That doesn't make sense, does it? The complex relationship between SNR and integration. Part II.
Вставка
- Опубліковано 28 гру 2024
- Dr. Detail has been bothering me for a while about some SNR claims. Basically, some people in forums and in videos are claiming that 4x the signal only doubles speed. But is that true? In my last video, I demonstrated that 4x the signal = 4x the speed, both in raw terms and in gathering SNR. But the conclusions seem like a paradox (at any given integration time, an f2 system only has double the SNR of the f4 system). In this video, we take on that paradox!
Become a buymeacoffee member to help choose which products I review: www.buymeacoff...
Affiliate Links:
Skywatcher HEQ5 Mount amzn.to/3NVrVQi
Skywatcher 80mm F7.5 Doublet amzn.to/47mbfZi
ZWO EFW mini filter wheel amzn.to/48OxhoN
Optolong LRGB Filter Set (1.25") amzn.to/47r23Tw
Orion 0.8x Reducer amzn.to/47s64a4
USB RJ45 Cable for Controlling Mount amzn.to/4aPlIzu
Raspberry Pi to control mount amzn.to/48mTBWH
R Pi case with touch screen amzn.to/3HvdNKl
ZWO Autofocuser amzn.to/3Skg0hX
ZWO 294MM Pro Camera: bit.ly/3tZmjOo
HEQ5 Mount: bit.ly/3O7VQVz (mine is Orion Sirius, which is equivalent)
ZWO Filter Wheel: bit.ly/3Soo5lw
Optolong Filters (LRGB, SII, Ha, OIII): bit.ly/3SlEPtJ (the Ha and OIII narrowband filters are better than what I have)
Skywatcher Evostar 80mm: bit.ly/4bfVjev (equivalent lens)
USB RJ45 Cable: bit.ly/48FRj50 (connects mount to laptop directly; mine is a different brand, but this should do)
ZWO Autofocuser: bit.ly/425XUDB
0.8x Reducer: bit.ly/3tQlbg7 (mine is Orion, but this should be equivalent)
Join this channel to get access to perks:
/ @deepskydetail
Why can't objects with non-constant acceleration hold a conversation? ............................
They're always going off on different tangents!
Also, you can support me on buymeacoffee here: buymeacoffee.com/deepskydetail
@@deepskydetail because one of them is my wife?
Ok, seriously now.
You've got me so confused....
Just tell me ....
Do I want an f/2 or an f/4 telescope.
Just let me know. PLEASE.
I'm begging you.
@christopherleveck6835 The f/2 scope, all things being equal, collects photons 4x faster and also collects "SNR 4x faster" (of things like nebulae). Some people have claimed in other forums/videos that 4x the signal only doubles speed. They use 2x SNR argument as the reasoning. But, because doubling SNR with the slower scope always means you need to quadruple imaging time, the scope that has "2x the SNR" is still 4x as fast. Just like 4x the photons is 4x as fast.
If your main concern is collecting SNR as quickly as possible, then you'd most likely want the f/2 scope. Without knowing exactly what your goals are, I can't really say what you should or shouldn't do.
As a summary. Looking from SNR as a goal. Four times faster optical system will take all neccesary subs four times faster.
Looking from time perspetice when you have fixed amount of time to collect light. Four times faster opitcal system will take pictures 2 times faster.
Choose your poison :)
Yes! With the caveat that to get 2x the SNR you have to 4x the imaging time :)
Love the boat analogy! This is super educational. I've been forwarding your videos whenever anyone asks me about SNR, Integration time, etc. because as well as I can explain them, your videos get there 4x faster 😀
Thanks, Naz! I'm really glad it helps! I hope people find it useful, and let me know if things are unclear (or wrong!)
:)
Another fantastic series. Good job!
Thanks so much, I'm glad you liked it!
Excellent explanation. Thank you for taking your time doing this video. You made it very understandable. Take care.
I'm glad you found it helpful :)
I have to admit that I got utterly confused. Are the two videos just about contrasting these observations: (1) the duration required to collect S photons and achieve SNR=sqrt(S) with a f4 scope is 4 times the duration required to collect S photons with a f2 scope; and (2) for an observation of a given duration, the f2 scope gathers 4 times as many photons as the f4 scope and therefore twice the SNR? Or did I miss something?
It can all be pretty confusing. You've got a good summary of my ideas. The reason for the video is I've seen some fairly big names claim that f2 is only 2x the speed compared to the f4. They use "SNR" as the argument. But that argument is flawed. To understand why it's flawed is not obvious (i.e., the speed and deceleration of the systems is not constant).
So I wanted to show that the f2 scope is really 4x faster no matter how you look at it (assuming constant pixel size), from SNR gathering or actual photons hitting the pixels.
Also note: I don't necessarily think SNR should be used to examine the speed of s telescope. It's not intuitive. And I talked about that in the last video.
I'd also like to just say that your point 2 (the way I'm reading it and maybe i'm misinterpreting it) is saying exactly the same thing as point 1. 2x SNR is 4x the speed (either in photons per unit time or SNR per unit time) in the video's examples. SNR is not the same as SNR per unit time, just like distance is not the same as speed.
@@deepskydetail you are absolutely right, my two points are saying the same thing, but with 2 different perspectives (the reference being either SNR or duration).
What do you mean by "speed"? What is SNR per unit of time? What does 1 dB/s represent? How do you measure it? Since you seem to need both photons per unit of time and SNR per unit of time, what's the relationship between 1 photon/s and 1 dB/s?
That's what the first half of the first video deals with. SNR is not a measure of speed. The second half of the video then tries to force it into a measure of speed by looking at SnR gained per sub frame (i.e. "playing the devil's advocate" section).
OMG I think you made it worse....
I'm REALLY trying here.
It’s an interesting comparison. I run f/10, f/7, and f/1.9 on the same scope, which makes it pretty easy to get the differences. I think what might have helped with clarity would be to equate differences in terms of exposure lengths. That being said because at f/1.9 I would be pushing toward saturation FWD at long exposure lengths as I would take with f/7, for example. I think you covered it in the overall time comparison to speed but in practice I would have a great many more shorter exposures at f/1.9 than that of f/7. E.g. 240 vs 96 within the same nightly session time. Now, granted, at f1.9 I’m done, with f/7 I’m going to need 192-288 exposures to get to the same SNR. In that sense, the total number of exposures is nearly equivalent, depending on conditions.
In either case, we get to the same conclusion that f/1.9 is faster than f/7 in overall integration time to reach the same SNR.
Just saying the same thing in a different way 😊
Cheers!
Doug
Thanks, Doug! I'm always amazed at how versatile the Edge HD is! Great comparison! I agree, total exposure time would have been better. Just as long as people get that each sub was the same length of time, I'll be happy! :)
@@deepskydetailok, I have a non hd C11. I have a .63 focal reducer....
I can take more but shorter exposures with the scope set up with the reducer but need longer exposures and more of them without it?
And assuming that's correct....
Is that when using similar gain settings?
My biggest problem is being in the Pacific NW.
Here in Oregon I only got 14 nights so far this entire year that were even kind of worth shooting.
And while I can practice my automation stuff on a less than clear night...... When it's time to take a real picture I don't have time to screw around .....
And I've got a decent refractor too.
So I'm trying to be ready for the perfect night if it ever comes, and I can't be playing around with new settings.
I'm well past the point of trying things out.
I feel like I'm so close to making this all work. But I'm at a point where I feel like I'm over thinking it maybe.
And my images aren't getting better, they are getting worse.
I remember thinking I was never going to figure out how to get a 10 second exposure in focus with no star trailing....
Now I have taken 30 minute subs, all night long, in 3 filters with pinpoint stars and all the calibration frames.......
And I'm not getting the detail I should be because I'm not getting the settings correct.
I don't want to seem like I'm whining but there has to be other guys like me that have sorted out the technical aspects of getting the images but the theory is so over my head I feel like im drowning in formulas and math I struggled through in high school.
What I really need are settings that should give me a workable solution that I can tweak.
And a process to walk through tweaking them.
A baseline that gives me a result that can be analyzed and then a "now do this" to improve the next round. One setting at a time.
Because right now I'm trying to adjust gain and exposure time and total number of exposures and trying to make tweaks to all of them with each new session.
I can't believe how far I've come in a very short time.
And I needed this. I spent time in the middle east and nights laying in bed looking at the ceiling are horribly long.
Having something to do at night that I am happy to do alone has been a real gift.
I'm right on the cusp of greatness.
Seriously.
I made the most out of the few nights I had this year.
I shot 54 hours of the Veil Nebula.
LRGB and narrowband SHO.
And it looks pretty good ....
It could be sooo much better.
In photography, f2 is only twice as fast as f4 (one stop). Why is it 4 times in this video?
If you keep ISO the same, how much longer do you have to expose an image stopped down to f4 compared to f2? The answer is four times longer. F2.8 is one stop higher than f4. f2 is two stops higher.
@@deepskydetail yes, you are absolutely correct. Apologies, I didn't remember correctly and completely forgot about f2.8 :))
No worries! I find I have to remind myself all the time. f/stops have always been weird to me.
Another great video
Thanks! I'm glad you liked it. :)
Straight over my head 😅 what if I've got a seestar s50? What's the snr rate😊
That's a good question. It all depends! But the Seestar (I think) is an f/5 scope. You'd have to compare it with another scope with the same camera on the same target to find the "SNR rate." Although, as mentioned in my previous video, I don't think SNR is a good measure to figure out speed.
The video assumes people are taking their camera and putting it on another scope with a different f-ratio. So, it's really complicated!
100 % agree
Who controls the channel, Doctor Detail or Deep Sky Detail?
That's proprietary information 😅
@@deepskydetail 👀
You (and other channels) give too much importance to the "speed" and f/ , mixed with SNR. Its just a hype.
The only tradeoff is SNR vs. Resolution . A telescope might be faster but useless for planetary imaging.
Stop saying that its confusing, its not. All you need is to read the book "Deep Sky Imaging Primer" of Charles Bracken
If it's not confusing, then why are people in forums and other yt videos claiming quadrupling signal only increases speed by 2x, when that is clearly wrong?
Speed is indeed very important. If you say otherwise, you may never been through imaging really dim deep sky objects with F/10 telescope with really narrow dual narrowband filter, being pushed to gain 300 in order to get some signal through.
I repeat, it depends on what you want to do and the trade-off speed vs resolution, in parallel with your $budget. In 0:22 of the video it already says "confusing" and then goes on with a topic which is largely discussed in CloudyNights for example. The issue can be understood if one takes time to study it and do some Excel exercise, im not saying its easy. People want to understand this, QMechanics and general relativity watching YT videos while drinking a beer...Please read the book I mention in my first comment, its a gem