M109 image - Request for Advice

[oopfan]

https://s3.amazonaws.com/oopfan-astroph ... rgan+B.jpg

There are multiple issues here:
1. Raining noise caused by declination drift.
2. Donuts caused by ice crystals forming on the objective.
3. Bloated stars.

I'd like advice, if I may, on the last of the three items but first let me explain the first two:
1. My mount is 50 years old. I have a stepper motor on R.A. but no drive on DEC. On the R.A. I replaced the synchronous motor with a stepper motor that is driven by a Raspberry Pi with Periodic Error Correction built-in. I use what I call "passive guiding". I have no guide scope or camera. I rely totaly on the precision of the polar alignment and the machining of the mount. Last night I did polar alignment using SharpCap at around 11pm with polar error less than 15 arc seconds (rated Excellent.) By the time I got the opportunity to image M109 the time was 2am after finishing imaging M81. The ambient temperature plummeted from +20F to +12F. I think that I lost polar alignment when slewing the scope to M109. Guiding was excellent for M81.
2. I wasn't using a dew heater, so now I know better next time.

My question is about item #3: Bloated stars. Those three bright stars in the image are 9th magnitude blow torches. M109 has a surface brightness of +22. That's a lot of dynamic range! I used the Brain to choose the exposure settings. As usual I chose a relatively star-free region of the image (the upper-left hand corner.) The Brain recommended an exposure of 9.3 seconds and gain of 362 (the camera is an Altair 290M.) So...

Would you recommend that I halve the gain in order to double the Full Well, or should I leave the gain alone and halve the shutter speed? Or does it really matter, six of one half dozen of the other?

Thanks very much,
Brian

[admin]

Hi Brian,

sounds like you are trying to get the maximum dynamic range in the stack (ie the biggest range between the brightest thing that is not quite completely saturated and the faintest thing that can be seen). The 'brain' has a mode for this - just switch the 'aim for' dropdown to 'Max Dynamic Range' instead of 'Unity Gain'

Capture.PNG (39.77 KiB) Viewed 1771 times

This is an actual plot for a Altair 290 sensor, and you can see that for this camera the best dynamic range is at gain=200 where the HCG mode kicks in (which drops the read noise). For lots of other sensors the max dynamic range is at minimum gain.

If you still want to go further in terms of getting less saturated stars, adjust the 'Read Noise Limit' dropdown. What this controls is how much contribution from read noise is allowed in the total final noise of the stacked images. When this is set to 5%, it means that the total stack noise is allowed to be 5% higher than the minimum theoretical stack noise that you could get by not stacking at all and just taking 1 enormously long exposure.

Change this control from the default of 5% to 10% and you will see that the suggested exposure time approximately halves, at the same time giving you about an extra stop of dynamic range in the stack. That's because the noise level is only going up a small amount - an extra five percent or so, but the brightest thing that is not saturated is about twice as bright because you will be taking twice as many images of half the exposure time.

I'd estimate that you will get about an extra 1.5 stops of dynamic range by making these two changes.

cheers,

Robin

[oopfan]

That's great, Robin, thank you! You can never have enough dynamic range when imaging galaxies, in my opinion.

I really need to tackle this "raining noise" problem. Of course the proper solution would be to buy a better mount. Until then however I think I will place a reticle on top of a star, zoom in, and capture 10 minutes of data, and then see if the star deviated more than a few pixels. If it did I'll go outside and tweak the slow motion controls and then come back inside and capture another 10 minutes. It's not perfect but it will definitely improve things.

By the way I reprocessed M109 to eliminate the raining noise. Now I'm not embarrassed to show it to anyone. I lost some dynamic range but there's no such thing as a free lunch!

https://s3.amazonaws.com/oopfan-astroph ... rgan+B.jpg

PS: When I was imaging it last night I could see an additional tiny "fuzzy" that wasn't in the C2A DSO catalog. I thought that perhaps it was a comet. Today when I processed it I saw that there are 3 additional fuzzies: tiny galaxies far, far away.

[admin]

Yes, I can definitely see two other faint fuzzies in there which is nice, and you have good definition on M109 too.

If you are live stacking then you can look at the drift graph to see how the image is moving rather than having to mess around with reticules.

cheers,

Robin