Why the Bananascope kicks ass

[Rocket Pooch]

Hi,

Just doing some calculations this morning on the Banascope and also some of the kit I have to give an idea what makes it soooo good.

Here are some facts,

Typically I image subs for 20 minutes a sub the same image (assuming you use the same camera) would be 3.86 minutes.

With the F4 scope I have the 20 minute exposure becomes 7.89 minutes.

This means the Bananascope is 5.18 times faster than my ED80 with the reducer and 2.04 times faster than the F4 newt.

Now, I'm definately switching to fast newts for imaging even if it kills me.

As a side point, Mark can help here, the DSLR is less sensitive than my QSI I have not looked at calculating it, but it mittigates some of the slowness in the scope, so I think the F4 newt and QSI should give me a good overal imaging solution moving forward.

Chris

[MarkS]

Chris,

Good analysis.  It was a train of thought very similar to this that led me to buy the Bananascope in the first place.

One point you may have overlooked is that the central obstruction is a massive 44% in terms of diameter so it reduces light collecting ability to about 80% but this is still 4x faster than the ED80+Reducer (instead of 5x).   By the way, it also means that CCD dark currrent is no longer an issue for me.

As for QSI vs Canon, the Canon CCD is certainly less sensitive - if you ripped off the RGB Bayer matrix, its Quantum Efficiency would be around 2/3 that of the Kodak KAF-8300.   But the 8300 has a smaller pixel area (5.4nm vs 6.4nm) which  reduces the light collecting ability of each pixel by 30%.  So measured on a pixel for pixel basis, there is very little to separate the 8300 and the Canon. 

But the Achilles heel of the Canon is the Bayer matrix.  Using an identical imaging train,  the 8300 will outperfrom the Canon for colour imaging because LRGB acquisition works more efficiently than Bayer matrix acquisition - however the difference between the 2 techniques is not that great (though I've yet to calculate it) and the Bayer matrix does have the advantage of simplicity.  For narrowband imaging, the 8300 is miles ahead - absolutely no question - because only 25% of the pixels are working with the Bayer matrix.

I definitely think your plan of using a fast Newtonian is a great idea once you have solved any collimation/focuser issues and assuming the corrector gives nice stars over the field of view.  Collimation at F4 is critical and at F2.8 it's almost impossible - that's why the Tak is built like a tank - otherwise it would be impossible to keep collimated.

All in all, I reckon the QSI on an F4 scope will give very similar full-colour images to the Canon/Tak.  For narrowband images it will definitely outperform the Canon/Tak.

Mark

[Mike]

Can the matrix be removed? Would it be a decent CCD using colour filters as normal do you think or is it not worth the effort?

[Ian]

The matrix is an intrinsic part of the ccd.

However, it might be fun to try and source the equivalent mono sensor and replace it. If one exists, of course...