Vertical orange and blue banding in ASI585MC images

[skipgazer]

I've been imaging successfully for two years on my Sky-Watcher Explorer 150P with a modified Nikon D750 - but I now have a new ZWO ASI585MC astro camera so I feel like I'm having to learn things all over again. I have taken some great Moon images, really pleased with those. But an issue that is plaguing my deep sky attempts is obvious, precisely vertical, mostly orange and blue banding in the images. I cannot work out what is causing it. I have attached an overstretched (to highlight the issue) example: 566 x 20", gain 252 offset 10, dithered - but I've tried different settings and it makes no difference. I have stacked in Astro Pixel Processor, which I usually use, and also in Siril - again, makes no difference, the banding is still there. I managed to use Siril's banding reduction to eliminate it, but I'd rather it wasn't there at all. People have suggested maybe my flats are the issue, but it's the same with or without flats, and my master flats look fine to my eye. Any thoughts? There's a UV/IR cut filter and MPCC III coma corrector in front of the camera, fwiw. Is it something about the 585MC?

[alacant]

 Hi

Nice image.

A few bits we need:
-power supply
-light source for flat frames
-exposure for flat frames.

For now, simply run the whole pp sequence through Siril's vertical banding before registration.

Cheers

 

 

Edited January 27 by alacant

[WolfieGlos]

I’m currently imaging with my 585mc with a triband filter, gain 252 180s subs. I’ll let you know if I get any of this banding.

I’ve used it briefly back in November with a near full Moon with a UV/IR cut and it didn’t show up. Both that session and the one I’m running right now, are with my 102ed refractor. 

A few thoughts;

1) Could it be high level cloud passing in some of your frames? Does it still show up if you try stacking, say, 10 frames from the start of the session? 

2) How often are you dithering, and what settings are you using? The vertical band almost makes it look like walking noise, but I wouldn’t expect to see it running vertically.

3) What’s the offset value for? I’m fairly new to my 585mc astrocam myself, and I’ve just checked my imaging settings in NINA and it’s set to 1.

[Adam J]

11 hours ago, skipgazer said:

I've been imaging successfully for two years on my Sky-Watcher Explorer 150P with a modified Nikon D750 - but I now have a new ZWO ASI585MC astro camera so I feel like I'm having to learn things all over again. I have taken some great Moon images, really pleased with those. But an issue that is plaguing my deep sky attempts is obvious, precisely vertical, mostly orange and blue banding in the images. I cannot work out what is causing it. I have attached an overstretched (to highlight the issue) example: 566 x 20", gain 252 offset 10, dithered - but I've tried different settings and it makes no difference. I have stacked in Astro Pixel Processor, which I usually use, and also in Siril - again, makes no difference, the banding is still there. I managed to use Siril's banding reduction to eliminate it, but I'd rather it wasn't there at all. People have suggested maybe my flats are the issue, but it's the same with or without flats, and my master flats look fine to my eye. Any thoughts? There's a UV/IR cut filter and MPCC III coma corrector in front of the camera, fwiw. Is it something about the 585MC?

so I barely see it. 

Are you using dark frames? Flat frames? what are you calibrating the flats with? 

If darks how well temperature matched are they? 

If yes can you post them please. 

Are you dithering, what's your dithering scale?  Can you post a histogram. 

What are your stacking settings, rejection etc. 

Adam

 

[Adam J]

On 26/01/2024 at 16:29, alacant said:

 Hi

Nice image.

A few bits we need:
-power supply
-light source for flat frames
-exposure for flat frames.

For now, simply run the whole pp sequence through Siril's vertical banding before registration.

Cheers

 

 

Power supply will be USB, would hope a USB 3 port is being used through. 

[LandyJon]

21 hours ago, WolfieGlos said:

3) What’s the offset value for? I’m fairly new to my 585mc astrocam myself, and I’ve just checked my imaging settings in NINA and it’s set to 1.

Offset basically pre-fills each pixel with a small amount before starting to capture light, I dont think its really necessary with these CMOS sensors, but your 1 or the OPs 10 out of a full well of 47,000 are a minimal amount.

More importantly with your sensor like many others they have a HCG on the gain at 252, so be sure to use gain there or above as there's a dramatic reduction in read noise.

On 26/01/2024 at 11:54, skipgazer said:

it's the same with or without flats,

You don't mention darks, have you done them ? does it show on them ? remember darks need to be done at the same temp as your lights, particularly with an uncooled camera you'll want to do a few darks for each imaging session as the camera can't be set to a temp it'll be running slightly warmer or cooler dependant on the ambient temp of the night.

[900SL]

Gulf syndrome? 

[skipgazer]

Thanks everyone. Like I said I'm familiar with a Nikon DSLR so am accustomed to using flats, flat darks and darks in stacking my lights, which I've been doing successfully in Astro Pixel Processor. The new ZWO camera has thrown me so I'm still learning what's different, and wasn't expecting this banding - which I guess is probably fixed pattern noise in the sensor? I can process it out but I'd rather it wasn't there in the first place. It's not discernible in the individual lights: stacking brings it out. Not what I'd expect. On some of the questions: my setup is (as it has always been): camera on Explorer 150P focus tube through Baader coma corrector (with 37.5mm spacer to get the right back focus). Power supply for the camera is USB 3.0 from the Stellarmate Plus controller [I've just ordered a tester to check the output but it should be more than enough for the camera] which is power via its transformer from a 12v Jackery power bank; fed via ethernet to Ekos on the iMac in my garden office, which is a quick and very reliable connection. 

Darks: usually shot (as in this case) at the start of the session, so the setup is entirely the same, just with the rig's coverall on as well as the cap on the OTA (and a dark cover on its primary end). 

Flats: I've always used an A4 LED light panel. With the 585MC I've got that on its lowest (of 10) brightness levels, through three sheets of 80gsm A4 paper and double thickness white T-shirt, which let me shoot 1" exposures. 

The master dark and master flat produced by APP look fine to me. (attached). EDIT: also attached a sample bias frame - I shot 30. If I can see anything, it's some horizontal stripes but certainly not vertical bands. 

Dithering: between each frame @ 500ms pulse (unguided dithering in Ekos) with a 1" delay before the next frame. I do not think this is working correctly given the walking noise I'm seeing, but I think that's a separate unresolved INDI issue which I'll pursue elsewhere. Anyway the walking (rain) noise is very different from the regular vertical banding that's at issue. 

Like I say, maybe the camera is just like this, but I was surprised to see such obvious vertical stripes. 

 

Edited January 28 by skipgazer

[alacant]

Hi

You can't control the sensor temperature, so best to lose any type of dark frame. Try without? Instead, simply remove the bias.

To lessen banding from light panel flicker, use an extra t-shirt and increase the brightness. Anything over 2s exposure should be fine. EKOS' flat frame wizard is your friend.

HTH

 

 

 

 

Edited January 28 by alacant

[skipgazer]

Well, here's a short stack (less than eight minutes) on a bright target, with the middle stretched. No flats, no darks, so they're not the problem. Lights and bias only. (Oh and a full Moon, as you see in the corner). Have I just got a dud, or are they all like this? I'm used to seeing noise from my DSLR; I'm not used to seeing stripes! 

Edited January 29 by skipgazer
to mention the Moonlight

[alacant]

.

 

 

 

 

 

Edited January 31 by alacant

[WolfieGlos]

1 hour ago, skipgazer said:

Well, here's a short stack (less than eight minutes) on a bright target, with the middle stretched. No flats, no darks, so they're not the problem. Lights and bias only. (Oh and a full Moon, as you see in the corner). Have I just got a dud, or are they all like this? I'm used to seeing noise from my DSLR; I'm not used to seeing stripes! 

Wow, that's really bad with no calibrations at all.

Is the camera still in the free returns or warranty period? If you purchased it from FLO I would get in contact with them, they might be able to offer specific advise or possibly exchange it for a one that works if it is indeed faulty. I had to do this with a ZWO electronic focuser last year and they were really good about it - even dispatched the replacement on the same day of receiving the duff one (which was a dead motor on arrival).

One thought I've had; did you shoot this with an offset value again? I know @LandyJon said above that it would make minimal difference (thanks for the explanation by the way ), but that's the only real difference I can see between my 585mc and yours in the capture settings (scope and coma corrector aside - mines a 102ED refactor with 0.8 reducer).

From my earlier post, I captured 1 hour of subs on the Horsehead with the 585mc + Askar Ha/OIII dualband filter with the Full Moon the other night - more of a test than anything else - but it's nearly the same area of sky as your M42, and below is what I got (with flats and darkflats) - no vertical banding is present. This was 20 x 180s at gain 252, with a basic stretch.

[LandyJon]

2 hours ago, alacant said:

no filter

I think this might be a good next step, take the filter and CC out of the equation.  Try another short stack, forget the bias too, just lights, don't worry about spacing, just get the camera direct on to the scope and see what happens, if it's still there you've ruled everything out, if it's gone you can re-introduce things one by one.

Reason I suggest is I've just done a  crazy stretch to your first image and the banding looked more of a grid and the only time I've seen anything like was due to reflections from a filter being too close to the sensor.

Worth a try before looking to return it.

 

[skipgazer]

Interesting thought on the filter and CC, thanks. Worth a try. I have completely nailed my flats now, using the Ekos automation and exposing for 4". They've made no difference. I also thought it might be a power issue because I was using a 12v splitter into the power bank. Removed that. Made no difference. So, yes pared down to basics next. Unfortunately we have clouds forecast forever . . .