M81 Bode's Galaxy in 11.6 hours

[turfpit]

This just demonstrates one of the challenges faced by imagers. The good windows on objects can be quite small in reality.

Interesting analysis Minos - 't tests' - had me reaching on the shelf for my 50 year old copy of Advanced Level Statistics by Francis.

Going for a lie down now ......

Dave

[oopfan]

Hi Minos,

Thank you for your analysis. This topic is an intriguing one and requires further study. There is one thing that is bothering me though. Increased light pollution due to human activity for a couple hours after astronomical twilight suggests to me that my Bortle class is temporarily raised. For example, after midnight when all are asleep my Bortle class is 5 but I can certainly imagine that my Bortle class might be 6, temporarily, while human activity waxes and wanes after sunset.

If I capture my red stack during this temporary hike in Bortle class then I would expect that the stack's signal-to-noise ratio would be lower than it otherwise would. This contradicts what I found in APP. I found that I had a strong red stack compared to blue. In fact I had to attenuate the red stack in order for the blue stack to be seen.

How do we square this circle? I think the answer is that M81 is circumpolar. Recall in my earlier post that an imaging session began at about 47 degrees above the horizon and ended at 62 degrees (that is within 1 hour of hitting the meridian). Normally I do most of my imaging at a different location in the backyard where I enjoy an east-facing window. There it is not uncommon for me to begin imaging at 40 degrees and finish greater than 80 degrees. The atmosphere becomes really steady at high altitudes. I use SharpCap to aid in guiding. It's really quite remarkable how small and bright stars get. High altitude imaging yields high signal-to-noise ratio (SNR) stacks.

Recall that I capture blues later in the session, so its SNR is quite high compared to the red stack taken earlier. I must admit that lower atmospheric extinction helps boost the red stack but I think it is fair to say that my blue stacks probably have greater SNR than the red stacks.

Now let's return to M81. There our peak altitude was 62 degrees not 80+ degrees so M81's blue stacks suffered from lower SNR. And since I began imaging at 47 degrees, not 40, the red stack's SNR was boosted.

So in an exaggerated nutshell:
1. Normal imaging: Low SNR reds, High SNR blues.
2. Circumpolar imaging: High SNR reds, Low SNR blues.

This would explain why my red stack was stronger than the blue stack, and why I had to attenuate it in post-processing.

Reminds me of an email exchange I had with a top imager last year. I explained how I white balance my camera and filters using a G2V star. He was familiar with the technique and uses it himself. And then he asked me how I modify the ratios in real-time as the altitude of the object changes. I explained how I capture reds first, then greens and blues. Hmm, now I see the importance of what he was saying.

Brian

[oopfan]

Minos,

Extending my thoughts from my last reply, it is not too difficult to develop a solution to calculate a stack's total SNR in real-time as frames are captured. The easiest solution would be a spreadsheet. All that is required is to enter the altitude of the DSO when a frame arrives. The spreadsheet does the rest. I always run C2A planetarium software while I image. C2A gives a real-time readout of the altitude of the DSO. My job would be to wait for a new frame and then glance at C2A and enter the altitude into the spreadsheet. Of course there are other solutions that are completely hands-free but require a bit more work to program. In a multi-session project the goal is to hit a target SNR on all three color channels. The number of frames will most likely be different. Notice in my original post that I captured 90 frames each of red, green, and blue. That would now change. In the case of M81 it might be 75, 85, and 90 for red, green, and blue, respectively. This way I have balanced SNR going into post-processing, mitigating potential problems like I experienced.

Brian

[oopfan]

Minos,

Great news! I programmed the spreadsheet. My hunch was right. Here are the results:

Comparison of Stack SNR of Circumpolar vs East-Facing DSOs.jpg (27.38 KiB) Viewed 1655 times

I pulled the timestamps from my image files and used C2A to calculate the altitude, then plugged the altitude into my new spreadsheet which takes account of atmospheric extinction in its calculation of SNR.

Any East-Facing DSO would be M31, for example, which I imaged this past September. There is a brilliant blue star cloud in one of the arms that came out very well. Notice in the attachment how the Blue Stack has a higher SNR than the Red Stack (10.711 > 10.634).

Notice how the colors are reversed for M81. The Red Stack is higher than the Blue Stack which explains why M81 was skewed towards the red with only hints of blue in the outer arms. To correct this I used APP to attenuate the Red Stack.

I will use this spreadsheet while imaging going forward. I'll target a single SNR for all channels. I'll capture frames until the spreadsheet tells me to quit.

Brian

[mAnKiNd]

Brian, this is a great adaptation you've surmised for calculating SNR according to where you're pointing in the sky.

With regards to differential bortle scales depending on compass and altitude pointing, I agree with this conundrum. I don't think it accurate to account for a singular bortle scale value according to zenith, as our situtations involve imaging through varying levels and sources of light pollution. One way I've thought to account for this, is to strap a unihedron SQM meter to the scope, pointing in the same direction I'm imaging at and then periodically take measurements of the sky brightness. Obtaining a range of values as they change throughout the night, would then enable one to reference those values to bortle scale ranges. This way, for example, you can state that you started imaging in a bortle 8 sky and ended up with a bortle 6 value at the end of your session as the target reaches it's apex. OR better yet, provide the range of SQM values from start to finish with a mean or median. I haven't come across anyone trying this to validate my assumption, but on the surface, this seems like a decent idea if one cares about collecting such values. I plan on getting an SQM meter soon and trying this out.

Well done once again for continuing to develop your SNR calculations in your attempts to quantify how much data is actually needed to achieve an image of a certain standard and white balance - math does not lie.

Best,
Minos

[oopfan]

Minos,

I too was considering purchasing Unihedron's SQM. It will be interesting to hear whether or not your neighbors' outdoor lights interfere with the readings.

I do have a column in my spreadsheet for Sky Brightness in units compatible with Unihedron (mag / arcsec^2). It would be a simple matter of recording the sky brightness along with the altitude. Actually, altitude can be calculated given a timestamp, the object's coordinates, and your latitude and longitude.

Brian

[mAnKiNd]

Cool, we can start a thread on it when one of us eventually gets one...just keep in mind that acquiring astrotoys takes time with me

[oopfan]

I've created a new blog post that sums it up nicely:

https://astrotuna.com/m81-bodes-galaxy-in-11-6-hours/

Thanks everyone for your interest and participation!

Brian