I here describe my experience with the DEO-TECH OWL EF/MFT lens adapter. The adapter is available mainly through astronomy web stores and seems to be currently distributed by Optolong. It is also available on special order at B&H photo video and possibly other large shops. Development of the adapter was initially funded through Kickstart crowd funding.
The adapter with one drawer is fairly expensive at 135 €. Filter drawers for 48 mm and 52 mm filters are available separately. I bought a Canon EF (D)SLR lens mount to micro four thirds (mirrorless) camera body mount adapter as it is the most versatile and a few additional filter drawers. Of the common full-frame DSLR mounts the EF mount has the shortest mount-to-film/sensor distance allowing rather easy adaption of lenses for Nikon Ai/G, Olympus OM, M42, and other (D)SLR mounts using additional adapters.
First limitation of this adapter is that it’s diameter near the MFT mount is too large preventing its use in Olympus OM-D camera bodies. However, the metal is thick enough to allow the use of a lathe to correct this problem. I found out about the problem and solution from a user comment at B&H web store, before buying the adapter. I had the adapter modified at a local workshop.
Using filters between the objective and camera body has several advantages, especially in the case of wide-angle objectives. The first advantage is that in the case of some fish-eye objectives with very wide angle view it can be impossible or very difficult to attach filters to the front because of the bulbous front lens element. The light enters the front of such lenses over a very broad range of different angles with respect to the flat surface of a filter, introducing various types of vignetting depending on the type of filter. The angle of incidence of light on a rear-mounted filter tends to be much closer to 90 degrees to the filter surface independently of the focal length. Additionally, a rear-mounted filter can, especially in the case of “crop-sensor” cameras, be of a smaller diameter. In the case of MFT, 30.5mm is more than large enough to avoid vignetting. As one could expect, smaller filters can be much cheaper than larger ones.
Some wide-angle lenses have at there back a holder for very thin gelatin filters, while some telephoto objectives, especially catadioptric objectives may have threads at their camera end for mounting glass filters. This is likely because adding a filter between an objective and the sensor or film shifts the focal plane. This shift is larger for short focal length (wide-angle) objectives than for long focal length objectives. This brings us back to the OWL adapter and use of filters. Most objectives intended for photography can focus past infinity and also at close distances. For objectives of long focal length the shift of the focal plane due to the use of filters can be taken care by refocusing. For objectives with short focal length this is usually not possible as the shift is so large.
In the case of adapting the Sigma 4.5mm 1:2.8 circular fisheye objective in Nikkon AF mount adding a glass filter brought the farthest possible focus to approximately 4 or 5 m. This explains why this objective with which it is impossible to use front-mounted filters, only accepts very thin gelatine filters at its back. As I am using a Nikon G to Canon EF adapter to adapt this lens I was able to easily add about 0.45 mm of shim to the adapter to move the objective away from the sensor, which restores focus at infinity with glass filters of a thickness close to 1 mm as well as retaining close focusing and a reasonable match to the focusing distance scale on the lens. Without a filter behind this lens focusing is not possible with the shimmed adapter. One could always use a filter, as a clear protector filter will also shift the focus plane, or use an adapter that is not shimmed when a filter is not used behind the lens.
I used Dymo text- embossing tape as shim and a very cheap Nikon G to Canon EF adapter partly made of stamped sheet metal. I removed the ring on the objective side by removing six small screws, added two layers of suitably cut pieces of tape, sticking them in place to the flange ring and attached back the ring and tightened the screws in stages alternating between screws on opposite sides of the ring. The tape is 0.225 mm-thick so the separation added in this case was 0.45 mm.
Roger Cicala of LensRentals wrote two posts in 2014 about the effect of glass filters present on the sensor of digital cameras on the performance of objectives. The first post deals with the technical details while the second post discusses the implications. He has some days ago published a new post about lens-mount to sensor distances in photographic cameras where he also discusses the question of the thickness of sensor-mounted filters. Based on this, one should expect that thick rear-mounted filters will deteriorate lens performance more than thinner filters. One can also expect that the 4 mm-thick sensor-filter-stack of MFT (Olympus and Panasonic Lumix) and Sony mirrorless cameras deteriorates the performance of objectives designed for film. The problems are more likely to affect the edge of the image circle, and consequently the smaller MFT sensor might have a bit of an advantage. A thin-filter conversion service is available for Sony cameras which improves the performance of adapted legacy objectives designed for film cameras.
The end of the story is that I can now use the Sigma 4.5 mm circular fisheye lens for infra-red (IR) photography and obtain images of reasonably good definition and in focus. This experience, also highlights why when digital cameras are converted to full spectrum or infrared, by removing the UVIR-cut filter on the sensor, it is important that it is replaced with a suitable filter or clear quartz or glass of the same effective thickness (which depends both on the actual thickness and the refractive index of the glass) as the original if modern objectives are to be used, while a thinner filter or no replacement at all may work better with objectives from the film era. Different cameras, and especially cameras with different mounts, tend to use “sensor stacks” of thicknesses varying between less than 1 mm and slightly more than 4 mm. Objectives are designed to match, and obviously objectives designed for film cameras assumed no glass between lens and film. Some of the consequences of adding glass cannot be compensated for, but the focus-plane shift can by a simple modification of a lens adapter.
Swapping filters is relatively easy, but the drawer lacks a “grove or rail system” to guide insertion. Once inserted, the filter drawer stays securely in place. I haven’t noticed any light leaks. As is the case with most adapters, except for the very expensive ones from Novoflex, the black finish both inside and outside is more reflective in the IR than in the visible (it is not IR-black which could potentially contribute to hot spots and reduced contrast in IR images). While the drawer is not in place, dust can enter and reach the camera sensor, so swapping filters in a dusty environment has almost the same risk of dirt getting to the sensor as swapping objectives under the same conditions.
In conclusion the adapter works as advertised, provides clear advantages and also it has some disadvantages related to the use of rear mounted filters. Using a small diameter filter can be expected to decrease reflections and vignetting, can introduce a correctable focus plane shift and possibly some additional effects on adapted objective + camera performance.