Compare 'light gathering capability' 200mm reflector vs 100mm refractor

[Robculm]

Hi, apologies for what's probably (to some) a very obvious question!

I'm tentatively looking at APO's, such as the Esprit 100ED as an 'alternative' to my 200PDS. The Esprit 100ED in particular as I'm looking for a slightly bigger FOV (but not too much), it's a good weight (little less than the 200PDS so should be OK on my HEQ5) and price isn't 'too' crazy.

Back to the key question though, is my understanding correct that I basically get about x4 less 'light gathering' with the 100ED than with the 200PDS. OK, presumably a little less as there's no secondary obstruction, but even so, it's quite a significant number. I'm sure there are some factors related to the f number etc that I'm missing here?! But in general, does this mean I need x4 imaging time to achieve similar results?

Main reason I'm considering this that I'm constantly battling collimation / star shape with the 200PDS. But I don't want to spend £2k on a new set up only to be dissapointed with what I can achieve! Aside star shape, I feel I'm getting some reasonable images!

Would very much appreciate advice / comments please.

Note, here's my latest example, 2h of data on M33. So this is my 'starting point' as such and of course looking to improve upon!

Edited November 11, 2021 by Robculm

[vlaiv]

41 minutes ago, Robculm said:

But in general, does this mean I need x4 imaging time to achieve similar results?

No.

With 100mm of aperture you will need x4 imaging time to match SNR of 200mm only if you image at same resolution.

Imaging resolution depends on pixel size, focal length and any binning used.

Speed of the system can be described as "aperture at resolution". When comparing two setups - you need to compare them by value obtained with aperture surface and "resolution surface" - or how much sky surface each pixel covers (real or binned one).

Multiply the two to get number that you can use to compare things.

To put things in perspective - since you'll be using the same camera (one from your signature?) Esprit 100ED has four times less aperture (by surface) than 200PDS, but pixel size covers x3.3 more sky area (1000mm / 550mm or 1.818181... squared). In the end Esprit 100ED will have 3.3 * 0.25 = ~82% of the speed of 200PS.

 

[emyliano2000]

I was curious about this too as I will be using the same sensor on my AT106 at 690mm focal lenght and my 8" f4 newtonian. 

Emil

[Stickey]

Hi. 

That's one beautiful M33. Congratulations 👏.

But you say the problem is you want a bigger FoV. How much bigger? There are a number of online FoV simulators that you could use to work out what scope aperture would give you the desired FoV with your camera. 

As Vlaiv explained,  it's maybe not just as simple as aperture size. In addition to FoV simulators,  there also one that calculates Pixel Scale of a camera/scope combination. Apologies, I can't remember the name but you should be able to find it easily enough

[Ohgodwherediditgo]

2 minutes ago, Stickey said:

Hi. 

That's one beautiful M33. Congratulations 👏.

But you say the problem is you want a bigger FoV. How much bigger? There are a number of online FoV simulators that you could use to work out what scope aperture would give you the desired FoV with your camera. 

As Vlaiv explained,  it's maybe not just as simple as aperture size. In addition to FoV simulators,  there also one that calculates Pixel Scale of a camera/scope combination. Apologies, I can't remember the name but you should be able to find it easily enough

https://astronomy.tools/

[Robculm]

Thanks for the comments.

Hi Vlaiv, the headline number (~82%) sounds good - Thanks. But still not really clear to me how we get to that.

Yes, using the 800D, so 3.7x3.7um pixel size.

With the 200PDS, forgot to mention, I have the Skywatcher coma corrector, which is also a 0.9 reducer, so actual f is ~900mm. The astrotools FOV calculator tells me that's a resolution of 0.85x0.85" pp, FOV 1.4x0.95deg.

For the 100ED, 1.39x1.39" pp (FOV 2.32x1.55).

So around x2.67 more sky area, and around x1.93 resolution per pixel.

I understand that from a sampling perspective, .85x.85 is 'oversampling' and I'm binning 50% in processing (software) , 1.39x1.39 is a much better sampling rate and possibly doesn't need binning?

When you say can only compare if imaging at the same resolution, can I basically say that if I imaged the same target, let's say this M33 image for example, and I cropped the 100ED image to the same size as I have here from the 200PDS, I'd have lower resolution and ~ x4 worse signal to noise? So the 100ED only makes sense on larger targets where I'm using the 'whole' image?

Apologies again, maths is not my strongest suit!

Thanks,

Rob

 

 

Edited November 11, 2021 by Robculm

[vlaiv]

9 minutes ago, Robculm said:

Hi Vlaiv, the headline number (~82%) sounds good - Thanks. But still not really clear to me how we get to that.

Ok, so here it is in a bit more detail.

Aperture surface determines how many photons are captured. That part is clear - x4 larger aperture by size - means x4 more photons gathered (in reality that is not quite true because there is central obstruction and losses in optical train - like 94% reflectivity of each mirror and transmission of glass and reflections on air/glass surfaces - but those are nuances).

Other important bit is "effective pixel surface", and when I say effective - I mean how much of the sky is covered by single pixel. This is important because all photons coming from that part of the sky - land on that single pixel and are put on the same "heap" - more photons, larger the heap and stronger the signal - better SNR (all of this is per unit time - and we can't "compress" time - it flows the same for both scopes so we don't mention it).

In order to see what sky surface single pixel covers - you take sampling rate  / imaging resolution which is: 3.72µm pixel size on 550mm FL and 3.72µm and 900mm FL which give as you already mentioned:

0.85"/px and 1.39"/px

And we simply calculate how much there is sky surface in square that is 0.85" x 0.85" and 1.39" x 1.39" and take ratio of those surfaces.

1.39 * 1.39  / 0.85 * 0.85 = ~2.67

(without 0.9 reduction it was 3.3)

Fact that you are using OSC sensor - DSLR does complicate things a bit, but yes, you are right - you should bin x2 (or use super pixel debayer) in each case.

0.85"/px is oversampling for 8" scope and 1.39"/px will be oversampling for 4" scope under most circumstances.

You can and should bin both by x2 (or use super pixel mode) - but that won't change speed ratio.

20 minutes ago, Robculm said:

When you say can only compare if imaging at the same resolution, can I basically say that if I imaged the same target, let's say this M33 image for example, and I cropped the 100ED image to the same size as I have here from the 200PDS, I'd have lower resolution and ~ x4 worse signal to noise? So the 100ED only makes sense on larger targets where I'm using the 'whole' image?

What I'm saying is - imagine for example that you have x1.1 instead of x0.9 coma corrector - so you are working at 1100mm FL with 8" scope.

Now you have 1.4"/px for Esprit and 0.7"/px for 8" Newtonian. You bin x2 Newtonian data, but leave Esprit data as is and you will get the same sampling rate / same resolution (word resolution is used in many contexts and here I use it to mean sampling rate and not total pixel count).

In this case - you will end up with same sampling rate, potentially the same level of detail captured (seeing dominates difference in aperture) but SNR will be different - and it will be different by exactly the difference in aperture surface - because "pixel sky surface" will be the same (same sampling rate) - so that ratio will be 1 and will not "contribute".

Hence expression for speed "aperture at resolution" - because if you fix your resolution (again - sampling rate and not pixel count or other meanings of word resolution), it is only difference in aperture size that counts.

Of course - there will be difference in one more aspect - FOV captured with Esprit will be x4 larger because focal length is twice shorter.

If you want to capture that large FOV - esprit is actually not slower at all but equal. In order to capture that larger FOV with 8" scope - you'll need to do 2x2 panels. and you can therefore only spend 1/4 of time on each panel - so that cancels with x4 light gathering surface.

What small scopes are good for - is wide field stuff. There they are not slower than large scopes - but if you want to match sampling rate and you can fit whole target in FOV with larger scope - then large scope wins.

Makes sense?

[Robculm]

Thanks Vlaiv, yes, that's much clearer 🙂

Key point is that I need to think very carefully about the targets I want to image & that I should probably persevere trying to optimise my 200PDS performance for now!

Thanks again, really appreciate your time explaining in such detail.

Cheers,

Rob

[Robculm]

Coming back to this topic, but from a slightly different perspective.

I'm contemplating getting a Barlow lens to 'increase magnification' for smaller targets. Possibly planets (something I've not really attempted yet), but certainly smaller deep sky targets.

So how does this work in terms of gathered light / image resolution etc... The Barlow (lets say x2) gives me 1/4 of the field of view, but does 'all' of the gathered light make it in to that 1/4 area or am I again missing something on the mathematics here!

And in general, thoughts on using a Barlow (this would be with the 200PDS) for imaging? Don't seem to see much reference to this, so not sure if it's common?

Many thanks

Rob

[vlaiv]

5 minutes ago, Robculm said:

I'm contemplating getting a Barlow lens to 'increase magnification' for smaller targets. Possibly planets (something I've not really attempted yet), but certainly smaller deep sky targets.

That works good for planets but it is questionable thing for smaller deep sky targets. Do keep in mind that imaging planets require very different approach to imaging DSOs. It is so called lucky imaging technique - where one captures tens of thousands of frames in rapid succession (much like movie - and is often referred as to movie) - exposures are kept very short - like 5ms or so - all with goal of freezing influence of atmosphere in mind, and eventually choosing only the best frames out of all captured (often as few as 5% of total).

With DSO imaging - atmosphere is dominant component of possible resolution and here you can't really circumvent its influence like in lucky imaging.  That often means that you'll get larger but blurry image when using a barlow - much like using "software zoom" - or just enlarging image in software.

10 minutes ago, Robculm said:

So how does this work in terms of gathered light / image resolution etc... The Barlow (lets say x2) gives me 1/4 of the field of view, but does 'all' of the gathered light make it in to that 1/4 area or am I again missing something on the mathematics here!

Adding barlow to scope will increase working resolution by factor of x2 (in case of x2 barlow) - as focal length of system is extended by factor of x2. If we go by "aperture at resolution" as being speed - we can easily see that this combination will be "slower" simply because we changed resolution part.

This has to do with above - do you really need higher resolution - or to put it more precisely - can you achieve that higher resolution given your sky conditions and mount performance?

As is - you are already very high with resolution. Your camera has 3.72µm pixel size and 8" F/5 scope has 1000mm of FL - that gives you 0.77"/px. That is already much more than your mount and sky support.

With 8" in most circumstances you'll be able to achieve say 1.5"/px - 2"/px. On very special nights - you'll go below 1.5" - but how much really depends on conditions and your mount performance. With premium mounts one can hope to achieve 1-1.2"/px on the best nights - but not below that.

Adding x2 barlow will just make image larger without any real detail.

15 minutes ago, Robculm said:

And in general, thoughts on using a Barlow (this would be with the 200PDS) for imaging? Don't seem to see much reference to this, so not sure if it's common?

From above you can conclude: For planetary - yes, it's sort of mandatory given that your scope is F/5, for DSO imaging - no, as you are not exploiting pixel scale that you already have (and you need to bin data to get best results from your setup).

[Robculm]

Thanks Vlaiv, an excellent explanation as always 👍

Ha, yes, I've seen from other posts that the planetary imaging is 'very different'. It's not something for which I have a great desire, but they jupiter is always so bright & tempting! Maybe I will explore the Barlow (perhaps a higher power) just for some 'visual' kicks for now.

To be honest, I think I'm just getting itchy fingers as the weather is so darn rubbish and I can't get a decent imaging session on anything at the moment!!!

Thanks again,

Rob

[vlaiv]

5 hours ago, Robculm said:

Thanks Vlaiv, an excellent explanation as always 👍

Ha, yes, I've seen from other posts that the planetary imaging is 'very different'. It's not something for which I have a great desire, but they jupiter is always so bright & tempting! Maybe I will explore the Barlow (perhaps a higher power) just for some 'visual' kicks for now.

To be honest, I think I'm just getting itchy fingers as the weather is so darn rubbish and I can't get a decent imaging session on anything at the moment!!!

Thanks again,

Rob

I see that you have ASI120mm in your kit list - you can use that to do some mono shots of Jupiter and Saturn. Moon and Venus are also fair game for trying out planetary type imaging.

All of these can be imaged with barlow and ASI120mm, and that is probably the best way to get into planetary imaging - just to try it out and see how you fare.

[Robculm]

Thanks 😀. I hadn't considered that option. 👍