Anyone tried square aperture mask?

[vlaiv]

I was sitting and pondering telescope resolution due to light diffraction, as one tends to do on a quiet Saturday afternoon

I'm still struggling to get the grasp of it in terms of QED. The way I like to think about it, which might be completely wrong, but it is used in some computer simulations, is via Heisenberg uncertainty principle. Once photon passes close enough to an edge it gets spatially constrained in one direction so position is well defined (in that direction), this means that momentum must be "smeared" out. And since momentum vector intensity is related to frequency and well defined - it there fore must change direction component of momentum vector - hence spread out (perpendicular to the edge).

At first, idea was to remove aperture and have only mirror in order to minimize airy disk - but I don't think that it will work, probably same principle as above applies to the mirror edge. Then I started thinking about the structure of airy disk - since we have circular aperture and photons spread perpendicular to edge - one easily sees how diffraction forms airy disk - spot like pattern - with peak intensity at the center (most photons hit aperture away from the edge and just land in center) and falling of (only small percentage of photons is really close to the edge enough to have their position so constrained they end up wildly changing momentum vector).

Then I had an idea. What would happen if we could shape distribution of momenta vectors in such a way to get better resolution in some cases. I know that apodizing mask is used to do that - but still in circular fashion.

Simple square aperture mask, at least if one follows above reasoning, would have really interesting, and possibly very beneficial effect on double star work. If we consider theoretically perfect optics - such square aperture mask would make all light that ends in airy disk and diffraction rings actually go into spikes - star image would change shape from airy disk into single high intensity point with spikes - this could vastly improve double star detection by just insuring that spikes don't fall on an image of the second star (simple enough - done by rotating the mask - it will rotate spike positions).

Of course, if this works, then "diffraction limited" term used to characterize optics quality no longer applies so resolution would be dominated not by aperture but rather smoothness and precision of optics figure.

So If anyone with experience in double star work, with decent optics could try this out and post findings, I would really appreciate it.

I, myself, certainly plan to do so, but I don't really have much experience in double star work (more like none at all ), and weather is real blocker at the moment. I might get a chance over next few week, but forecast is not looking great.

One would normally be restricted by seeing, and this would show a benefit only in the best conditions (ones that provide a good airy disk in normal scope).

One more thing I should mention, this simple mask design is suited for refractors. For scopes with any sort of central obstruction, a bit different mask should be constructed - square mask with square central obstruction (aligned sides), also if there is spider support - best configuration is to orient mask so that square sides are parallel to spider stalks.

[cotterless45]

I have used a very minimal hexagonal mask. Read years ago ,that in theory it works with refractors. What does work with Newtonians is either a straight through secondary vane ( older Orion Optics)  ,or a curved vane.

The standard four vanes greatly interfere with the light path, making binary observation difficult.  A lot of my observing is on fainter companions and I'm not keen on restricting too much aperture, Nick.

[vlaiv]

2 hours ago, cotterless45 said:

I have used a very minimal hexagonal mask. Read years ago ,that in theory it works with refractors. What does work with Newtonians is either a straight through secondary vane ( older Orion Optics)  ,or a curved vane.

The standard four vanes greatly interfere with the light path, making binary observation difficult.  A lot of my observing is on fainter companions and I'm not keen on restricting too much aperture, Nick.

That is great to hear, so my reasoning is not completely off. Hexagonal mask should have same effect as triangular one - six spikes 60 degrees apart, but provide more light gathering surface.

Can you remember if it had any impact on airy disk size and did it work at all?