M81 and the Integrated Flux Nebula (IFN)

A place to share images that you have taken with SharpCap.
Forum rules
Please upload large images to photo sharing sites (flickr, etc) rather than trying to upload them as forum attachments.

Please share the equipment used and if possible camera settings to help others.
User avatar
oopfan
Posts: 1321
Joined: Sat Jul 08, 2017 2:37 pm
Location: New York
Contact:

M81 and the Integrated Flux Nebula (IFN)

#1

Post by oopfan »

I've attached an animated GIF that alternates between two images. (Please leave a comment if your browser is not rendering it as such.)

The benchmark image comes from AstroBin:
https://www.astrobin.com/iru2mp/0/

The benchmark was captured at Bortle 2 with 14 hours integration time and an ASI1600MM. My image was captured at Bortle 5 with 1 hour integration time and an Atik 314E CCD.

I hope you can see the IFN in my image.

NOTE: In the attached animated GIF, my image looks strange. I severely stretched it, and then applied a good amount of de-noising. Fortunately, the de-noising does not harm the large-scale structure of the IFN in this region of M81.

EDIT: I modified the Black Level making it easier to see:
M81-IFN-30-2-0-ANIME-BL.gif
M81-IFN-30-2-0-ANIME-BL.gif (166.97 KiB) Viewed 1649 times
My technique was to find the exposure that raises light pollution (LP) to half of my camera's Full Well Depth (FWD) which is 6700 electrons or 32000 ADU. That exposure was 700 seconds at bin 3. According to my calculations, the IFN is only 43 electrons or 205 ADU above LP. That represents a difference of only 0.64%. Ordinarily 205 ADU would be easily seen in a "normal" image, but in this image there is a lot of noise due to light pollution which drives the signal-to-noise ratio (SNR) down.

You may ask, why not use a shorter exposure so that light pollution isn't such a large factor? Good question. The answer is that it also means that we are also capturing fewer IFN photons. We want to put as much "distance" between the IFN and LP. Since noise grows as the square root of signal, it behooves us to grow the exposure as much as possible to achieve the highest SNR. The thing that prevents me from achieving an even higher SNR is my camera's relatively small FWD of only 13400 electrons. A better option is a camera with a FWD of 50000 to 100000, and a large aperture scope to hoover-up more photons per unit time so that the exposure time isn't ridiculously long.

You may ask, why not use a shorter exposure and just stack more? Even with a low-noise CMOS camera, Read Noise is a small percentage of LP noise. True, if you lower exposure, then LP noise becomes less of a factor, but now you are in the same situation as above: the IFN SNR becomes so low that it is barely detectable.

What is the solution? Travel to a site with darker skies where LP noise becomes less of a factor.

Brian
User avatar
Menno555
Posts: 1053
Joined: Mon Apr 20, 2020 2:19 pm
Location: The Netherlands
Contact:

Re: M81 and the Integrated Flux Nebula (IFN)

#2

Post by Menno555 »

Very cool Brian!!
Clear description too and good result!
Here still waiting for clear skies which will happen in ... oh, just did see maybe in 4 days :)
Decided also to go for bin 2 or 3. Stars will get somewhat bloated then but that doesn't matter in this case.
Btw I found a cool survey/overlay in Stellarium which shows (roughly) IFN regions, see screenshot. You can see it like this in grayscale with the IRAS-IRIS Band 4 - 100µm or with the Planck-2015-Pol-Dust or in color with the IRAS - IRIS HealPix survey color.
Based upon that, I think I will go for closer to Polaris. There are some relative "dense" parts there with some nice (very) small and interesting galaxies in it. Plus it's easier for me with tracking for the long exposure.
It will probably give nothing but it's fun to try :)

ifn_stellarium.jpg
ifn_stellarium.jpg (322 KiB) Viewed 1624 times
User avatar
oopfan
Posts: 1321
Joined: Sat Jul 08, 2017 2:37 pm
Location: New York
Contact:

Re: M81 and the Integrated Flux Nebula (IFN)

#3

Post by oopfan »

Hi Menno,

Hey that's a neat overlay! Polaris region looks brighter, but just be careful. The farther down towards the horizon you look, the more atmosphere there is between you and the IFN. It's interesting how deceptive that AstroBin image is. Clearly the author employed some post-processing trickery to make it look like the IFN and M81 look approximately the same brightness. My image is closer to reality. No curves adjustment. I was surprised how faint it is. It's just my guess but I believe it is closer to 26 mags.

How good is your dew shield? Is it one of those long, extended ones? I ask because I had a gradient that could only be explained by my flat-fielder device. In the attached photo, the flap is in the open position. It is difficult to see, but the electroluminescent panel is affixed to the underside of the flap. My theory is that stray light from the living area in the house is reflecting off of the glossy surface of the panel and entering the scope. When I image M81, I have to look over the house. The easy fix is to rotate the flat-fielder 180 degrees so the hinge is on the bottom.

Good luck!
Brian
Attachments
The Flatinator.png
The Flatinator.png (220.69 KiB) Viewed 1622 times
User avatar
oopfan
Posts: 1321
Joined: Sat Jul 08, 2017 2:37 pm
Location: New York
Contact:

Re: M81 and the Integrated Flux Nebula (IFN)

#4

Post by oopfan »

To everyone,

I want to reassure everyone that I am using the same formulas as the Brain, although it may not seem that way. I am 100% confident that the Brain would recommend the same exposure if I could tell it that I am "extremely intolerant of noise." Why? Because the IFN is very faint and my light pollution is relatively high. If the IFN was brighter, then I would say I am "somewhat tolerant of noise." If it was even brighter, then I'd say I am "tolerant of noise."I don't have control over the IFN's brightness, but I do have control over light pollution by traveling to a dark site. At Bortle 2, I would say I am "tolerant of noise." However, here at home at Bortle 5, I have to say I am "extremely intolerant of noise." It is no more complicated than that. I am not challenging the effectiveness of the Brain. We are following the same Physics.

Brian
User avatar
Menno555
Posts: 1053
Joined: Mon Apr 20, 2020 2:19 pm
Location: The Netherlands
Contact:

Re: M81 and the Integrated Flux Nebula (IFN)

#5

Post by Menno555 »

Brian

If you say "extremely intolerant of noise", do you mean that when using the Smart Histogram Brain, you would enter a SNR of 1% ?
But the new Brain in the Beta SC gives new "challenges" with the added features. I did try that with the option for "darkest" (so, saying that I want to capture the most faint things), and then it returns a suggestion with a relative long exposure but also with a high Gain of around 230 (of 600). This contradicts the lower gain you calculated, right?
Or is that also dependent of the aim for Unity Gain or Highest Dynamic Range?

On Polaris: for me that's at around 53 degrees, so rather high. And the IFN is at the moment above Polaris, so that's even a bit higher.
And yes, I have a long flexible dew shield with a black velvet like inner layer.
User avatar
oopfan
Posts: 1321
Joined: Sat Jul 08, 2017 2:37 pm
Location: New York
Contact:

Re: M81 and the Integrated Flux Nebula (IFN)

#6

Post by oopfan »

Menno,

I haven't looked at the specs of any of the new 16-bit CMOS cameras, but one of the problems with older models is that they were mostly 12-bit, with some 14-bit models. My old Altair 290M is 12-bit. At its lowest gain setting, the conversion factor is 3.62 electrons per ADU. That does a poor job of capturing faint detail with short exposures, but it does a good job of capturing bright nebulae and stars.

It is never a good idea to choose an exposure that puts the detail you want to capture at the bottom end. Let's look at an example. Say you have a faint nebula that emits 1 electron per second. At one-second exposure you capture one electron, but you won't see it because the gain is 3.62 electrons per ADU. Next, let's double the exposure to 2 seconds. You may start to see it every other frame on average but it is only two levels. Next, let's boost the exposure to 60 seconds. Now you've got 16 levels (60*1/3.62). The nebula has depth of detail. You could raise the gain setting to 1 electron per ADU. That gives you 60 levels, even more depth of detail. However, you are giving something up. More stars are saturating, and even bright nebula. There comes a point where raising the gain further negatively impacts the faint nebula. Sensors are not linear up to the point of saturation. At around 50000 ADU the response curve flattens, and slowly creeps up to saturation at 65535. You want to avoid this.

Now we need to talk about light pollution. For argument's sake, let's say that LP is 100 electrons per second. At one-second exposure you capture 100 electrons of LP and 1 electron of faint nebula. So, LP is 100 electrons, and LP+Nebula is 101 electrons. You are still only going to see 1 electron of detail in the nebula. Now, start raising the gain. You will get more levels as before, but raising it further ends up saturating the sensor.

It is a delicate balance between raising the gain and extending the exposure. The advantage of raising the gain is that it yields more levels of detail for a given exposure. The advantage of raising the exposure is that it yields a higher SNR for a given gain setting. However, one must be mindful of your light pollution. Raising either the gain or exposure can lead to sensor-wide saturation. Robin can speak more about this than I can, but my hunch is that the Brain is solving simultaneous equations that yields the best combination of both.

My CCD has a single gain setting: 0.21 electrons per ADU, so my entire world revolves around exposure control. CMOS has both exposure control and gain control, so it is that much more difficult to properly utilize without the aid of software.

EDIT:

Menno, use the Brain with the IFN. Use the equivalent of "extremely intolerant of noise". Capture a single frame. Sample the ADU of the sky in that image. Write it down. I'm curious to know the ADU level. Capture an hour, and process. If you don't see the IFN, I suggest raising the exposure and lowering the gain.

Something to think about: perhaps the IFN isn't a good target since we don't know if your LP is just too high. Perhaps you should consider a slightly brighter nebula. I could search for one that I think you might have an easier time of.

Brian
User avatar
Menno555
Posts: 1053
Joined: Mon Apr 20, 2020 2:19 pm
Location: The Netherlands
Contact:

Re: M81 and the Integrated Flux Nebula (IFN)

#7

Post by Menno555 »

oopfan wrote: Fri Apr 09, 2021 2:36 pm ....
EDIT:

Menno, use the Brain with the IFN. Use the equivalent of "extremely intolerant of noise". Capture a single frame. Sample the ADU of the sky in that image. Write it down. I'm curious to know the ADU level. Capture an hour, and process. If you don't see the IFN, I suggest raising the exposure and lowering the gain.

Something to think about: perhaps the IFN isn't a good target since we don't know if your LP is just too high. Perhaps you should consider a slightly brighter nebula. I could search for one that I think you might have an easier time of.

Brian
Oh, don't worry about that :)
I know my limitations (well, my equipment's limitations :) ) and that there is a very high chance of "failure" with this. But I don't care. I am going to try and if it doesn't work, it doesn't work. And if it does, I'm happy :D
I mean, I never ever imagined that I could capture so much in my ARP299 deep field from my balcony and high LP skies, yet I capture galaxies of mag 22 and 6 billion lightyears distance or a quasar of mag 19 and 11.4 billion lightyears. Granted, it's just a very faint little blob and a small blue pinpoint, but it's there.
I'm just a hands on guy: I do it, learn, experiment, see what works and doesn't work, try to use valuable information like you are giving, and so on. And still with no clue what so ever on SNR calculations/formulas ;)

On the ADU in a single frame. I wouldn't know how to sample the ADU of the sky? Or do you mean the overall average ADU in the histogram, the white line?

Menno
User avatar
oopfan
Posts: 1321
Joined: Sat Jul 08, 2017 2:37 pm
Location: New York
Contact:

Re: M81 and the Integrated Flux Nebula (IFN)

#8

Post by oopfan »

Hi Menno,

The histogram should look something like this:
LP Histogram.jpg
LP Histogram.jpg (9.19 KiB) Viewed 1489 times
Brian
User avatar
Menno555
Posts: 1053
Joined: Mon Apr 20, 2020 2:19 pm
Location: The Netherlands
Contact:

Re: M81 and the Integrated Flux Nebula (IFN)

#9

Post by Menno555 »

Ahhh, okay. Yes, clear! :D
timh
Posts: 515
Joined: Mon Aug 26, 2019 5:50 pm

Re: M81 and the Integrated Flux Nebula (IFN)

#10

Post by timh »

Coming in rather late on this discussion..

The animated GIf and overexposed M81 image works really well Brian - you can definitely see that the IFN is faintly there

My deepest images at F 4.5 and Bortle 6-7 do show a good extended detail on the outer arms of M81 but not a sniff of the IFN - which is not surprising if it is down at magnitude 23/ 24 or so - and along with the fact that the part of the IFN near M81 is not the brightest region. In addition it would, I am sure, require completely blowing out and overexposing everything else which I didn't try. Maybe the way to approach getting the IFN in the same picture as one that includes M81 is to go for high dynamic range compositions with a wide range of exposure lengths?

Anyway, having just been away on a break where the skies were Bortle 4 it was obvious that 2 Bortles makes for a big difference in what is detectable - At my normal offset I needed exposures 2-3 times as long under darker skies just to move the histogram away from the left hand axis. Fainter details were visible much more quickly some of which, for any feasible total imaging time, would probably be inaccessible at Bortle 6 (although actually the seeing at the dark sky site was very poor which limited what I could capture on this particular occasion).

I think that I am coming to the view that 'portability' is one of the most desirable features of any astro set up - because dark skies really help and I am sure that you are right about that being key to detecting the IFN in this case.

PS. If you set the % tolerated noise using the BRAIN algorithm very low then it does exactly what you would expect and recommends longer exposures. However it also confirms that sky brightness is the real killer. For any given total exposure time (irrespective of the length of the single frames) the threshold detection level for faint low electron flux features (Y axis) is mostly dependent on background sky brightness. At high brightness, the total time required to detect fainter objects just becomes unfeasibly long - and of course the ADU capacity gets filled up and the dynamic range diminishes. The BRAIN is a very useful tool to do 'what if' experiments with and it is interesting to play with the effect of sky brightness.

Tim
Post Reply