Front vs. back filter optics question

I was in a curious mood yesterday and did some experimentation. I have a 14mm Samyang f/2.8 ultrawide lens that I sometimes use on an Olympus Pen F with an adapter giving a 20mm equivalent focal length. This len's bulbous front element means it doesn't have a front filter thread. It also has a slightly bulbous back element and doesn't take internal/rear filters either. Anyway, I have some small series 6 filters that I normally use with my TLR. I wanted to see what would happen if I taped one on the back as if using a back filter. The filters fit pretty much perfectly in that space between the EF-mount adapter and camera.

What I found was that the infinity focus point changed quite a bit, somewhere around ~1.2 meters marked on the lens barrel. My question is why does that happen and why doesn't it happen when front filters are used?

And would you predict a degradation in sharpness if the lens was used this way? Everything looked fine, at least to my eye through the viewfinder.

Final third question. I have an infrared filter in this series 6 size. What are the odds this complicated ultrawide lens with a zillion elements is usable for infrared photography?

Thanks.

Here's the answer to your first question.

loccdor said:

What I found was that the infinity focus point changed quite a bit, somewhere around ~1.2 meters marked on the lens barrel. My question is why does that happen and why doesn't it happen when front filters are used?

Click to expand...

The simplest explanation is probably that the camera/lens system was designed to operate in air, only, inside of the camera. When you insert a filter inside of that space the focused image will be formed at a slightly greater distance. This is a result of the filter material having a higher refractive index than air. (Light behaves as though it were "slowed down" inside of a higher index material. )

The diagram from monopix demonstrates the concept.

FWIW the same thing also happens in front of the lens; it's just that the subject distance is so much greater that the small distance shift due to the filter is usually insignificant.

loccdor said:

And would you predict a degradation in sharpness if the lens was used this way? Everything looked fine, at least to my eye through the viewfinder.

Click to expand...

I would say that theoretically the filter will displace different colors to slightly different distances (the refractive index of the filter "glass" varies slightly with the color of the light, red vs green vs blue. Does it make a practical difference? I'm a little doubtful that one would notice anything in general purpose usage. But I really don't know for sure.

FWIW high quality gelatin filters are so thin that the image shift is essentially negligible. So they would be better suited, from this standpoint, for use behind the lens (inside of the camera body).

Gels get dirty and damaged easily, and need rather cumbersome adapters back there. Few lenses do well optically with rear mounted filters unless they are specially designed for that application. Tape has its own drawbacks, especially for fragile gelatin filters which can be expensive to begin with.

But lots of us have tested this kind of postulate out of curiosity, including me.

For ordinary glass filters, the focus shift is about 1/3 of the thickness of the filter. (This is because typical glass has index of refraction n ~ 1.5, air has n=1, and the optical path length is proportional to 1/n.)

A typical filter is perhaps ~1.5mm thick, so you have ~0.5mm of focus shift, which is very noticeable behind the lens (in image space). But 0.5mm is negligible in subject space unless you are working at macro distances. That is why you don't notice the focus shift with a filter on the front, and if you did notice you would just refocus and take it out.

For a thin flat piece of glass, if the cones of light passing through it are narrow angle and are not at a sharp angle relative to the glass, the aberrations it introduces are negligible. So filters of good optical quality have little effect on the image in the front, since the beam of light from a point on the subject to the filter is very narrow angle (very slow in effective f-ratio). The cone of light from the lens to the film is much faster (f-number equal to whatever aperture you are using), and the off-axis angle of beam to filter might be large, IOW the filter is tilted relative to the beam. This in principle introduces an off-axis aberration, I think it's astigmatism, but whether it is significant I'm not sure, and it likely depends on the lens.

If you have a lens that is designed to take rear glass filters, like certain telephotos and fisheyes, leaving the filter off will cause unacceptable focus shift and may also cause an off-axis aberration, since the lens was designed to work with the filter in place.