Weired Bokeh with Bigma and choice of filter

[ArvoJ]

3. The filter is tilted with respect to the front lens. Then you would get a striped diffraction pattern - again creating the striped obstruction.

Note - for the diffraction theory to hold - the front lens of have to be nearly teh same curvature as the filter, i.e. the front lens needs to be flat.

Tilted (or wedged?) filter seems reasonable cause. Actually it doesn't need to be parallel to any lens element - it has two [badly coated] [almost] parallel sides itself. I attempted to trace rays (on paper) - seems to create double images (in out-of-focus areas) just due to the internal reflections, amplified by interference between direct and double reflected rays.

--
Arvo
Sigma/Foveon information collection and little gallery:
http://www.stv.ee/~donq/sigma.htm
http://www.stv.ee/~donq/images.htm

 

[Roland Karlsson]

Tilted (or wedged?) filter seems reasonable cause. Actually it doesn't need to be parallel to any lens element - it has two [badly coated] [almost] parallel sides itself. I attempted to trace rays (on paper) - seems to create double images (in out-of-focus areas) just due to the internal reflections, amplified by interference between direct and double reflected rays.

You are riight - a wedged filter is the most likely cause.

One interesting trivia is though that I have read several times both in this and in the Pentax forum about Bigma double images for out of focus parts of the image.

I have even been in a heated discussion in the Pentax forum - where a poster accused me of accusing him of forging an image. He had taken a fantastic image with two hard to catch birds on the same branch. There was a very strong double image in the background though. And - I told him so with a smilie

He then interpreted my post plus the smilie as a hint that maybe the birds were a part of a double exposure. He even wrote me an angry email - and he never believed me when I said I had absolutely no such intention.

Cest la vie.

--
Roland

support http://www.openraw.org/
(Sleeping - so the need to support it is even higher)

X3F tools : http://www.proxel.se/x3f.html

 

[Truman Prevatt]

It's interference - it is not from diffraction. When a wave that - is "split" is incident on a sensor - one with a phase offset from the other they will produce an interference pattern. If they are 180 deg out of phase - the energy will cancel (destructive interference). The distances involved for are very small for example for green the 1/2 wave length (for destructive interference) is about 250 nanometers. When they are in phase is a multiple of 360 degrees (which is a multiple of 500 nanometers for green) the energy will add. This can happen for multiple reasons. Shining a light on a board with two slits is one. A slightly variable index of refractions in a lens or filter coating is another. The enlargement of a long focal length lens will show the interference quite well which is why you see it more in long focal length than short focal lengths. It is fairly easy to calculate the variations of the index of refraction and focal length that could produce an interference pattern but I'm to lazy to do it.

Slight difference in curvature in the wave front from a lens vs. a filter is another. There is a physical cause - associated with the lens and/or lens filter combination. However, what you are seeing is an interference pattern on the sensor.

Truman

But ... I am not sure it is interference. It might be.

But ... if it was interference due to the the different curvature of the filter and the front lens - then the pattern should be circular - not striped.

I can see three possible reasons for the striped bokeh.

1. The filter has som striped obstruction. You can e.g. get that kind of result if you hold your fingers in front of the front lens. Try it and see for yourself. It works just fine. But - that is of course absurd. The owner would see the obstruction.

2. The surface of the filter is wavy with some broad - but very slight - striped waves. Those waves could - due to interference or some optical refraction effect - create the obstructed stripes.

3. The filter is tilted with respect to the front lens. Then you would get a striped diffraction pattern - again creating the striped obstruction.

Note - for the diffraction theory to hold - the front lens of have to be nearly teh same curvature as the filter, i.e. the front lens needs to be flat. An alternative to this is maybe hat another surface in the lens is optically flat seen from the outside. But the deeper into the lens - the less likely.

--
Roland

support http://www.openraw.org/
(Sleeping - so the need to support it is even higher)

X3F tools : http://www.proxel.se/x3f.html

--
Truman
http://www.pbase.com/tprevatt

 

[Roland Karlsson]

It's interference - it is not from diffraction. ...

Interference and diffraction is basically the same thing, or rather diffriaction is caused by interference. If you have an aperture you can see that aperture as an infinite number of point sources where the waves from each point source interfres; then you get diffraction.

But this is besides the point.

I am 100% sure that the striped bokeh is caused by something striped that blocks the path. Thats how you get striped bokeh even if the in focus image is not striped. And for long tele lenses this thing that blocks the path can even be in front of the lens.

Now - there is of course nothing physically visible that blocks the path. That would be easily seen.

It more or less has to be interference - and then something that causes a striped interference. Like Arvo said - its probably a wedge formed filter.

--
Roland

support http://www.openraw.org/
(Sleeping - so the need to support it is even higher)

X3F tools : http://www.proxel.se/x3f.html

 

[Truman Prevatt]

Diffraction results in an interference pattern. However, interference is not diffraction. http://en.wikipedia.org/wiki/Diffraction . The famous Michaelson-Morley experiment used the properties of the interference fringe to show that there was no luminiferous aether by bring two beams together using a half silvered mirror. There was not diffraction anywhere in the house - but there was an interference fringe.

There are many ways - other than diffraction - to produce an interference pattern. http://en.wikipedia.org/wiki/Interference_%28wave_propagation%29

--
Truman
http://www.pbase.com/tprevatt

 

[Roland Karlsson]

Diffraction results in an interference pattern. However, interference is not diffraction.

There are many ways - other than diffraction - to produce an interference pattern. http://en.wikipedia.org/wiki/Interference_%28wave_propagation%29

I am sorry Truman - but this is very much my domain

Your descriptions are not wrong but very confusing

Interference and diffraction are both just effects of wave propagation. Neither of them are fundamental in any way. Actually - they are more or less the same thing. But most importantly - the basis is wave propagation - nothing else.

If you assume wave propagation and then integrate over all possible propagations - then the result will be effects like diffraction and interference.

Diffraction can be seen as interference between an infinite number of wave sources in an aperture - if you so like. It can also - much simpler - be seen as the actual waves from an infinite number of wave sources.

--
Roland

support http://www.openraw.org/
(Sleeping - so the need to support it is even higher)

X3F tools : http://www.proxel.se/x3f.html

 

[Robert Krawitz]

Shot yesterday:

Shot today - in different lighting conditions:

Everything in this shot is sharper. The other difference is that it was shot at F9.0 and +0.30EV.

Apologies for jumping in 5 years(!) after the fact, but I had exactly the same problems with my 50-500 f/4.5-6.3 -- double images and bokeh with strong diagonals (at a different angle, though). I've never seen that with any other lens. I finally figured this out last weekend when trying to test out the lens; I'm just a bit unhappy with myself that I spent $250 to get Sigma to try to fix it (and that I had suffered almost a year with awful image quality).

I've never seen this effect (obvious image degradation with a filter) with any other lens. That includes the Sigma 70-200 f/2.8 OS, even with a Canon 2x TC. The filter I had on was cheap (Vivitar Series 1), but it shouldn't be causing that kind of effect. I tried another filter, which degraded the image quality noticeably but without the strong double image effect.

One thing I've noticed is that filters do not screw on smoothly. I can turn them about 3/4 of a turn before they tighten, but it doesn't feel like they "stop" in place. I wonder if at least some of these lenses have problems with misalignment of the filter thread such that the filter plane is not perfectly perpendicular to the lens axis.