Eclipse sequence - Dark, even, horizontal banding on the really short exposures

[BillyBoyBoy]

Hello all... first post for me here.

New user of SharpCap, drawn to it for the ability to script for bracketed exposures with a CMOS. I'm loving SharpCap!!

For the capture of totality with the total solar eclipse, I used linuxkidd's wonderful Eclipse Sequence Generator tool and it set me up very nicely to edit and build upon a great sequence. I have a ZWO ASI6200 that I'm using through a 140mm aperture refractor and using .SER format. I can get all the exposures I want to get in around a minute or so, then I repeat it until totality ends. For insurance if the corona is huge, I'm running a second scope with the same sequence in a 51mm and/or 71mm scope with an ASI2600MC, piggybacked on the 140mm.

Here's the issue I'm having though, and since SharpCap enables use of a CMOS camera with scripting for this purpose, and part of the bracketing approach with the solar corona is to include very short exposures, I'm wondering if the issue has been experienced by anyone.

My exposure sets start at .001s and run all the way to 5s. In the shortest exposure sets (.001s, .002s, .004s, and slightly in .008s), I am seeing banding. Not random banding, but more like stripes of evenly-spaced dark horizontal bands. My research thus far seems to indicate this may have to do with the camera not having a global shutter. It has a rolling shutter, and I'm wondering if this is the culprit. I've seen this with shooting flats too short sometimes, but I always thought that was due to the emission frequency of the flat panel, resolved of course with longer exposures. In this case, my light source was the Sun. Is this still just a rolling shutter thing? The clouds of course arrived before I could try the Banding Supression, so I don't know if that is relevant or has an impact.

The other question is - having not done this before - whether exposures as short as 0.001 through 0.008 are even contributive to capturing any coronal detail? With a 140mm aperture, I was wondering if I'd need them that short due to the bigger aperture, and additionally not sure for the 51/71 apertures.

Any insight?

Thanks!

Bill

[admin]

Hi Bill,

you are correct that unexpected banding is most likely related to the camera being a progressive scan camera. There are typically two main causes in my experience

1) Artificial light sources that flicker at 50/60hz - shouldn't be relevant in your case as you were using the sun as illumination

2) voltage fluctuations in the power supply to the camera

I've only ever noticed the second cause showing up in stretched images or at very high gains. You can see an example of how this might look in the SharpCap documentation here - https://docs.sharpcap.co.uk/4.1/#Banding%20Suppression . That might help you decide if what you were seeing was the same effect. If you were using the display stretch option in SharpCap to stretch the image then turn that off to check how significant the effect is without a stretch being applied.

I can't offer much advice on the exposure times as I personally have not done any eclipse imaging. I will point out thought that it is the f-ratio of the scope that will be the important thing to consider when working out if one scope is going to show brighter images than another - not the aperture.

cheers,

Robin

[oopfan]

Hi Bill,

I have photographed two total eclipses: Africa in 1973 and Mexico in 1991. I used Kodachrome slide film each time, so it is a lot slower than modern CMOS cameras. In Africa I used a 60mm f/11 refractor and in Mexico a SLR with a 200mm telephoto lens at f/5.6.

In both cases my bracketed exposures started at 1/125 sec and went up to 2 sec. 1/125 sec was good for prominences and the inner chromosphere. However, if I could do it all over again I'd work on the guiding because the long exposures were the most visually stunning.

My best advice is to shoot Bailey's Beads at 1/1000 sec but keep an eye on the histogram! It will tell you a lot about what exposures you should advance to once you get into the meat of totality.

Me, being an oldster, I still prefer bracketing over stacking for eclipses, but that is your choice. The fundamental problem is there are no do-overs. If you're lucky you get one opportunity per year, usually at high-cost of traveling. I say KISS (Keep It Simple, Sonny) and learn from this eclipse and then apply what you've learned to the next eclipse.

Edit: Don't forget to take off the solar filter once Bailey's Beads begin.

Good luck!
Brian

[Real14]

I am seeing banding. Not random banding, but more like stripes of evenly-spaced dark horizontal bands.

Hi,

Are those bandings parallel or do they have a common circular center.

They could be Newton rings and there you can tilt the camera to0g et rid of them.

Post an image to really see what your problem is.

[BillyBoyBoy]

Thanks, all.

Robin, I suspect the rolling shutter to be it. I'll try the Banding Suppression, but I expect that I'm trying to squeeze blood out of a turnip.

Brian, I can't imaging doing this with film, and I'm going to guess manual tracking! I don't know how closely the exposure times with film translate to a CMOS, but 1/1000 sounds pretty reasonable. Bear in mind, I am bracketing, but I am shooting hi-res video for each exposure, so I'll be stacking for each exposure time.... the best of both worlds. Hopefully, that gives me better results on each exposure time's image.

Real14, they're not Newton rings. It's straight horizontal banding. I deleted the images, not thinking I might want to include one in a post like this. I really think we're just trying to do something that a camera without a global shutter just can't do.

I'm still doing testing, and I'll post my findings with Banding Suppression and if I get any of that aforementioned turnip blood.

Clear skies! I'll definitely be posting my results once I muddle my way through all the data.

Bill

[BillyBoyBoy]

On the ASI6200MC, it appears the banding stops at 0.008s, which is a quarter of the 32ms that ZWO indicates is its shortest exposure time. I tested 0.001s all the way to 0.007 and banding is present throughout. And the banding rolls as I blink through each set of the same exposure. That's the rolling shutter.

6200MC Banding at 0.005s.jpg (112.8 KiB) Viewed 277 times

Bill

[admin]

Hi Bill,

that's very regular and not the same as the more random banding that 'banding suppression' was designed to help cure. It's also clearly visible in an unstretched image which is not typical either. It certainly looks like the banding you would get from artificial lights, but a daylight scene rules that out.

I checked the spec for the camera and the exposure is meant to go down to 32 microseconds (not milliseconds), so well below the point where you are finding this issue.

A couple of final things that come to mind to try...

* Swap out the 12V power supply to the camera for an alternative to see if that has any effect
* Try running the camera at a smaller ROI, which might just let you get to lower exposures before hitting this effect.

cheers,

Robin