Barnard 150 Esprit vs RASA

[gorann]

17 minutes ago, Rodd said:

I was speaking more in terms more of resolution than speed.  Aperture is aperture, regardless of F ratio.   Regarding apertuire alone, I would raher have a diffraction limited 6" refractor than a 7" RASA.  The 11 has more of an obstruction, but it has more aperture--so it will still be greater as far as aperture.  In terms of resolving stars, a fast system with a small focal length will not be as good--even though it will be much faster.  One really can't compare the RASA to a 6" refractor because they do completely different things.   The refactor will resolve stars and details at a longer focal length while the RASA will collect lots of signal over a wider FOV with a shorter focal length.  The real compariosn is between a RASA and a fast refractor (4"), or large refractor reduced (I am not aware of a 6" refractor that has a native FR of 3 or 4).  An apt comparison is the FSQ 106 with .6x reducer (F3)--or the VSX 100 (F3).  I suppose the epsilon would be a good comparison too.--but not your typical 6" APO.   The key point is to compare systems of equal resolution (or nearly) in arcsec/pix.  

Yes, I think we agree that the RASA 8 and a 6" refractor both have their places and do slightly different things. I think the RASA shines when time is limited. Regarding a RASA 11, it will collect about twice as many photons, but deliver them over a much larger imaging circle (the area in the circle is 2.7 times larger) so if you want to  collect a decent part of the photons you need a 24 x 36 mm sensor. With a 24 x 18 mm sensor you will collect most of what a RASA 8 is collecting but only a fraction of what a RASA 11 is collecting, so unless you have a big sensor camera I don't think there is much to gain with a RASA 11. It also weighs nearly three times more (19.5 vs 7.7 kg)  and costs the double.

[Rodd]

20 minutes ago, gorann said:

so if you want to  collect a decent part of the photons you need a 24 x 36 mm sensor. With a 24 x 18 mm sensor you will collect most of what a RASA 8 is collecting but only a fraction of what a RASA 11 is collecting,

I agree that the 11 will cover a larger FOV.  However, the 11 will collect more photons per pixel.  More photons are collected everywhere--not just at the edges.  That is why a larger aperture scope provides higher resolution (at similar focal lengths) than smaller aperture scopes.

[gorann]

1 hour ago, Rodd said:

I agree that the 11 will cover a larger FOV.  However, the 11 will collect more photons per pixel.  More photons are collected everywhere--not just at the edges.  That is why a larger aperture scope provides higher resolution (at similar focal lengths) than smaller aperture scopes.

not sure about that Rodd, you may be right, but the photons collected will be spread out over more pixels as I understand it. Maybe our SGL oracle @vlaiv can explain?

[Rodd]

2 minutes ago, gorann said:

not sure about that Rodd, you may be right, but the photons collected will be spread out over more pixels as I understand it. Maybe our SGL oracle @vlaiv can explain?

No--the photons that go into a feature of a nebula will always go to that feature of the nebula from any scope regardless of aperture.  Lets say the nose of the horsehead.  in a 24" scope, more photons will go into the horses nose than with a 4" scope.  That is why it will appear more detailed

[vlaiv]

3 minutes ago, gorann said:

not sure about that Rodd, you may be right, but the photons collected will be spread out over more pixels as I understand it. Maybe our SGL oracle @vlaiv can explain?

I think that we are talking about couple of things here, so let's leave details (resolving power or the telescope) aside for now and concentrate on SNR, shall we?

Here, things are fairly simple - we need to fix one variable and see how other variable contributes. Let's fix sampling rate first.

If we fix sampling rate - so both scopes sample at the same resolution - be that for example 2"/px - given all else is equal (sky quality, sensor QE, read noise, filters used ...) - larger aperture wins. Here we should really calculate clear aperture equivalent - by taking into account central obstruction, air/glass surfaces (hopefully coated) and reflectivity of mirrors (regular, enhanced, dielectric ...).

If we fix aperture - larger sampling rate wins. At same aperture 2"/px will be faster than 1"/px.

If we fix pixel size (by using same camera and not binning) - F/ratio wins, faster F/ratio will be faster than slower F/ratio.

Fixing FOV is not meaningful for example - as it will depend on further things (like how many pixels you divide that FOV into and what aperture you use).

Back on resolution / resolving power. In long exposure imaging, 3 things determine resolution of the image provided that distortion of optics is below this level. Seeing, quality of tracking / guiding and aperture size.

I purposely used term "distortion of optics" rather than "diffraction limited". Most telescopes are diffraction limited only on or close to central axis and once you start introducing field flatteners / coma correctors and such - you are compromising diffraction limited performance. In any case - as long as blur from optics is significantly smaller than that of atmosphere / guiding above holds

There are optical systems where above does not hold - like camera lenses - which might be sharp for day use but are far from diffraction limited. I suspect that RASA is similarly not as sharp - which is not really much of a problem for it's intended use - low resolution wide field imaging.

Did I help with above? Why do I have feeling that although what I've written is correct and informative - it did not help much to this discussion?

 

[tomato]

1 hour ago, vlaiv said:

Did I help with above? Why do I have feeling that although what I've written is correct and informative - it did not help much to this discussion?

 

Vlaiv, since I have been on SGL(2014), I cannot recall a single post you have made that has not made a helpful contribution to the discussion. When you talk about the physics of optics and imaging, I listen. I don’t always understand it, but I do listen👍

[gorann]

5 hours ago, vlaiv said:

I think that we are talking about couple of things here, so let's leave details (resolving power or the telescope) aside for now and concentrate on SNR, shall we?

Here, things are fairly simple - we need to fix one variable and see how other variable contributes. Let's fix sampling rate first.

If we fix sampling rate - so both scopes sample at the same resolution - be that for example 2"/px - given all else is equal (sky quality, sensor QE, read noise, filters used ...) - larger aperture wins. Here we should really calculate clear aperture equivalent - by taking into account central obstruction, air/glass surfaces (hopefully coated) and reflectivity of mirrors (regular, enhanced, dielectric ...).

If we fix aperture - larger sampling rate wins. At same aperture 2"/px will be faster than 1"/px.

If we fix pixel size (by using same camera and not binning) - F/ratio wins, faster F/ratio will be faster than slower F/ratio.

Fixing FOV is not meaningful for example - as it will depend on further things (like how many pixels you divide that FOV into and what aperture you use).

Back on resolution / resolving power. In long exposure imaging, 3 things determine resolution of the image provided that distortion of optics is below this level. Seeing, quality of tracking / guiding and aperture size.

I purposely used term "distortion of optics" rather than "diffraction limited". Most telescopes are diffraction limited only on or close to central axis and once you start introducing field flatteners / coma correctors and such - you are compromising diffraction limited performance. In any case - as long as blur from optics is significantly smaller than that of atmosphere / guiding above holds

There are optical systems where above does not hold - like camera lenses - which might be sharp for day use but are far from diffraction limited. I suspect that RASA is similarly not as sharp - which is not really much of a problem for it's intended use - low resolution wide field imaging.

Did I help with above? Why do I have feeling that although what I've written is correct and informative - it did not help much to this discussion?

 

Thanks Vlaiv,

so what is your conclusion regarding RASA 8 vs RASA 11. The RASA 11 has 2 x the light gathering area but spreads it at 2.7 x the imaging circle area. To me that means less photons per pixel (with the same camera) and much smaller FOV for that same camera. Would it still give more resolution?

[vlaiv]

37 minutes ago, gorann said:

Thanks Vlaiv,

so what is your conclusion regarding RASA 8 vs RASA 11. The RASA 11 has 2 x the light gathering area but spreads it at 2.7 x the imaging circle area. To me that means less photons per pixel (with the same camera) and much smaller FOV for that same camera. Would it still give more resolution?

I new I did not get the question

Ok, got it now.

First we need to understand how telescopes work - I know, such a phrase, but indeed - above question and reasoning suggest that there are still some things that people overlook.

Here is "simplified" model.

Telescope is device that brings light coming from certain angle to a single point. All light that falls on its aperture coming from a single angle will end up in a single point. This is important thing to understand.

All photons coming from angle A1 that are collected by aperture (size matters here) will end up in single point. All photons coming from different angle A2 will end up in a different point. Here is diagram explaining this perhaps better:

Green lines represent photons coming from one direction, while red lines represent photons coming from a different direction - different angle.

Important point:

Every single point on focal plain receives photons from whole aperture. Points on focal plane do not split aperture somehow, each of them gets full aperture - as if aperture represents funnel for every single point on focal plane that works in parallel. What makes the difference is angle at which photons arrive. This is important thing, remember it for later. Aperture is just size of the funnel. Larger funnel, more photons each point on focal plane gathers simultaneously.

Larger FOV will not spread light from same aperture in comparison to smaller FOV - it will just add more points to gather light "in parallel" (like having more processors in a computer).

What about pixels, you ask?

Pixels join/spread angles not aperture. Angles in the sky. Every single point gets photons from just one and only one angle - exact number like 3.212331313.....

Pixel on the other hand collects all photons from part of the sky - all distinct angles that end up in that region. Say you have 1"/px. Single pixel will gather angle 3.21212... and 3.33434 and 3.543434...  everything between third and fourth arcsecond. Another pixel will gather everything between fourth and fifth arc second and so on.

More sky one pixel covers - more photons from all different "individual angles" will add up to form a signal in that pixel. Higher the signal - better the SNR.

But pixel size is not alone that determines how much of the sky is mapped to single pixel - there is focal length as well.

Now that we understand this, let's see if we can compare RASA8 and RASA11. For purpose of comparison we will assume that both RASA8 and RASA11 have same F/speed when all losses (central obstruction and coatings and all things) are accounted for - let it be F/2.2 or F/2 or F/2.6 - it does not really matter much.

Case 1 - using the same camera without pixel binning:

- They will be equally fast

- RASA8 will image larger FOV then RASA11

- RASA11 will resolve more, under assumption that optical aberrations in both scopes are the same (not manufacturing defects - but design spot diagram) simply because it has more aperture (airy disk is smaller and any aberration will be correspondingly smaller) - provided that mount and seeing are the same and there is no under sampling in both scopes.

Case 2 - different pixel size - matching resolution in arcsec/px:

- RASA11 will be faster

- FOV covered will depend on each camera sensor size and respective focal length - any combination is possible RASA8 > RASA11, RASA11>RASA8 and RASA11 = RASA8

- RASA11 will resolve more provided that there is no under sampling, if there is under sampling - both will resolve as allowed by pixel scale

Case 3 - matched FOV - this is really of no consequence because it tells us nothing

Case 4 - matched FOV and pixel scale - well that is just special sub case of case 2 (which can have any FOV relations) - same conclusions as Case 2

Does it make sense and does it now answer the question?

 

[gorann]

8 minutes ago, vlaiv said:

I new I did not get the question

Ok, got it now.

First we need to understand how telescopes work - I know, such a phrase, but indeed - above question and reasoning suggest that there are still some things that people overlook.

Here is "simplified" model.

Telescope is device that brings light coming from certain angle to a single point. All light that falls on its aperture coming from a single angle will end up in a single point. This is important thing to understand.

All photons coming from angle A1 that are collected by aperture (size matters here) will end up in single point. All photons coming from different angle A2 will end up in a different point. Here is diagram explaining this perhaps better:

Green lines represent photons coming from one direction, while red lines represent photons coming from a different direction - different angle.

Important point:

Every single point on focal plain receives photons from whole aperture. Points on focal plane do not split aperture somehow, each of them gets full aperture - as if aperture represents funnel for every single point on focal plane that works in parallel. What makes the difference is angle at which photons arrive. This is important thing, remember it for later. Aperture is just size of the funnel. Larger funnel, more photons each point on focal plane gathers simultaneously.

Larger FOV will not spread light from same aperture in comparison to smaller FOV - it will just add more points to gather light "in parallel" (like having more processors in a computer).

What about pixels, you ask?

Pixels join/spread angles not aperture. Angles in the sky. Every single point gets photons from just one and only one angle - exact number like 3.212331313.....

Pixel on the other hand collects all photons from part of the sky - all distinct angles that end up in that region. Say you have 1"/px. Single pixel will gather angle 3.21212... and 3.33434 and 3.543434...  everything between third and fourth arcsecond. Another pixel will gather everything between fourth and fifth arc second and so on.

More sky one pixel covers - more photons from all different "individual angles" will add up to form a signal in that pixel. Higher the signal - better the SNR.

But pixel size is not alone that determines how much of the sky is mapped to single pixel - there is focal length as well.

Now that we understand this, let's see if we can compare RASA8 and RASA11. For purpose of comparison we will assume that both RASA8 and RASA11 have same F/speed when all losses (central obstruction and coatings and all things) are accounted for - let it be F/2.2 or F/2 or F/2.6 - it does not really matter much.

Case 1 - using the same camera without pixel binning:

- They will be equally fast

- RASA8 will image larger FOV then RASA11

- RASA11 will resolve more, under assumption that optical aberrations in both scopes are the same (not manufacturing defects - but design spot diagram) simply because it has more aperture (airy disk is smaller and any aberration will be correspondingly smaller) - provided that mount and seeing are the same and there is no under sampling in both scopes.

Case 2 - different pixel size - matching resolution in arcsec/px:

- RASA11 will be faster

- FOV covered will depend on each camera sensor size and respective focal length - any combination is possible RASA8 > RASA11, RASA11>RASA8 and RASA11 = RASA8

- RASA11 will resolve more provided that there is no under sampling, if there is under sampling - both will resolve as allowed by pixel scale

Case 3 - matched FOV - this is really of no consequence because it tells us nothing

Case 4 - matched FOV and pixel scale - well that is just special sub case of case 2 (which can have any FOV relations) - same conclusions as Case 2

Does it make sense and does it now answer the question?

 

Thanks a lot Vlaiv! Then you confirm what Rood argue🥴, that the RASA 11 will resolve more with the same camera due to a larger aperture. But it is at the cost of a smaller FOV. Then the question is if it is noticible, worth the money, worth the smaller FOV and all the photons lost outside the sensor😀.

[vlaiv]

1 hour ago, gorann said:

Thanks a lot Vlaiv! Then you confirm what Rood argue🥴, that the RASA 11 will resolve more with the same camera due to a larger aperture. But it is at the cost of a smaller FOV. Then the question is if it is noticible, worth the money, worth the smaller FOV and all the photons lost outside the sensor😀.

I think it is worth it, provided that one has the mount and money for it and wants to go down that route.

Fact that RASA11 has larger corrected and illuminated circle means that it has potential to be faster scope for the same FOV.

Faster imaging + added resolution, what is not to like?

Thing is - you don't have to pair RASA11 with the same camera you pair RASA8, but you can keep the pixel size the same, or change it to suit your needs.

Here is an example of fields of view - red is RASA11 with ASI6200 (as it has 43.3mm imaging circle), yellow is RASA8 with ASI294mm (as it has 22mm imaging circle).

RASA8 + ASI294mm = ~2.4"/px

RASA11 + ASI6200 / bin2 = ~2.5"/px

Second one will obviously be faster due to added aperture. How much will additional resolution show at ~2.5/px - well that depends on spot diagram of the telescope.

I managed to find official spot diagram for RASA11:

Box size is 9µm and at 620mm FL that equates to 3". Add to that seeing and tracking issues - I don't think that RASA11 can deliver 2.5"/px sharpness, but 3.75"/px certainly (that would be x3 bin of ASI6200 or about 10µm pixel size).

[gorann]

Interesting @vlaiv!

One comment about the imaging circle of the RASA 8. 22 mm is what they promise to be fully corrected but Celestron say that the usable field is 32 mm with "only minimal performance loss at edge of FOV". That appears to be true at least in my sample. I use it with an ASI2600MC APS-C that has a diagonal of 28.3 mm and there is very litte vignetting and stars are quite ok in the corners. So a comparison of FOV with the RASA 8 and ASI2600 vs RASA 11 and ASI6200 gives identical FOVs.

 

At FLO the RASA 8 system (scope + camera) will cost you 3732 GBP and the RASA 11 (scope + camera) 7838 GBP. Then there is the possible additional cost for a mount that can handle about 20 kg of scope and camera (RASA 11) compared to 10 kg (RASA 8).

Here is my first light with the RASA 8 and ASI2600MC, which happened to be 2.6 hours of M31 (and before I realized that I could arrange the camera cables in a circle to avoid star spikes)

Edited December 12, 2020 by gorann

[vlaiv]

1 hour ago, gorann said:

At FLO the RASA 8 system (scope + camera) will cost you 3732 GBP and the RASA 11 (scope + camera) 7838 GBP. Then there is the possible additional cost for a mount that can handle about 20 kg of scope and camera (RASA 11) compared to 10 kg (RASA 8).

RASA11 cost does not seem justified - since you can purchase two RASA8 scopes for same amount of money, put them in parallel operation and get even faster system

After examining both your image and spot diagram that I posted - I've somewhat changed my opinion of RASA systems.

I believe them to be very good scopes for their intended use - they are like Samyang 135 F/2 of astronomy world. Indeed, they are not diffraction limited (in sense that you would not expect planetary images with resolution that 8" or 11" is otherwise capable of providing) - but they are sharp enough to produce very good wide field images.

In fact, I think I was wrong to say that they would not be sharp at 2.5"/px, as your image above at 2.7"/px clearly shows - that is sharp enough and can certainly compete in sharpness with fast systems otherwise used to image M31 at that resolution (like 80-100mm APO scopes, 130mm hyperbolic newtonians or mosaic with 6" F/4 regular newtonian or similar).

[gorann]

1 hour ago, vlaiv said:

RASA11 cost does not seem justified - since you can purchase two RASA8 scopes for same amount of money, put them in parallel operation and get even faster system

After examining both your image and spot diagram that I posted - I've somewhat changed my opinion of RASA systems.

I believe them to be very good scopes for their intended use - they are like Samyang 135 F/2 of astronomy world. Indeed, they are not diffraction limited (in sense that you would not expect planetary images with resolution that 8" or 11" is otherwise capable of providing) - but they are sharp enough to produce very good wide field images.

In fact, I think I was wrong to say that they would not be sharp at 2.5"/px, as your image above at 2.7"/px clearly shows - that is sharp enough and can certainly compete in sharpness with fast systems otherwise used to image M31 at that resolution (like 80-100mm APO scopes, 130mm hyperbolic newtonians or mosaic with 6" F/4 regular newtonian or similar).

Thanks Vlaiv! I have actually thought of the possibility of a dual RASA 8 rig, since my iOptron CEM70 that currently holds my RASA 8 should be able to support a second one. But I hold my horses and save my coins until there is a mono APS-C produced and which I could put on the second scope.

Edited December 12, 2020 by gorann

[gorann]

On 10/12/2020 at 15:38, tooth_dr said:

Nice comparison.  I agree that is quite a result for just over 3 hours with the RASA8 scope, the colour has really been captured.   What about using the luminance of the Esprit with the colour of the RASA?  

 

On 11/12/2020 at 13:49, Laurin Dave said:

The total area of glass collecting light is nigh on identical at 25529mm2 for the Esprit duo and 25057 mm2 for the RASA (taking into account the 90mm diameter camera).   You could speed up the colour capture by going mono on the Esprit100 or as Adam has suggested use RASA colour data with Esprit150 luminance... be very interested to see that..

Dave

ps... my ASI2600MC and Askar200 should arrive tomorrow... what is the world coming to!

 

So acting on your requests (thanks!), I have now made an image with Eprit luminance on top of the RASA RGB. You find it here:

 

[maxchess]

Hi, I have just got a RASA 8 that I got so I could maximise my time under UK cloudy skies. Still cloudy so only one quick image so far. Going back to the first two images that started this thread I notice that  for the refractor the "ASI071MC collected the RGB (56 x 5 min). Totally 8.2 hours" for the RASA 8 it was  " 19 x 10 min, so totally 3.2 hours " .  So I wondered what the impact of the very long exposure time of 10 min per exposure might be?  Could the detail be blown out? Maybe star bloat is the reason for poor separation of smaller stars noticed by another user. Many RASA8 users are going for 20-30 second exposures which is the equivalent to somewhere around 180 secs for a F 6.3 refractor in photon gathering (rating the RASA 8 effective f ratio as f2.2 adjusted for central obstruction).   Shorter exposures also give you the benefit of reduced tracking error.

To me the benefit of the RASA 8 is all about photon collection. I believe I can get  around 4 hours worth of photons in something like 30 mins compared with a f 6.3 refractor. I have other scopes for other FOV but for a given sensor this covers me for the medium wide FOVs.

Or I could be completely wrong.

Max 

Edited August 30, 2021 by maxchess

[vlaiv]

18 minutes ago, maxchess said:

To me the benefit of the RASA 8 is all about photon collection

Not sure if you got it right there. Do you think that RASA8 will magically collect more photons than any other 8" scope?

Number of photons falling on 8" of aperture does not change when you change the scope.

8" F/8 RC will be equally fast as RASA8 if you utilize the same "effective" pixel size in both (pixel size in units of arc seconds in the sky and not micrometers). Only difference will be in FOV between the two. That is the strength of RASA8 - covering more sky in the same amount of time than would be possible with longer FL scope. Tradeoff is sharpness / detail. Longer FL scope would be able to resolve more if it has smaller spot diagram (and in most cases that is true). You can't do planetary imaging with RASA8 for example.

22 minutes ago, maxchess said:

So I wondered what the impact of the very long exposure time of 10 min per exposure might be?

Two major things to consider when choosing exposure length are:

- impact of read noise. You need long enough exposure so that one of the noise sources that changes with time (shot noise, LP noise or thermal noise - each grow with exposure time) swamps fixed amount of read noise.

- full well capacity. You can and will saturate pixels in single exposure. That happens with bright stars and sometimes even with bright parts of targets. Luckily, there is simple way to deal with that - just take small set of short exposures at the end of session and use those values scaled to fill in saturated parts of original exposure.

There are other considerations that are not as important - like thrown away data due to tracking issues or maybe wind (sudden wind gust can ruin either 30s or 15minutes - depending on your sub duration). Some algorithms work better when they have larger number of subs to work with (sigma rejection and similar).

[maxchess]

4 minutes ago, vlaiv said:

Do you think that RASA8 will magically collect more photons than any other 8" scope?

I think it's about matching equipment to targets and FOV. Then for me maximising imaging opportunity.

 If I am looking for a wide FOV equivalent to 400mm for given sensor then by choosing a 400mm RASA at f2, rather than a say a William Optics GT-71, FL=420, f5.9,  I can capture the same number of photons 7 times faster for a specific FOV/ set of targets. So with the RASA I can do in 34 mins the equivalent of 4 hours integration time with this specific refractor.

I also have an 8" C8 for small FOV targets but that has a native focal length of 2000mm at F10, which I can bring down to a 1200mm with a focal reducer, (or even use a hyperstar,) but I use that for a different range of targets. 

In between I have a Williams132 with a focal length of 900mm (720 with FR) .  If I could afford it I might buy a  RASA 14 with a FL = 790mm but that's about £13,000

Exposure length is very interesting Robin of Sharpcap provides detailed calculations and an interesting video that shows that effective exposure times are often a lot shorter than commonly believed. Obviously heavily equipment dependant.  But with the RASA it is so fast that there is a real danger of exceeding the full well capacity.  

 

 

[Rodd]

I too prefer the top version.  However--there are many factors at play--obviously sky conditions.  But also, settings (exposure times, gains--all that)--the different systems will have different optimum settings.  Finally, processing plays a huge role in the appearance of the final image.  It would be interesting to see a sub comparison--and a linear stack comparison.  

But regardless, it looks like you are having a ball with your scopes.