Want a step up from qhy5liic, what colour camera would you recommend?

[vlaiv]

Just now, yelsac said:

Oh dear I seem to have opened a can of worms here......

No, it's not your fault. It's mine.

I keep insisting on some things that I should probably let go ...

In any case, I think my advice is very sound. I'll explain details briefly (all verifiable things).

- if you increase "magnification" while capturing (by using barlows), for same exposure length, signal strength goes down. This is well known fact and can be even tested visually - increase magnification - image will become dimmer. This is because light spreads over larger surface so there is less photons per unit surface.

- Lower signal = less SNR. This is fairly obvious since SNR is signal/noise and if we decrease signal we decrease ratio of it and noise

- Telescope can resolve only so much and that depends on aperture size.

That all adds up to very sound conclusion that you should not make image at capture time larger than it needs to be (than telescope can resolve) - because your subs will either have lower SNR - which makes it harder for software to determine sharpness and stack properly and impacts overall final SNR result which prevents you to sharpen image as much as you would like, or it forces you to stack more subs than you should (lower quality subs) or it forces you to use longer exposures to recover signal - which is also not good in most circumstances because we want to freeze the seeing. In most cases this happens around 5ms. If atmosphere is really stable - this time can be longer. Some planetary imagers enjoy very good and stable skies and this allows them to use longer exposures without having seeing induced motion blur. This in turn lets them capture more zoomed in image without as much ill effects mentioned above, however that does not mean they are going to capture additional detail as one can't defy the laws of physics.

What you do after capture really does not matter, so if you decide to drizzle or scale image up - that is fine. Just be aware that you can't add detail that way (that would be like enlarging image and expecting it to remain as sharp - that does not happen).

[neil phillips]

4 minutes ago, yelsac said:

 

Oh dear I seem to have opened a can of worms here......

Well I know we don't always agree which I suppose keeps life interesting but I really appreciate both of your comments & imput. You both have so much experience that is so valuable to ones like me, so many thanks for your advice.

Clear skies

We disagree from time to time. Nothing to do with you. Vlaiv does a lot of good on the forum.  And has my respect for all the good he does. But opinions can vary sometimes and that's all.

You may find something interesting here

WinJUPOS tutorials (t-online.de)

 

[vlaiv]

2 minutes ago, Varavall said:

Hi

That's all very interesting. Could you post images you have made with each of those cameras so we can see what you are getting at?

Which part confuses you?

Here is QE graph of ASI462:

From it - it is obvious that peak QE is around 820nm and that such camera is useful for IR imaging. It also shows that QE is lower in visible part of spectrum.

Here is published graph of ASI662:

It is absolute QE graph and peak QE is over 90% which is also quoted by ZWO.

Similarly here is graph for ASI678:

Blue is strongest with peak at 83%

Rest of specs are also available, and I personally don't doubt them (much) as I have tested several ASI cameras and specs (mostly) match those published (read noise is reported a bit optimistically low in my opinion - but not by much).

[Varavall]

18 hours ago, vlaiv said:

Which part confuses you?

Nothing. Are you coming at this as a theoretical scientist and not as an experimental scientist like myself? Theory is fine if it was an ideal World, but I like to see the experimental data, i.e. images taken with those cameras using the same optical train in the same location on the same night. 

[vlaiv]

51 minutes ago, Varavall said:

Nothing. Are you coming at this as a theoretical scientist and not as an experimental scientist like myself? Theory is fine if it was an ideal World, but I like to see the experimental data, i.e. images taken with those cameras using the same optical train in the same location on the same night. 

Don't confuse theory being right or wrong with your lack of understanding of it or its domain of application.

If you find that theory is "wrong", it can be one of few things:

- you are applying the theory in the wrong way (lack proper understanding of it)

- you are applying the theory outside of its domain of application (you are applying theory correctly but to the wrong case or maybe wrong initial conditions)

- you found genuine evidence that theory is wrong. That is major thing and cause for "celebration" - you have made genuine advancement in science, but that does not happen very often. When this happens you should definitively publish a paper about it.
 

All above is that I've written is easily verifiable / testable

When stacking, only noise contribution that depends on number of stacked subs rather than total imaging time is read noise. If one wants to use shorter subs - one will benefit from lower read noise, as total SNR will be better with lower read noise camera. If we had camera with zero read noise - it would not matter if we took 10000 x 10ms or 20000 x 5ms or 100000 x 1ms exposures as far as SNR goes (if we stack the same percentage of each frames).

QE is rather self explanatory in terms of benefit - more signal - better SNR. Again - something easily testable.

Using high FPS camera is just common sense really, nothing to test there. Higher FPS - more frames captured, with same percentage of good frames - more frames to stack, better SNR.