15 replies to this topic

[The_boots]

Since I'd been imaging for the monthly challenge on the few clear night I've had, my poor refractor has been languishing. I recently acquired a set of Askar HO and SO filters (ColoUr Magic C Duo) filters to try out some narrow band(ish) nebula work, so I swapped my filter drawer to the refractor, redid the wiring and headed out. 

My camera freaked out while taking calibration frames (it just was sending blank white and black frames) that resulted in my having one single useable dark, since I only realized what was happening after 10 minutes of pulling my hair out trying to get the newly attached EAF into focus. 

Nevertheless, a power cycle of my camera, and things smoothed out. Everything looked ok on the previews, except for the mount performing worse than expected (Wave 150i that was getting sub 0.5" guiding on a fully loaded C8 was struggling to stay below 1" on a 94/550 refractor), but it's still a new mount, and this is my wide(er) setup, so I didn't think it would matter.

Then I saw the results today.

 

 

I've never really had issues with walking noise before, but all of a sudden it was absolutely EVERYWHERE. So now I'm looking for solutions. I'm guessing the data is likely to be a total loss without a ton of extra work, but I want to make darn sure I figure out the source before I lose another night (last night was a pretty decent night!!).

The full setup in use was: 

SW

Wave 150i on an Innorel RT90C 

ASIAir

Sharpstar 94 edph with 0.8 reducer/flattener, effective f4.4 and 414mm

ASI 294MC Pro (MAN that sensor is noisy compared to the 533mc pro I've been using, but I thought the fov increase was a good idea...)

ZWO f4 mini guide scope with an ASI224MC for guiding.

40x2m subs for each filter

 

My current suspicion is that today when I was doing some initial checking for the source of the noise, I noticed that there was a rattle in my guide scope. The focus rings are secure, so I think the lens cell might be loose? Could that disrupt the guiding enough to cause this noise?

 

 

[fewayne]

Is this a processed image, or a single sub? It's awfully dim. Walking noise is essentially a stacking artifact resulting from pattern noise in the sensor getting averaged along with your signal; since it isn't random, it gets worse the longer the integration time.

 

I am unclear from your text whether you succeeded in producing darks. You do know that those don't have to be taken in the field with a cooled camera, right?

 

I'm stepping beyond my expertise here, but IIRC the 294 is among the few cameras still in common service that require dark flats, rather than just bias and flats. And I also seem to recall that folks struggle with too-short flat exposures with that sensor, resorting to dimming the flat light source to get long enough exposures to work consistently. Again, I could be wrong here.

 

What I do know is that posting all the details of your capture and calibration frames will probably help diagnosis by people much more familiar with the 294 than me.

Edited by fewayne, 24 June 2025 - 03:11 PM.

[The_boots]

Is this a processed image, or a single sub? It's awfully dim. Walking noise is essentially a stacking artifact resulting from pattern noise in the sensor getting averaged along with your signal; since it isn't random, it gets worse the longer the integration time.

 

I am unclear from your text whether you succeeded in producing darks. You do know that those don't have to be taken in the field with a cooled camera, right?

 

I'm stepping beyond my expertise here, but IIRC the 294 is among the few cameras still in common service that require dark flats, rather than just bias and flats. And I also seem to recall that folks struggle with too-short flat exposures with that sensor, resorting to dimming the flat light source to get long enough exposures to work consistently. Again, I could be wrong here.

 

What I do know is that posting all the details of your capture and calibration frames will probably help diagnosis by people much more familiar with the 294 than me.

This is the starless image partway through processing. I used this one because I thought it did a decent job showing the walking noise.

I produced ONE usable dark, the others did not come out. My flats should be ok took them first-- before capping the scope for darks and biases), although they look a little TOO flat... Everything past dark 1 (including all bias frames) are no good. There's a weird gradient instead of the expected pattern of noise and amp glow. 

I'm not sure which specific details will help, so I am putting all the files into Google Drive here.

I'll say that the noise pattern itself is there, no matter what calibration frames I use. The issues with the camera were resolved with a power cycle.

[bignerdguy]

To remove walking noise, you need to dither your shots. Usually every four or five shots at the very minimum need to be dithered, but preferably dither every one. That way, the walking noise gets averaged out. Flats and darks won't really resolve that problem. That's more of an issue with the static noise and the imager being the same across the even every picture because it's something that's inherent to the sensor, not specific noise that can be generated from interference or from outside sources. Dithering will usually remove that. I have the same problem I can and can't. If I don't dither, I get a lot of walking noise. That's a pretty common problem with DSLR's and some other types of cameras.

[michael8554]

Walking Noise shows up best (worst ?) on a stacked image.

 

Here's the image after a quick stretch:

 

 

Plenty of noise, but is it Fixed Pattern Noise aka Walking Noise ?

[The_boots]

Walking Noise shows up best (worst ?) on a stacked image.

 

Here's the image after a quick stretch:

 

25June.jpg

 

Plenty of noise, but is it Fixed Pattern Noise aka Walking Noise ?

Yes. The first image I posted is stacked and has the stars removed to make it easier to see. Instead of little dots, there are pairs of vertical lines of color. 

 Based on the description of walking noise I've read and some of the example images, it sure seems like that's what this is. Do you disagree?

[The_boots]

Here is a zoomed in version. It's incredibly bad... The noise itself I expect, it's the fact that it makes the paired line shapes that makes me think that there's something else going on. 

 

I don't think this pattern is inherent to the setup, so I'm going to stick with my instinct that the noise is from a progressive guiding error caused by the guide scope's loose lens cell. I think that as the the attitude of the telescope shifted, the lens slid-- which gradually misled the ASIAir into wandering. Someone please correct me if that's not plausible.

Attached Thumbnails

• 

[licho52]

I don't think this pattern is inherent to the setup, so I'm going to stick with my instinct that the noise is from a progressive guiding error caused by the guide scope's loose lens cell. I think that as the the attitude of the telescope shifted, the lens slid-- which gradually misled the ASIAir into wandering. Someone please correct me if that's not plausible.

You need to dither early and often.

[The_boots]

You need to dither early and often.

Or just tighten the lens cell so it isn't loose any more?

[Pgblack]

Or just tighten the lens cell so it isn't loose any more?

Yes tighten the loose lens, but you still need to dither to help the stacking algorithms to remove noise, hot pixels etc.

[The_boots]

Yes tighten the loose lens, but you still need to dither to help the stacking algorithms to remove noise, hot pixels etc.

Are there any guidelines or rules of thumb for frequency and spread of dithering?

[bignerdguy]

Are there any guidelines or rules of thumb for frequency and spread of dithering?

Not sure about frequency but dithering you generally want a minimum of around 5-8 pixels or more in random directions.  Most astro-software can do the dithering automatically and will dither in multiple directions.

[WadeH237]

Here is a zoomed in version. It's incredibly bad... 

Can you please clarify exactly what we are looking at here?  Is this a single sub, or is it a stacked image?

 

Walking noise is an artifact of stacking.  If you are seeing this in a single sub, it cannot be walking noise.

[markalot63]

It's not a guiding problem.

 

> My camera freaked out while taking calibration frames

 

You've got a cable problem. 

[The_boots]

Can you please clarify exactly what we are looking at here?  Is this a single sub, or is it a stacked image?

 

Walking noise is an artifact of stacking.  If you are seeing this in a single sub, it cannot be walking noise.

It's a close up of the stacked image. The subs each have a dot of noise instead of the pattern. I tried stacking two subs and the dots of noise were pairs of dots instead of the pattern here. 

 

It's not a guiding problem.

 

> My camera freaked out while taking calibration frames

 

You've got a cable problem. 

I mean, there are multiple problems-- the calibration frames exploding might be a cable thing, but it resolved with a power cycle, so I think it's more likely a camera problem.

I think the noise pattern itself is a guiding problem, but I thought the lack of darks might have made the noise more visible. So I was trying to explain why I wasn't able to use my calibration data from that night.

[Pgblack]

Are there any guidelines or rules of thumb for frequency and spread of dithering?

I'm certainly no expert but what works for me is every third frame for 180s exposures and every fourth for 120s exposures. What's important is to do it often enough to allow the stacking algorithms to work. With very short exposures you can dither less frequently and with very long exposures maybe every frame. It may also depend upon how quickly or otherwise your mount settles after a dither and hence how much lost time is acceptable to you.