Are we there yet?

[Pete]

Wednesday, December 2, 2020

There’s a forecast for a break in the overcast early evening so took the opportunity to implement possible fixes to Vignetting banding that I came up with earlier today.

The entire optical train was wrapped to exclude possible light leaks, and lengthy flats were run: 20 X 3 sec. Setup was optimized running T shirt flats in the twilight.

Flat capture screen.jpg (294.99 KiB) Viewed 2709 times

201202 flat.jpg (93.98 KiB) Viewed 2709 times

And since this record documents all, here’s the dark:

201202 dark.jpg (623.51 KiB) Viewed 2709 times

Tonight’s target – IC 5146 – C19 – The Cocoon Nebula

Frames were processed through Images Plus in three different configurations 1) Used the prestacked live image FITs file 2) Stacked the 19 individual FITs files in IP and then processed 3) Same as #2 only also added dark and flat to mix. #1 was far superior, #2 was a distant second, and #3 sucked.

Here we go. 19 X 2 min, -15°C, gain at 200, darks & flats, 2 sec guiding

IC5146 annotated.jpg (880.54 KiB) Viewed 2709 times

Conclusions & lessons learned:

Yikes!!! Very minimal vignetting!!!. No banding!!! One or both of the fixes (longer exposure on flats and wrapping the camera against light leaks) seemed to fix things.

Also there’s the thing about CMOS noise being different from that of a CCD. With CCDs the longer the exposure the better the signal to noise ratio. But with CMOS one balances shot noise against sky glow, and stacking shorter exposures may be cleaner than stacking long exposures. Most recently I’d been running more than an hour at 5 minutes. This is only 2 minute exposures at a 38 minute total. And it’s far superior.
Perhaps I’ve arrived at where I wanted to be.

A note of gratitude to Roger M for his support. I might have given up without it.

Hgp 2 Dec 2020

[bluemax]

Pete, congrats on a nice image. I am especially interested in the comparison you made of CCD VS COMOS. What camera did you use?

My favorite comment about imaging is: It is a most frustrating yet rewarding thing!

[mark.m]

Hi, Pete:
(Great set of posts you've been sharing!)
For a QHY cousin of your camera, QHY has the following to say about horizontal banding:

"5. Horizontal banding Optimize
It is a common behavior for CMOS sensor which contains some horizontal banding. Normally speaking the random horizontal banding can be removed with multiple frame stacking so it will not effect the image. The periodic horizontal banding can not be removed with stacking so it will bring some problem for the image. For QHY600, by adjust the USB traffic in both Single Frame mode or Live Frame mode, you can adjust the frequency point of the CMOS sensor driver and it can optimize the horizontal banding appeared on the image. This optimized is very effective to remove the periodic banding in some conditions ."

After trying to demangle the English in their statement, I take away the following:

1. Horizontal banding contains both a random component and a non-random (periodic) component
2. The random component will average out when you stack images. The more images that you stack, the lower the residual random component.
3. The non-random component has an underlying characteristic "noise frequency." This noise frequency is sensitive to USB transmission speed. Anything that changes the USB transfer speed will affect the appearance of the non-random banding component. For some USB transfer speeds, the periodic component of the banding will appear to go away.
4. Some camera control software gives you a way to tweak USB transfer rates.

Of course, take all of this with a grain of salt because I've never imaged with a CMOS sensor (yet).
- Mark M

[Pete]

Turns out that CCD and CMOS are a lot more different than I knew. The rules of thumb that we apply to CCD are quite different from those used for CMOS.

Banding and grids are inherent in CMOS but can be avoided. At least I avoided them last night. I've still a lot to learn but at least I may be over the hump.

The SharpCap program recommended by ASI for imaging with their cameras was written by an Englishman - Dr. Robin Glover.

Here's Robin's 1 hour lecture that covers the special things to be considered when using CMOS. https://www.youtube.com/watch?v=3RH93UvP358

Some of the items of note are that one should specifically "tune up" control software for a CMOS camera. And his mathematically complex but credible argument that shorter CMOS exposures are better. Based upon one-time experience with my setup, 2 minutes was better than 5 minutes.

Sony is the last manufacturer of CCD chips and they're ending their production in 2021. So like it or not we gotta convert eventually.

[NGC7000]

Pete, thanks for the link to Dr Robin. Looks very interesting, and I'm excited to watch it in the near future.

Thanks,too, for the detailed documentation of your image and flat process. I have some ongoing flat result issues with my Nikon D5300 that I did not encounter with my Canon T3i although banding there WAS an issue), and some of what you mentioned here could be very useful.

Tom

[AstroGeek]

Wow, Pete, making BIG strides now. Looks great!!

[menardre]

Pete

Looks like you solved all of your imaging issues. Really nice shot of Cocoon Nebula.

Looking forward to seeing many more images from you !!

Thanks for the credit, but you were the one who persisted.

Roger

[Bruce D]

It's above my head standing on the sidelines Pete but my 2 cents- keep doing what you did for this image and you are golden- very nice!

[Pete]

Not complicated at all if you have a mentor to show you the way Bruce. I can talk you through it. It's just in knowing a few magic software settings.

If it's clear for the next meeting perhaps I can do a live session from the observatory, so all will be doing virtual imaging on their monitor as stuff happens. It actually is pretty easy and quick.

[bluemax]

I have SharpCap but have not used for acquisition so can't comment on Smart Histogram but, am enlightened by the concept of shorter exposures. The text of Robin's lecture talks about read noise related to exposure time so I plan to try for more/shorter ones. His examples compare 2e read noise cameras VS 10e.
My Atik 428EX is 5e read noise so, not too bad I guess. The Canon 450D Rebel I use (modded) has 3.6e so, wonder if the Canon at 3.6e
would be able to take shorter frames than the Atik. The Canon is CMOS while Atik is CCD. But according to the lecture, they are both good at what they do.
Maybe look for a 2e camera!
Once again we are educated and enlightened, thanks for the posting Pete!