Sorry for the confusion, but how is QE a factor here? Simply because I'm capturing more photons/second, so getting better SNR even though I'm using a slower scope?
And I agree, those newtonian astrographs look interesting. But does their longer physical length make them harder to mount/track than an SCT?
If we compare an 11" SCT with a KAF-16200 to a 10" newt with an IMX183. These two systems have very similar image scales and similar apertures, and are real systems you could get and use today. Both can deliver high resolution results that are comparable to each other.
The SCT system here has an advantage due to the larger aperture. The IMX183 has a very slightly larger image scale, but much higher quantum efficiency. The difference in Q.E. helps to counteract the aperture benefit of the SCT, effectively putting the two systems on equal footing.
The Newt system should still have higher throughput, thanks to the low read noise of the IMX183, it should be able to use shorter exposures, which reduces the cost of sub loss.
When it comes to tracking...image scale is going to be the biggest factor. An image scale of 0.45-0.5"/px very large, and that alone is going to put critical demands on tracking. The longer moment arm of the newt is a factor, but you really would not want to be under-mounted with either scope. Both are similar in weight, about 35-40lb with the scope alone, and several more pounds for additional equipment, dovetails, cameras, etc. So no low end mount is going to do, here. You would want something like a CGE, HDX, or a proper high end mount like an AP Mach 1, Paramount MyT, AP 1100, etc.
Does this imply that these large aperture scopes should generally be paired with such large pixels? And why?
When we start talking about 20-24" and larger scopes, the focal lengths tend to be pretty long regardless. Even with a CDK at f/6, a 20" scope has a 3000mm focal length. Outside of maybe solar system objects, small pixel CMOS cameras, even more moderately sized pixels, are going to be too small. Further, the field of view at 3000mm or longer with a Panasonic M or IMX183 sensor, even APS-C sized sensors, is going to be very small. So not only do bigger pixels pair better with big scopes like that, but you also want a big sensor. And read some of what John writes about his imaging experiences with the 16803 at just 14"...he usually gets a significant amount of data, often has to throw away a lot of it, and still usually needs to stack a lot of hours to get good SNR.
Could you explain why?
Continuing from above, and to answer this next question. Again...the image scale is the key factor here...doesn't much matter how big the pixels are, or even whether they have 60% or 75% or 84% Q.E...what matters is that each pixel only sees about half an arcsecond of sky, and you just need longer exposures and lots and lots of data to get good SNR at that kind of an image scale. At 0.2"/px, you would need four times the exposure (total exposure) as at 0.4"/px, and even more than that as at 0.5"/px, to get the same SNR. So bigger pixels are more effective here, if you want to be able to achieve a reasonable SNR in a reasonable amount of time. This is what John Hayes is saying. (Differences in Q.E. will ultimately pale in comparison to differences in aperture...if you had the means to get a 24" RC, and pair it with say a KAF-9000 with its 12 micron pixels, such a system would again have an image scale around 0.5"/px, and would be 4x as fast as say a 12" newt system that was also at 0.5"/px. Even with very high Q.E. you might gain about a 50% improvement over the KAF-9000, which is just not enough to overcome the benefit of the aperture. You could even build an even bigger system, with a 36" aperture, an even bigger sensor with even bigger pixels, and it would be even faster. Thing is you keep losing FoV here, and not everything is about getting the highest SNR out of the smallest FoV as fast as possible... )
Edited by Jon Rista, 06 November 2018 - 07:40 PM.