Pixel size effects with equivalent arcsec / pixel

[zicklurky]

Something that has been running through my brain today with regards to difference in pixel size with systems with the same arcsec / pixel.

Let's have 2 hypothetical systems, with sensors with the same number of pixels, eg. both sensors are 2000x2000 pixels:

System 1 - 5 micron pixels with a 500mm scope =  ~2 arcsec / pixel
System 2 - 10 micron pixels with a 1000mm scope = ~2 arcsec / pixel

Apart from the deeper wells of the bigger pixels, what other differences are present? Will the same amount of photons hit each pixel?

[Stub Mandrel]

> Will the same amount of photons hit each pixel?

Assuming the same aperture, yes. The longer scope has twice the f-ratio so a quarter of the light hitting each point, but the 10x10 pixels are four times as big.

 

In principle it's just like how a crop sensor DSLR needs a different lens to take the same picture as a full frame DSLR.

 

In principle on a pixel for pixel basis both setups are optically identical, but in practice the different setups would introduce non-optical effects that might make differences.

[jsmoraes]

What I think is:

Lacks information about the aperture, but ... if they have the same aperture ...

2000 x 2000 - 5 microns - 500 mm scope will have images with more brightness/contrast (shaprness) and thus will show better the details of an image.

for example: a crater of the Moon will seems to have more details of their walls.

[theastrodragon]

The systems will push the same number of photons/sec onto a CCD pixel.

The shorter FR one will be a bit more difficult to collimate and focus, the longer will be heavier and more liable to flex.

There are some minor issues with what theoretically goes on that makes a very small difference, but in the UK with typical seeing of 3" or so they are of intellectual interest only.

[zicklurky]

By same aperture, are we talking same f/ number or diameter of the objective?

[jsmoraes]

I thought aperture as diameter of objective (or primary mirror).

[zicklurky]

28 minutes ago, jsmoraes said:

I thought aperture as diameter of objective (or primary mirror).

Ahhh ok thanks. That makes sense.

[ollypenrice]

What would intrigue me would be to compare two sets of results captured at the same pixel scale, one from a smaller aperture/smaller pixel system and one from a larger aperture/larger pixel system of the same F ratio.

For example we could compare a 14 inch F6.8/Atik 11000 with a 10inch F6.8/Atik 314L. Both would give 0.78"P/P.  Ignoring the difference in chip size (by picking a small galaxy as target) how different would these results be? There will be a small difference in chip sensitivity but what really interests me is how much is gained by using the larger optics.  The Dawes limit obviously favours the big one. Theory says the big one resolves to 0.33 arcsecs and the smaller one to 0.46 arcsecs. Both lie below the pixel scale of the cameras... 

So, where does that leave us??

Olly

[Stub Mandrel]

It's not the same thing at all, but I have the impression that, with a planetary target like Jupiter, if you look at the 'better' images much of the difference between images seems to come from aperture, not the camera. With my 6" scope small storms in the temperate belts are detectable, but have no detail. With 10" and especially 14" scopes, the detail appears much better and I don't think the difference is focal length as my scope + barlow is  ~ 3m. But comparing with other 6"  aperture scopes I don't see a great difference between webcams and expensive cameras.

 

[vlaiv]

For two setups with different pixel sizes but same resolution I think following applies:

• Light per pixel depends on light throughput of optics (aperture and light loss in scope) so if throughput is the same - the same amount of light hits the pixel
• Smaller pixels due to physics may have disadvantage in QE and well depth (with pixel sizes of couple of um we are fast approaching photon size which is the order of 500nm (half of a um), and of course smaller pixel size - smaller physical volume for electron well.)
• If same resolution is achieved with different aperture sizes, then given the same seeing, larger optics will have sharper images. We often think that seeing overrides resolution limit of scope, but in reality it is a sort of multiplicative effect (convolution of seeing psf and system psf). So star image will be blurred with combined psf which is larger in case of smaller aperture (convolution is associative operation so we can think of it as being star blurred by seeing and then by system psf, or star being blurred with seeing psf and system's psf combined - or even we can switch seeing and system psf, as convolution is commutative as well).

Not long ago I started a thread with a proposal to define new way of specifying "light power" of the system, it included light throughput, pixel area (not linear size), and sensor QE but it did not get much traction