Mounting unusual lenses on digital cameras

lens recovered from Polaroid Sprintscan 35

Pleasing effects can often be found when using unusual lenses on modern digital cameras. Sometimes they give a highly aberrated image which is useful in creative situations, sometimes the history of the lens enhances the image in a meaningful way, and sometimes the unusual lens offers a quality which is not possible with typical lenses from first party manufactures.

The last instance is often the case with macrophotography. Many people use objectives to photograph small insects as they offer much higher magnification ratios and resolution than macro lenses. Occasionally, lenses used for obscure purposes in industry can find use in various areas of photography. Surplus aerial photography lenses, such as the Kodak Aero-Ektar are highly sought after for their high resolution and high speeds over large flat film planes. Occasional lenses with have excellent performance are found for macrophotography.

It is one of these which I have acquired recently from an old film scanner. Film scanners image a flat sheet of film onto a flat sensor at high resolution. Unlike lenses for general photography they are optimised for magnification ratios of around 1. Most photography lenses are optimised for a magnification of almost zero.

lens recovered from Polaroid Sprintscan 35

Mounting a lens can be easy. Adapters often exist allowing a camera lens to be mounted on other cameras. However, when we look at older lenses, or lenses which were designed for industrial use the adaptation is more difficult.

CAD model of the lens holder

In order to mount the lens securely, I designed a small device to securely hold the lens. The bore has 0.5mm of clearance for the lens barrel, and a flange at the base so that the lens can be reliably positioned the same distance from the sensor. This was 3D printed by a friend of mine. There is also a place for a grub screw to be installed so as to secure the lens in place. The base is so sized so that it may be bonded to a body cap. The injection moulding used on the body cap left the inside surface shiny, this was sanded to reduce reflections. The body cap can the be fitted to a set of bellows or extension tubes.

3D printed lens hold affixed to body cap and with a screw to hold the lens in place
Lens mounted to an old set of bellows. These bellows are backwards, so that the remaining rails are on the camera end.

I was surprised to realise that the top of the body cap was quite convex. This cased two problems. Firstly, the contact area which the lens holder made was rather small, and secondly, it would slide about as the glue dried. To compensate for the first issue I used quite a lot of adhesive. To compensate for the first I used a quick curing epoxy resin. This turned out not to be so quick curing, and I spent about 30 minutes poking the parts into alignment.

I intend to test the lens both ways around and at different magnifications. I don’t know exactly what magnification it will perform best at, presumably at it’s design magnification. However, it may surprise us. The lens is not a symmetric design, the front (dot end) has a convex element, the rear surface is plano.

I took the lens out for a short while and tried to photograph some insects. Unfortunately, I didn’t bring my flash light guide, so most of the picture turned out greatly under-exposed. The lens is not exceedingly sharp, at least, not at the magnifications which I tested it at. However, this is not a well designed resolution test (that will come later).

Full image (resized) and 100% zoom on a 36MP full frame sensor

As can be seen from the above frame, the lens is not sharp to the pixel. However, it shows nice contrast and has very little longitudinal chromatic aberration.

Not many insects would stay still for me. This guy did, but he was really small.

The insect above was very small. I’d be interested to know what species it is. Part of the issue with this photo is the due to a heavy exposure pull due to a lack of flash power. The Pentax K-1 isn’t known for its dynamic range, but this is pulled 3.5 stops, and I don’t think it is so bad. I tried a few different magnifications, but I didn’t keep track of it. The working distance is quite short, but this is also a lot higher magnification than my macro lens.

The resolution is probably what I should expect. The scanner that this lens came from was a 2,700 dpi scanner and the resolution of my sensor is 5,200 dpi, so it isn’t surprising that it out resolves the lens. However, image space resolution isn’t the only important property.