you guiding at all or is it very short subs stacked?

I did blink the files and there was movement, whether from backlash, dithering or flexure I do not know, but there is significant movement. I used 120s subs guided.

Funny thing is that I didn't get this issue before, it only started up recently. I might make a new profile in sgp to see if that has something to do with it.

[Whirlwind]

It might help if you posted the video here as it may help in analysing if there is any issues.

Have you changed the settings recently on the camera (e.g. gain / cooling etc?)

[TheMan]

23 minutes ago, Whirlwind said:

Have you changed the settings recently on the camera (e.g. gain / cooling etc?)

I don't believe I have changed anything, but i'll definitely try resetting everything to default and making a new profile. Also is -10 on the camera too low/high for good results?

23 minutes ago, Whirlwind said:

It might help if you posted the video here

I don't quite understand what you mean with that part.

Edited May 17, 2019 by TheMan

[Whirlwind]

58 minutes ago, TheMan said:

I don't believe I have changed anything, but i'll definitely try resetting everything to default and making a new profile. Also is -10 on the camera too low/high for good results?

I don't quite understand what you mean with that part.

In terms of the video, a lot of software allow you to produce a playback of the blinked images which would demonstrate how frequently and what is happening between each image.  

I'm unsure about colour CMOS cameras in terms of actual settings. I think noise in CMOS is meant to be generally quite low so excess cooling may be unnecessary?

It is strange that it has gone from being not a problem to all of a sudden being so which would suggest to me some change in the setup that is causing the issue.

 

[TheMan]

I was reading the zwo forums and some people have been having similar issues. One guy said he found a solution though. "I used SharpCap Pro to do a Sensor Analysis and then used the Smart Histogram tool to suggest an exposure for my sky conditions which are Bortle 4, 21.45 so quite nice. The exposure recommended was 255s at gain 120 with an Offset of 4. The drivers default to 28 and 30 (ASCOM/direct) on my laptop when clean installed. With an offset of 4 I get lovely data."  

It seems that the 294 mc pro is having a lot of issues with many people, i'm hoping zwo can resolve them if that is the case.

[TheMan]

Went out last night and resolved the issue. I used gain 120 offset 4 and used extreme dither. Walking noise is still there, but it is not noticeable.

 

 

Edited May 19, 2019 by TheMan

[Whirlwind]

Cool, great to see that you have found a solution.

I suspect what this is doing is reducing the dynamic range (so most of the walking noise is one/two levels and then the offset is slightly clipping this out leaving just the small remaining edges of the walking noise.  It may be worth playing with the parameters slightly starting from this point to see if you can change the gain / offset to see if there are any further improvements (for example to remove the remaining trace of the noise.  

[wimvb]

4 hours ago, TheMan said:

Walking noise is still there, but it is not noticeable

Is your dither working as it should? If you create a movie from your unregistered but calibrated subs, you should see the stars moving in roughly the same pattern as your dither movements. If they still move in a linear pattern, even if that pattern consists of forward and backward steps, your dither doesn't work. (In PixInsight you'd use the Blink tool for that.)

[atamuli]

On 17/05/2019 at 09:18, alan potts said:

I had something a bit like this from my Canon 40D, and found that it was me having ticked an incorrect box in DSS, I changed the darks to Median setting and it appeared to have solved it. I was told dither was the way to stop it without darks but have not been able to explore this yet due to weather.

Alan

Hello Alan,

I am exactly facing the same issue. Would you please share the settings that you use in your dark and bias settings in DSS?

Thanks in Advance,

AT.

[alan potts]

13 hours ago, atamuli said:

Hello Alan,

I am exactly facing the same issue. Would you please share the settings that you use in your dark and bias settings in DSS?

Thanks in Advance,

AT.

I am afraid I can't I have gone over to OSC Zwo camera now and have changed things a few times. I can't remember what I used other than in the setting there is an option of median with sigma clipping, I believe I used that. I am sure someone on here will have an answer for you. I may however be more useful to post an example and start a new thread, this one is rather old.

Alan