Are Fast Telescopes Really Better? | Demystifying Astrographs
00:00 - Introduction 00:27 - The Hype Around Fast Astrographs 00:59 - Focal Ratio 02:18 - How To Compare Telescopes? 03:43 - Two Telescopes with the Same Focal Length but Different...
The Space Koala Astrophotography by Luca Bartek
Honestly, best Astrophotography channel on YouTube. I don't say that because she's attractive, I say that because she explains things 10x better than anyone else.
This one needs watching a number of times,, she talks as fast as my wife, but at least this is interesting đ
I was always confused why we were talking about the speed of telescopes since that should only depend on the size of the aparature.
I think it's interesting that the photography world focuses entirely on focal length & completely disregards aperture, while astronomy is the opposite.
Thanks!
Ansel Adams had his highest quality photos at f64! Try that. My fast lenses are always less quality in the corners. Always.
F64 is insane haha! But yeah I believe it. Itâs always a tradeoff!
I had a RASA and really enjoyed it. Lots of fun imagining at f/2 but I was never happy with my stars. Weird shapes and issues with tilt. I just picked up a ASI2400MC and I am going to enjoy that on my f/7 C11. Big 6um pixels.
I live in PNW and get so many cloudy nights that I sold my F7 and got an F3 Newt. Gotta make the best of the few clear nights I have.
Congratulations. Just beacause I own a C8 and C11, I reaaly love and appreciate your explanationsđ
go team SCT!
Your explanations are just fantastic. Thank you Luca.
What I wonderful discussion, enough details and math to be intuitive and solid, without losing you.
As usual, brilliantly explained and easily understandable - youâre a laser scalpel for marketing BS
Yeah her understanding is superbly concise.
Great topic! Because of my DSO preferences I always liked the the ~800 mm FL (APS-C sized sensor) very much. And guess what, I sold my F4 and F5 Newtonian telescopes and stick to my small 107/700 Apo since its the easiest to use.
the best scope is the one you're going to use the most!
I greatly appreciate your deep knowledge and are willing to bring use near your knowledge level.
Your explanation really put everything into focus (pun intended). Thanks. Subscribed
Well I guess that we always have to compare apples to apples = same resolution, same field of view, same everything. All other comparison doesn't really make sense. And if we consider the sky is equally illuminated, meaning homogenous, then a narrower field of view collects less photons at a given aperture. No ? So we need more aperture to collect the same quantity of photons then a larger field of view at a given image circle/sensor size. One can take the equation the other way : considering the image circle. If we stretch the image circle of a given tube aperture, then as the light quantity is given by the aperture, the biggest the image circle, the less light / cm2. That is what happens with barlow lenses. As soon as one uses a barlow lens, we increase the image circle, so we lower the photons per cm2 and we increase the F-number. On a given sensor size, a given image circle, a given field of view (magnification) the largest the more photos collected. If you start to change any parameter in the equation then the comparison doesn't mean anything. In the other way, if your scope has an image circle for a FF size, if you use a reducer, meaning conecntrating the same quantity of photons in a smaller image circle (let's say APSC) then each cm2 of the image circle will contain more photons. If we consider the image circle is the same, then for the same image (resolution, framing) a faster telescope will collect more data in the same amount of time. But if we want the same framing from two differents field of view then we compare apples to oranges. So the conclusion is : all things neing equals, a faster telescope is always better. And we can add to that that a larger mirror will provide more definition. If we compare apples to oranges then things are differents.
I agree with a lot of what youâre saying except the small contradiction between saying âsame everything, same field of viewâ vs. âsmaller fov collects less lightâ. It doesnât when you focus on the same field of view. The wider FoV collects more light from *additional* parts of the sky but you throw all that away if you crop to the same FoV which is what you have to compare as you say. Between 2 telescopes of the same aperture, the faster one will outperform the slower one at any FOV that is larger than the native FOV of the slower scope (with the same sensor). If you go at the native FOV of the slower scope or smaller, they perform exactly the same way. In addition, the faster optics come with a few tradeoffs, such as needing special filters, more careful focus, collimation, better field flatness etc. more importantly, optically they provide worse contrast in the same spatial frequencies and therefore perform worse at resolving smaller structures. I recently did a video on this last part recently called the journey of a photon
magnificently informative video
Hi, thank you for this video. There is something I'm not really understanding for the "same aperture, different focal length" case. If they have the same aperture, considering an identical sensor, there should be as many photons per pixel in both case. So why is the longer focal length dimmer? The "photon density" is the same and there are more total photons of the target as it covers a larger part of the field of view, no?
I think you got it quite right - the point is exactly that we get exactly the same number of photons from the same light source (e.g. galaxy). However, the longer focal length magnifies this more i.e. distributes the same amount of light over a larger surface - this is why it becomes dimmer
@the_space_koala Ooh of course I think I got it now. What matters to us is the light from the DSO so if we spread it around the sensor with higher magnification, the photon density *from* the DSO is indeed decreased.Thanks a lot!
The earth's atmosphere transparency limits the resolution of the telescope to 1.5 to 2 arc. seconds in the most cases. In this context everything above 150 mm aperture is a waste of money. The only reason you would need a larger aperture is saving time.
This... is just straight up wrong. I don't normally comment all that much, but I really want to clear this up. Faster systems are always better (of course if we exclude all the difficulties with using a lower f-stop, so narrower focus, a need for tighter correction...). They gather more light. The snr increases faster on a faster telescope than a slower one. Here's why: 4:42, you state that the rate of light collection isn't the speed of the scope. What? Of course it is. You also state it has to do with the amount of light gathered. Yes, the amount of light gathered over some unit of time is the speed of the scope, by definition. You also fail to understand that both upsampling and downsampling change the snr of the image, by a huge factor, usually (for cmos cameras), 2, not 4, because of the randomness of read noise. In order to match the FOV of two images/the pixel scale that have a different FOV, you have to crop one (so remove data that the telescope collected). For your example of f/4 and f/8, you'd lose out on 3/4 of the light by cropping the f/4 image to the samo FOV as f/8. Yes, you would end up with the same image, but you have to account for the cropped data as well. I'm genuinely surprised no-one in the comments mentioned this. I don't fell like writing more, but your other points are dismantled by the same logic (binning and drizzling also increase/degrade the snr, therefore the singal)... Faster scopes ALWAYS gather more data for any unit of noise. With the f/4 and f/8 scopes, the snr of the f/4 will be 4x the snr of the f/8. That's because, while the read noise stays the same, the signal increases four-fold. There is a reason people don't use f/20 refractors, even theough they don't really suffer from abberations. This sounds like a hate comment, but treat it more like a wake-up call. Your logic is really skewed, what can I say. Go over all your points and try to think about what all the pixel operations do to the overall signal and noise.
I never mentioned SNR but talk about amount of light collected, that is because we are focusing on light gathering power only. At 3:05 I actually start out with a disclaimer stating that in todayâs comparison we are focusing on the light gathered and do not consider any losses related to binning/drizzling etc. I get your point about binning producing a worse SNR than a faster scope on the same pixels. However, on modern cmos cameras read noise is truly negligible (and on CCDs this statement doesnât even hold true). Itâs a property of whatever sensor you choose to use and cannot be generalized when weâre speaking about the performance of the optics. In either case in low-light conditions, the noise in our images is dominated by the randomness of the incoming signal itself. This is so much more significant than the read noise from the sensor, which becomes almost negligible in comparison. Binning is just a tool in the calculation to help show the relationship between the amount of signal gathered in each case - you can replace a 2x bin by larger pixels in your thought process if you like. More importantly, photon noise will be equal for the same area of the sky with both the faster and the slower telescopes if you use the same diameter. But if you *really* want to focus on read noise, you can argue that to get the same detailed image with a faster scope, you have to upsample by drizzling and so youâre degrading the SNR that way. Also not sure why youâre saying I miss the point of throwing away data while cropping, when that is actually my whole point. A faster telescope has a benefit only as long as you want the larger field of view >> the faster telescope means we get to see a wider FOV and this is my main point. Iâm definitely not missing this ;) What you wrote I say at 4:42 is not actually what I say at all. I say the faster scope in that example gathered more light simply because of the larger mirror, not because of the f/ ratio.
@the_space_koala Hello, about 4:42. A larger mirror/lens means a lower f/ ratio, given the same focal length. Saying a telescope has a larger diameter and saying it has a lower f stop is the same thing. Let's assume we are talking about telescopes of the same focal length here. The faster scope only has benefit if you want a larger FOV, of course, but there is no downside to using a faster system (again ignoring optical difficulties). And again, both binning and drizzling, when done in post-processing increase/decrease the signal by a factor of 2, not 4, as you mention. A ccd will have an increase of 4 when done in-camera, while a cmos won't (virtually all modern astro cameras use cmos sensors). You do say you won't consider the losses of drizzling at 3:05, but you totally should. These losses determine whether your point is correct or wrong. Most people (myself included) didn't even notice the "disclaimer", so how can a beginner know what reality looks like if you only show the most basic examples? You say "a property of whatever sensor you choose to use and cannot be generalized when weâre speaking about the performance of the optics". Then why are you talking about sensors at all? I get your point that sensor technology doesn't matter when looking at the gathering power of a scope, but you have to take in into account if you actually use sensors. I think we agree on virtually all points, I just think you failed to properly express your ideas in the video. After reading your comment, I can see you actually know what you're talking about. I still think you shouldn't have disregarded binning/drizzling losses, they are huge. The feeling I get from your video is that a faster scope is almost always as good a slower one, and I'm guessing most people would get the same idea. Of course we both know this is not true, you should've pointed out multiple times that you're talking about wanting the same FOV at different focal lengths. Again, I wasn't completely sure that even was your point on my first watch.
when you do the post-processing steps mentioned, you do increase/decrease the signal AND noise, not the signal-TO-noise ratio by a factor of 4, not 2. You are thinking of SNR. Since we (well I) said we're not considering losses coming from the digital side (read noise), when you do these operations the results will be the same as the original optics we were comparing to. For example, if I downsample my f8 scope to match my f4 scope (in terms of SNR only, not FOV) - the resulting SNR that is original signal +- variability of signal will be identical - as I collected the same amount of data with the same surface area of the same subject. Yes if you do want to start thinking in terms of read noise, you would increase the noise by binning, but if you actually look at numbers of sensors, this is (1) near negligible on last generation CMOS sensors (2) nonexistent on CCD sensors (3) you cannot generalize because each sensor will be different