Lens with two apertures

[baachitraka]

One of the inspiration to ask about two apertures is that I use to work with corrugated horn antennas with different profiles.

http://www.google.com/search?um=1&h...pl=9602l11547l0l15l10l0l1l1l0l234l769l2.1.2l5

The mathematics behind is quite complicated. I just wonder whether it is possible to shape the light waves through two more apertures...

Probably not. Apertures are usually placed at the nodal point of the lens. There is only one of those. An aperture placed in a different spot either won't have an effect or will become the de facto aperture, but in the wrong place. There are however many "tricks" available in optics, but in the end it's all math and physics, and there is rarely a free lunch with those two.

 

[polyglot]

The Sony lens is a soft focus lens, not a super sharp lens.

Sorry, but that's just not true. The STF is not a soft focus lens, it does not have deliberate spherical aberration, in fact it has remarkably low SA and is one of the very sharpest 135mm lenses you will find. It introduces no softness (reduction in resolution) to any part of the image, either in-focus or out-. It was originally a Minolta AF lens and Sony continues to make it, so it (including the current Sony version) will work with any Alpha/Dynax/Maxxum 35mm cameras people might have.

What it does is have an apodisation filter (dark at the edges, light in the middle) very close to the aperture. This means that when shot wide open, the bokeh has a gaussian rather than circular form, which means that you no longer have blur circles with sharp edges but big soft blobs with no visible edge whatsoever, which means that backgrounds become beautifully smooth.

With any lens (ignoring aberrations), a point source appears on the film as an image of the lens' aperture, scaled by how far out of focus it is. So a focused point is a point, a defocused point will appear as a circle/septagon/whatever with sharp edges. If you then put the apodisation filter next to the aperture, that circle is dimmed at the edges and you lose the sharp edges on your bokeh.

As to the original poster - your vision gets sharper because when you squint, you're manually stopping down your aperture. The eye is a very simple lens that works poorly wide-open. Our vision is sharp in bright sunlight (small pupils, high f/number) but in dim light, the pupils open up, expose more of the poorer-performing part of the lens and your vision gets softer.

 

[markbarendt]

Dang, wrong again.

 

[baachitraka]

Thanks for clarification, further STF has two apertures.

Sorry, but that's just not true. The STF is not a soft focus lens, it does not have deliberate spherical aberration, in fact it has remarkably low SA and is one of the very sharpest 135mm lenses you will find. It introduces no softness (reduction in resolution) to any part of the image, either in-focus or out-. It was originally a Minolta AF lens and Sony continues to make it, so it (including the current Sony version) will work with any Alpha/Dynax/Maxxum 35mm cameras people might have.

What it does is have an apodisation filter (dark at the edges, light in the middle) very close to the aperture. This means that when shot wide open, the bokeh has a gaussian rather than circular form, which means that you no longer have blur circles with sharp edges but big soft blobs with no visible edge whatsoever, which means that backgrounds become beautifully smooth.

With any lens (ignoring aberrations), a point source appears on the film as an image of the lens' aperture, scaled by how far out of focus it is. So a focused point is a point, a defocused point will appear as a circle/septagon/whatever with sharp edges. If you then put the apodisation filter next to the aperture, that circle is dimmed at the edges and you lose the sharp edges on your bokeh.

As to the original poster - your vision gets sharper because when you squint, you're manually stopping down your aperture. The eye is a very simple lens that works poorly wide-open. Our vision is sharp in bright sunlight (small pupils, high f/number) but in dim light, the pupils open up, expose more of the poorer-performing part of the lens and your vision gets softer.

 

[Steve Smith]

If you want sharp focus and minimal depth of field at all aperture settings then you need to fit a variable aperture in the same shape as a cat's eye. i.e (). No matter how closed up it is, the image is formed from the full diameter of the lens.


Steve.

 

[baachitraka]

Something analogous to elliptic horn(Antennas theory).

 

[polyglot]

Thanks for clarification, further STF has two apertures.

Physically, yes; functionally, no. They are right next to each other and only one aperture actually closes when the image is taken. As far as the optical design is concerned, there is a single aperture.

If you want sharp focus and minimal depth of field at all aperture settings then you need to fit a variable aperture in the same shape as a cat's eye. i.e (). No matter how closed up it is, the image is formed from the full diameter of the lens.

That's going to look really weird but probably cool. You'll have elliptical blur aligned with one particular axis of the image, which means that you'll have anisotropic defocusing - detail at a given position in the image will be subject to a different level of blur depending on its orientation. You could take a photo of a chainlink fence and make half the wires disappear and the other half remain! It can be a cool effect though: google for "heart bokeh" sometime, the use of a cardboard mask on the front of a lens with some point-lights (e.g. coloured christmas lights) in the background can produce some very cutesy effects.

Something analogous to elliptic horn(Antennas theory).

Well now you're just making shit up. Elliptical horns are used because they have different gain in each of the two planes, i.e. to form an elliptical gain pattern, i.e. a fan-beam instead of a pencil-beam. Antennae are not imaging devices, they do not have planes of focus and therefore they do not have defocused areas. The effect is also in the opposite direction: a wider horn causes a narrower beam whereas a wider aperture produces less DOF in an imaging sensor; the only thing in common is that they both obey the Rayleigh limit (diffraction) for their resolution/beamwidth limit. You can think of a horn as a lens with a single pixel (corresponding to its feed point) that's been stopped down well into diffraction - if you stop it down any further (make the horn narrower), the beam representing that one pixel will get even wider/fuzzier.

 

[baachitraka]

Too much imagination seems back-firing for me. @polyglot: Again, thanks for clarifying the difference.

May be I delete this thread if possible, since it does not give anything useful for the community.

 

[polyglot]

No need to delete anything and you should feel free to disregard me sounding grumpy

You are right that an elliptical aperture on a lens is like an elliptical aperture on a horn and they both have differing resolution in each dimension. It's just that the mechanism and physics is so very different and in fact causes variation in the opposite direction as each dimension varies.

 

[baachitraka]

I was only involved in programming different horn profiles, with almost no knowledge on its design. It was just a wild imagination that any such profiles is applicable in optical lens design.

That was the background to ask such a question... ;-)