28 replies to this topic

[Jape]

I recently came across the new HAC125 from Skywatcher, which is praised for its very fast f/2 optics and extremely affordable price point.
As I explored its features, I started wondering: to what extent is its compatibility with smaller sensors a limiting factor?

 

I imagined two different setups:
1- An HAC125 (f/2) paired with an IMX585 sensor (1/1.2”, 12.8 mm diagonal, 3840×2160 resolution, 2.9 μm pixel size)
2- An Askar 80PHQ (f/7.5) paired with an IMX571 sensor (APS-C, 28.3 mm diagonal, 6252×4176 resolution, 3.76 μm pixels), with 2×2 binning

 

Both setups have a very similar FOV (considering FL adjusted for crop factor) and can be used to image the same targets:
- FL around 830 vs 900 mm FF equivalent

- Resolution: 3840×2160 vs. 3126×2088
- Pixel scale: 2.4″/pixel vs. 2.6″/pixel

But there are also major differences:
- f/2 vs f/7.5
- Aperture 125 vs 80 mm
- Pixel size 2.9 vs 7.52 micron

 

 

My question for the experts is: which setup is truly “faster” in terms of achieving a better SNR in the same time?
I believe the second setup is MUCH faster because of the much larger pixels (almost 3x even ignoring central obstruction, transmission, QE) but I might be wrong: why would there be so much hype for the HAC125 speed, price aside?

Edited by Jape, 22 April 2025 - 05:05 AM.

[MartinNI]

F2 vs F7.5 is 14 times more light collected with F2. That only is really huge. Pixel base size difference is negligible, and probably even binning.

Edited by MartinNI, 22 April 2025 - 05:59 AM.

[MartinNI]

Some real world results from amateur review

 

https://www.deepskyp...he-sw-72ed.html

 

So, the benefits MAY not be so obvious, but they are certainly there.

[acrh2]

I recently came across the new HAC125 from Skywatcher, which is praised for its very fast f/2 optics and extremely affordable price point.
As I explored its features, I started wondering: to what extent is its compatibility with smaller sensors a limiting factor?

 

I imagined two different setups:
1- An HAC125 (f/2) paired with an IMX585 sensor (1/1.2”, 12.8 mm diagonal, 3840×2160 resolution, 2.9 μm pixel size)
2- An Askar 80PHQ (f/7.5) paired with an IMX571 sensor (APS-C, 28.3 mm diagonal, 6252×4176 resolution, 3.76 μm pixels), with 2×2 binning

 

Both setups have a very similar FOV (considering FL adjusted for crop factor) and can be used to image the same targets:
- FL around 830 vs 900 mm FF equivalent

- Resolution: 3840×2160 vs. 3126×2088
- Pixel scale: 2.4″/pixel vs. 2.6″/pixel

But there are also major differences:
- f/2 vs f/7.5
- Aperture 125 vs 80 mm
- Pixel size 2.9 vs 7.52 micron

 

 

My question for the experts is: which setup is truly “faster” in terms of achieving a better SNR in the same time?
I believe the second setup is MUCH faster because of the much larger pixels (almost 3x even ignoring central obstruction, transmission, QE) but I might be wrong: why would there be so much hype for the HAC125 speed, price aside?

Since the pixel scales are almost the same, the largest difference will be in the aperture diameter of the two imaging systems. The larger aperture will produce a brighter image by a factor of (125/80)^2, so it will also have a higher SNR. 

[Jape]

F2 vs F7.5 is 14 times more light collected with F2. That only is really huge. Pixel base size difference is negligible, and probably even binning.

 

Since the pixel scales are almost the same, the largest difference will be in the aperture diameter of the two imaging systems. The larger aperture will produce a brighter image by a factor of (125/80)^2, so it will also have a higher SNR. 

 

Pixel area (similar pixel scale of 2.4" and 2.6") is 8.4 vs 56.6 μm². It is 6.7 times bigger, while aperture area is only 2.4 times bigger - (125/80)².

That makes me think that the number of photons collected per pixel and the SNR should be higher in the 80phq/imx571 setup.

[acrh2]

Pixel area (similar pixel scale of 2.4" and 2.6") is 8.4 vs 56.6 μm². It is 6.7 times bigger, while aperture area is only 2.4 times bigger - (125/80)².

That makes me think that the number of photons collected per pixel and the SNR should be higher in the 80phq/imx571 setup.

 

I don't know how to explain this any better than I already have. That was the simplest math that comes from the idea that when you are imaging the same chunk of the sky in a single pixel, then the amount of light that comes in will be proportional to (aperture diameter)^2.

 

You can go to the first principles and derive all of the relevant math, but here's the gist:

 

relative brightness of pixel ~ (fratio1/fratio2)^2 * (pixels size2/pixel size1)^2

 

In your case, the relative brightness will be (7.5/2)^2 * (2.9/7.52)^2 = (1.45)^2 in favor of the larger scope.

[Jape]

I don't know how to explain this any better than I already have. That was the simplest math that comes from the idea that when you are imaging the same chunk of the sky in a single pixel, then the amount of light that comes in will be proportional to (aperture diameter)^2.

 

You can go to the first principles and derive all of the relevant math, but here's the gist:

 

relative brightness of pixel ~ (fratio1/fratio2)^2 * (pixels size2/pixel size1)^2

 

In your case, the relative brightness will be (7.5/2)^2 * (2.9/7.52)^2 = (1.45)^2 in favor of the larger scope.

I'm not an expert, but I believe that the F/ratio is good as a practical tool for comparing speed in terms of the scene as a whole.

But in terms of light collected from a given FOV of the sky (or a given subject), aperture is the only thing that matters. You said it in the first line of the quoted post (highlighted).

 

Imagine a 100mm f/7 telescope shooting M101. Then we add a 0.5x reducer, obtaining a 100mm f/3.5 with a much larger FOV.

The amount of light collected for M101 is exactly the same, the reduced telescope is not twice as fast.

 

 

If you replace F/ratio with aperture in your last equation, we have (125/80)² * (2.9/7.52)² = 0.36.

 

That means that the first setup (HAC125/imx585) would be 3 times slower.

Edited by Jape, 22 April 2025 - 09:43 AM.

[dx_ron]

I would guess that you also should factor in the fact that a ZWO533 camera body will block 39% of the area for the HAC 125

[acrh2]

Imagine a 100mm f/7 telescope shooting M101. Then we add a 0.5x reducer, obtaining a 100mm f/3.5 with a much larger FOV.

The amount of light collected for M101 is exactly the same, the reduced telescope is not twice as fast.

 

 

The reduced telescope is "faster," as it will collect 2^2 more photons per pixel, because it will image 4 times larger area of the sky in a single pixel.

[smiller]

I think I see several mistakes from most of the responses so far:

 

1) Most here are focusing only on the f-ratio, but the OP is changing not only the F-ratio but dramatically changing the size of the camera sensor: both impact how much of the sky your capturing in the same way, they’re roughly equivalent.  The OP was purposefully trying to bring the two systems back to roughly FOV equivalency by doing this.

 

2) This overemphasis on how much light each pixel gets is misguided and is a bit of a “CCD days anachronism”.  This was important back in CCD days where read noise was always there lingering but that’s not so true anymore.  For people that are still obsessing over pixel fill rates: Could you please explain why this actually impacts image quality rather than the total number of photons from the target you actually acquire, regardless of how it’s distributed across the pixel if it’s something other than read noise you’re worried about?  Is this about read noise?
 

3) And finally, there is little emphasis on the difference in aperture of the two scopes which is fundamental on how many photons per square unit of sky are captured.  We don’t care much how many photons per pixel are captured, we care about how many photons of the target we are capturing, regardless of how they are distributed amongst the pixels.

 

Jape is on the right track… there isn’t much difference.

 

Jape, here is my swing at the answer:

 

1) If you are in a situation where camera read and thermal noise are dominating or a strong factor, such as you’re taking short exposures in dark skies, or with a very narrow filter, then pixel fill rates are importantly and low f-ratio and large pixel cooled cameras in high gain mode with low read noise dominate.

 

But I’m assuming in your question that this is not the case that you are capable of taking reasonably long exposures such as 3 to 5 minutes if needed and both cameras are cooled.

 

2) assuming we got number one out-of-the-way then since they have a similar FOV, because you hobbled the HAC125 with a tiny camera, then they both capture about the same square area of sky.  So we’ll assume that they have an equivalent field of view for the remaining of this argument.

 

3) With the same field of view, then the primary factor is how many target photons you collect per second and in that case the larger aperture will win.  You will have to moderate the HAC 125 based on its central obstruction, but it almost certainly still has more photon gathering power, and it will win by that amount which is perhaps 2x.

 

4) the different pixel scales will primarily result in a higher resolution image for the smaller pixel scale system, especially if you don’t bin the system with the IMX571 because the pixel scales aren’t so small that you need to bin (I know you just binned to create an equivalent pixel scale, no problem).  So in reality you wouldn’t bin that camera and you would get a higher resolution image in most seeing situations

 

So, the HAC125 will have 2x the target SNR gain productivity while the refractor will have a more detailed picture due to the smaller pixel scale despite the HAC125 having a larger aperture.

Edited by smiller, 22 April 2025 - 10:47 AM.

[Jared]

I recently came across the new HAC125 from Skywatcher, which is praised for its very fast f/2 optics and extremely affordable price point.
As I explored its features, I started wondering: to what extent is its compatibility with smaller sensors a limiting factor?

I imagined two different setups:
1- An HAC125 (f/2) paired with an IMX585 sensor (1/1.2”, 12.8 mm diagonal, 3840×2160 resolution, 2.9 μm pixel size)
2- An Askar 80PHQ (f/7.5) paired with an IMX571 sensor (APS-C, 28.3 mm diagonal, 6252×4176 resolution, 3.76 μm pixels), with 2×2 binning

Both setups have a very similar FOV (considering FL adjusted for crop factor) and can be used to image the same targets:
- FL around 830 vs 900 mm FF equivalent
- Resolution: 3840×2160 vs. 3126×2088
- Pixel scale: 2.4″/pixel vs. 2.6″/pixel

But there are also major differences:
- f/2 vs f/7.5
- Aperture 125 vs 80 mm
- Pixel size 2.9 vs 7.52 micron


My question for the experts is: which setup is truly “faster” in terms of achieving a better SNR in the same time?
I believe the second setup is MUCH faster because of the much larger pixels (almost 3x even ignoring central obstruction, transmission, QE) but I might be wrong: why would there be so much hype for the HAC125 speed, price aside?

The problem is you have a bunch of different variables in your comparison, not just one. Ultimately, though, assuming similar fields of view and assuming similar arc seconds per pixel (either due to actual pixel size or binning), the scope with the larger aperture will go deeper in a given integration time. It also has the potential to yield more resolution due to less diffraction. Discount the advantage of the 125 a bit due to central obstruction and lower throughput if you like, but the aperture difference is still the one remaining factor since your scenario equalized arc seconds per pixel and field of view.

SNR on the 125 should be about 1.6x higher for a given integration time even after accounting for central obstruction.

By the way, central obstruction is overstated in the SkyWatcher spec’s. It is 42% by diameter not by area, so about 18% by area.

[Jape]

I think I see several mistakes from most of the responses so far:

 

1) Most here are focusing only on the f-ratio, but the OP is changing not only the F-ratio but dramatically changing the size of the camera sensor: both impact how much of the sky your capturing in the same way, they’re roughly equivalent.  The OP was purposefully trying to bring the two systems back to roughly FOV equivalency by doing this.

 

2) This overemphasis on how much light each pixel gets is misguided and is a bit of a “CCD days anachronism”.  This was important back in CCD days where read noise was always there lingering but that’s not so true anymore.  For people that are still obsessing over pixel fill rates: Could you please explain why this actually impacts image quality rather than the total number of photons from the target you actually acquire, regardless of how it’s distributed across the pixel if it’s something other than read noise you’re worried about?  Is this about read noise?
 

3) And finally, there is little emphasis on the difference in aperture of the two scopes which is fundamental on how many photons per square unit of sky are captured.  We don’t care much how many photons per pixel are captured, we care about how many photons of the target we are capturing, regardless of how they are distributed amongst the pixels.

 

Jape is on the right track… there isn’t much difference.

 

Jape, here is my swing at the answer:

 

1) If you are in a situation where camera read and thermal noise are dominating or a strong factor, such as you’re taking short exposures in dark skies, or with a very narrow filter, then pixel fill rates are importantly and low f-ratio and large pixel cooled cameras in high gain mode with low read noise dominate.

 

But I’m assuming in your question that this is not the case that you are capable of taking reasonably long exposures such as 3 to 5 minutes if needed and both cameras are cooled.

 

2) assuming we got number one out-of-the-way then since they have a similar FOV, because you hobbled the HAC125 with a tiny camera, then they both capture about the same square area of sky.  So we’ll assume that they have an equivalent field of view for the remaining of this argument.

 

3) With the same field of view, then the primary factor is how many target photons you collect per second and in that case the larger aperture will win.  You will have to moderate the HAC 125 based on its central obstruction, but it almost certainly still has more photon gathering power, and it will win by that amount which is perhaps 2x.

 

4) the different pixel scales will primarily result in a higher resolution image for the smaller pixel scale system, especially if you don’t bin the system with the IMX571 because the pixel scales aren’t so small that you need to bin (I know you just binned to create an equivalent pixel scale, no problem).  So in reality you wouldn’t bin that camera and you would get a higher resolution image in most seeing situations

 

So, the HAC125 will have 2x the target SNR gain productivity while the refractor will have a more detailed picture due to the smaller pixel scale despite the HAC125 having a larger aperture.

I was ignoring thermal noise, transmission, central obstruction and small differences in FOV and pixel scale for semplicity.

 

I agree with everything you say, except you don't take into account the size of the pixels: a much bigger pixel in one setup (collecting light from the same "small piece of sky" in both setups) should be capable of much more light gathering - is that correct?

 

If that is correct and with everything else equal (FOV, resolution, pixel scale), then the advantage for HAC125 due to aperture (125/80² = 2.44x light gathering) should be compensated and overturned by such difference in pixel size (7.52/2.9² = 6.72x).

Edited by Jape, 22 April 2025 - 11:27 AM.

[acrh2]

I think I see several mistakes from most of the responses so far:

 

 

Excuse me, sir, but everything I said is mathematically correct.

[acrh2]

I was ignoring thermal noise, transmission, central obstruction and small differences in FOV and pixel scale for semplicity.

 

I agree with everything you say, except you don't take into account the size of the pixels: a much bigger pixel in one setup (collecting light from the same "small piece of sky" in both setups) should be capable of much more light gathering - is that correct?

 

If that is correct and with everything else equal (FOV, resolution, pixel scale), then the advantage for HAC125 due to aperture (125/80² = 2.44x light gathering) should be compensated and overturned by such difference in pixel size (7.52/2.9² = 6.72x).

 

Here are the first principles that I mentioned earlier:

 

1) If were are talking about SNR of an image, then we have to talk about SNR per pixel. Here's an example, binning 2x2. When you are binning, your SNR will increase by a factor of 2, and the resolution will decrease also by a factor of 2. This is because when you are binning 2x2, the brightness of a single pixel will increase by a factor of 2^2 (due the pixel area being 2^2 larger.) And the SNR of a single pixel is proportional to the square root of the signal, so SNR would be sqrt(2^2) = 2. This is for a hypothetical ideal camera without read noise. Whenever you are binning, you are increasing the SNR at the expense of resolution. Conversely, when you are drizzling, you are increasing resolution at the expense of SNR.

 

2) Pixel brightness is proportional to (aperture diameter)^2. The aperture is what provides light gathering ability for a single pixel.

 

3) Pixel brightness is inversely proportional to (focal length)^2. The focal length is what is spreading light out for a single pixel. 

 

4) Pixel brightness is proportional to (pixel size)^2. Pixel size is responsible for collecting light at the camera for a single pixel.

 

So if you combine 2,3 and 4, you get:

 

Pixel brightness ~ (aperture diameter)^2 * (pixel size)^2 / (focal length)^2 = (pixel size / focal ratio)^2.

 

Again, this is for a hypothetical camera without read noise, and also ignoring the central obstruction.

[Jape]

The problem is you have a bunch of different variables in your comparison, not just one. Ultimately, though, assuming similar fields of view and assuming similar arc seconds per pixel (either due to actual pixel size or binning), the scope with the larger aperture will go deeper in a given integration time. It also has the potential to yield more resolution due to less diffraction. Discount the advantage of the 125 a bit due to central obstruction and lower throughput if you like, but the aperture difference is still the one remaining factor since your scenario equalized arc seconds per pixel and field of view.

SNR on the 125 should be about 1.6x higher for a given integration time even after accounting for central obstruction.

By the way, central obstruction is overstated in the SkyWatcher spec’s. It is 42% by diameter not by area, so about 18% by areae 

(I am ignoring central obstruction and transmission for simplicity)

If we equalize pixel scale and FOV and resolution, shouldn't pixel size be a significant variable?
Wouldn't a much bigger pixel collect much more light, everything else being equal?
 

[acrh2]

(I am ignoring central obstruction and transmission for simplicity)

If we equalize pixel scale and FOV and resolution, shouldn't pixel size be a significant variable?
Wouldn't a much bigger pixel collect much more light, everything else being equal?
 

If you equalized pixel scale, you have already taken into account the pixel size:

 

image scale = 206.3 * pixel size / focal length.

Edited by acrh2, 22 April 2025 - 11:59 AM.

[Jape]

If you equalized pixel scale, you have already taken into account the pixel size:

 

image scale = 206.3 * pixel size / focal length.

I appreciate your replies, but I still believe there is something missing.

 

Let's replace the 80/600 scope with a 125/600, everything else being equal.

Now we have:

- same aperture

- same FOV

- same pixel scale

- same resolution

- a much bigger pixel for the 125/600 compared to the 125/250

 

... and yet according to your calculations the 125/250 would collect more light (f/2 vs f/4.8).

 

I can't explain in detail why, but that seems completely wrong.

[acrh2]

I appreciate your replies, but I still believe there is something missing.

 

Let's replace the 80/600 scope with a 125/600, everything else being equal.

Now we have:

- same aperture

- same FOV

- same pixel scale

- same resolution

- a much bigger pixel for the 125/600 compared to the 125/250

 

... and yet according to your calculations the 125/250 would collect more light (f/2 vs f/4.8).

 

I can't explain in detail why, but that seems completely wrong.

 

If pixel size for the two imaging systems is the same, then the pixel brightness ratio will be (focal ratio 1 / focal ratio 2)^2.

[Jared]

(I am ignoring central obstruction and transmission for simplicity)

If we equalize pixel scale and FOV and resolution, shouldn't pixel size be a significant variable?
Wouldn't a much bigger pixel collect much more light, everything else being equal?
 

If we equalize pixel scale, FOV, and resolution, then pixel size is completely irrelevant. The slower scope sends a dimmer image per area, but the larger pixels exactly counterbalance that, so the end result is the same amount of light per pixel per unit time for a given aperture. The only difference left, then, is the aperture. 

[Jared]

I appreciate your replies, but I still believe there is something missing.

 

Let's replace the 80/600 scope with a 125/600, everything else being equal.

Now we have:

- same aperture

- same FOV

- same pixel scale

- same resolution

- a much bigger pixel for the 125/600 compared to the 125/250

 

... and yet according to your calculations the 125/250 would collect more light (f/2 vs f/4.8).

 

I can't explain in detail why, but that seems completely wrong.

Pixel size, by itself, does not mean anything at all--literally nothing. OK, I'm simplifying by ignoring things like full well capacity and read noise, but that isn't what we are talking about here. A bigger pixel doesn't buy you anything at all in and of itself. 

 

A 125/600 scope vs a 125/250 scope will, to a first approximation, perform the exact same if you have a sensor in the 125/600 that is scaled up by a factor of 2.4. If you have the same number of pixels, just 2.4 x 2.4 times larger pixels you would get the same amount of light entering the telescope from your subject, the same amount of light hitting your sensor, the same amount of light detected by each individual pixel, and the same angular resolution for each pixel. The increased focal length (slower focal ratio) will dim the image at the sensor. The larger pixels will brighten the image. They offset.

[psienide]

Using roofkid's etendue calculator:

 

Telescope/Lens Reducer/Barlow Camera/Sensor Binning Plate Scale in "/px FoV (w) in ° FoV (h) in ° Etendue Etendue (QE respected)

Askar 80 PHQ                          1             ASI2600           2                       2.59             2.24            1.50          336                                   306
SW HAC125                             1              ASI585            1                      2.39             2.57            1.44          576                                   525

Edited by psienide, 22 April 2025 - 02:04 PM.

[CraigT82]

Can always have a go with John Hayes’ excel calculator… works well for comparing two scopes or more scopes. Put one scope you want to compare in System 1 column then the other scope in any other column. It’s pretty self explanatory.

 

Attached Files

•  SNR Ratio Calculator.xlsx   15.07KB   5 downloads

[smiller]

I was ignoring thermal noise, transmission, central obstruction and small differences in FOV and pixel scale for semplicity.

 

I agree with everything you say, except you don't take into account the size of the pixels: a much bigger pixel in one setup (collecting light from the same "small piece of sky" in both setups) should be capable of much more light gathering - is that correct?

 

If that is correct and with everything else equal (FOV, resolution, pixel scale), then the advantage for HAC125 due to aperture (125/80² = 2.44x light gathering) should be compensated and overturned by such difference in pixel size (7.52/2.9² = 6.72x).

Yes, if you are just concerned about the pixel... a giant pixel is like subsampling an image after capture: You tradeoff pixel level resolving power for pixel level SNR but this does not increase the SNR of the target:  That is dictated by the total number of target photons you collected which is the same in either case.

 

As stated in by another in another forum: "The reduced pixel level SNR associated with a higher resolution sensor is simply because the light/scene generated SNR is divided between more photosites"

 

You can do the same thing by standing further away from a displayed image:  You lower resolving power but the smaller number of pixels your eye receives and is able to resolve will have lower noise.

 

Excuse me, sir, but everything I said is mathematically correct.

I don't think I said anyone did wrong math.  It wasn't the math, it was the emphasis on the SNR of a pixel versus the SNR of your target using a normalized metric.  When we do AP, are we trying for the best pixel or the best image of the target?

 

If it's the best SNR pixels, then yes, you are correct, it's all about the pixel.  Big pixels is better SNR... of the pixel.

 

But I think the goal is the best SNR of the target.  In that case, if you capture more photons, even if it's spread across more pixels due to a much finer pixel scale, that can still render a better SNR of the entire target even if the individual pixels are higher SNR because they are smaller "bites of the image".   As long as you aren't read or thermal noise limited, it doesn't matter if you divide the image into 100 pixels, 10,000 pixels, 1,000,000 pixels or you just perfectly represent each captured photon as a "pixel", these all have the same image SNR even though you've subdivided it differently. 
 

If the quality of an AP target result is the end goal, then I think any SNR discussion across systems should either normalize to the same pixel scale or normalize to the same sized patch of sky (I.e, the target).

 

To that end, your first answer was the simplest and most direct one:

 

"Since the pixel scales are almost the same, the largest difference will be in the aperture diameter of the two imaging systems. The larger aperture will produce a brighter image by a factor of (125/80)^2, so it will also have a higher SNR."

 

And to expand upon this first answer, you can even say this if the pixel scales are different because the larger aperture (if it has a smaller pixel scale) will still have a higher target SNR even if the pixel scale SNR is lower.  To prove this, you can just subsample the image, after it is captured and stacked, down to the pixel scale of the smaller aperture scope and.... ta dah... you are back to the ratios of the apertures representing the SNR ratios, per your comment in blue.

 

Again, I was trying to switch the emphasis on the pixel level SNR to target level SNR.

 

Just look at these two equal sensor size full frame camera examples:  ASI6200MC with 3.76um pixels and ASI2400MC with 5.94um pixels.   I don't see experienced astrophotographers talking about the ASI2400MC producing higher SNR results of a target.   They know that they both capture the same number of photons of the target, it's just one splits them amongst more photosites.   People shouldn't worry too much about the SNR of each photosite. 

 

This is why people generally don't recommend binning, even if you are somewhat oversampled, as long as you aren't drowning in data or stacking time.    You can just downsample an image later if you desire.  This capture pixel SNR thing isn't a factor in the consideration.

 

Again any SNR discussion across systems should either normalize to the same pixel scale or normalize to the same sized patch of sky (I.e, the target).  Pixel level SNR discussions become meaningful if you are looking at the impact of pixel level SNR drivers like read noise and thermal noise.

Edited by smiller, 22 April 2025 - 04:56 PM.

[imtl]

I was ignoring thermal noise, transmission, central obstruction and small differences in FOV and pixel scale for semplicity.

 

I agree with everything you say, except you don't take into account the size of the pixels: a much bigger pixel in one setup (collecting light from the same "small piece of sky" in both setups) should be capable of much more light gathering - is that correct?

 

If that is correct and with everything else equal (FOV, resolution, pixel scale), then the advantage for HAC125 due to aperture (125/80² = 2.44x light gathering) should be compensated and overturned by such difference in pixel size (7.52/2.9² = 6.72x).

You keep going back to the point of larger pixels. What matters is the sampling. And the sampling in both cases is almost identical. That is when the pixel size differences end their contribution to the story. From here it is aperture as stated by Jared and ARCH2. Pixel size is irrelevant once you established the sampling.

[Jape]

A 125/600 scope vs a 125/250 scope will, to a first approximation, perform the exact same if you have a sensor in the 125/600 that is scaled up by a factor of 2.4. If you have the same number of pixels, just 2.4 x 2.4 times larger pixels you would get the same amount of light entering the telescope from your subject, the same amount of light hitting your sensor, the same amount of light detected by each individual pixel, and the same angular resolution for each pixel. The increased focal length (slower focal ratio) will dim the image at the sensor. The larger pixels will brighten the image. They offset.

Thank you for your answer, it helps me understanding - it is still somewhat counterintuitive though.

 

If I understand all the answers the SNR depends on (ignoring transmission, obstruction, QE, thermal noise...):

- (aperture)² * (px size)² * (1/FL)² 

 

that can be written as

- (1/FR)² * (px size)²

- (aperture)² * (px scale)²