Pentax Super-Takumar 28 mm f/3.5
Pentax made 28 mm f/3.5 lenses in M42 threaded mount from the early 1960s to the late1970s (aohc.it). Small-scale production of these lenses may have continued into the 1980s (takumarguide.weebly.com).
My interest in these legacy lenses is mainly in testing the most promising ones in digital near-UV imaging. I am publishing information about the external appearance of these lenses on this site, in order to allow the clear identification of which of the many models and variants I have tested in UV imaging. Performance in UV may be different in other variants and models, especially if they use different optics and/or different coatings.
Variants of Super-Takumar 28 mm f/3.5
These lenses were made in four recognized variants, which use at least two different optical schemes. A summary of the differences among variants and pictures of the optical schemes can be found on pentaxforums.com.
Based on the above criteria for lens candidates to test in UV imaging, variants 1 and 2 are attractive because of their relatively thin and single-coated elements, but the use of one cemented doublet is of some concern. Variant 3 is potentially interesting because of the single-coated elements and the absence of cemented groups. However, the very thick third element is a potential problem that may substantially reduce the transmission of NUV. In fact, in this lens roughly three-quarters of the physical length between front and rear optical surfaces is glass, instead of air. Variant 4 is probably best avoided because it is multi-coated.
This page deals only with tests of variant 1 and variant 3.
All variants of this lens use metal lens barrels with no plastic parts. These lenses feel unusually heavy compared to modern ones, and some of the parts are made of brass.
Focus throw is only about 3 mm. Both lenses produce sharp images when stopped down to f/8 or f/11, both in VIS and UV. They are reasonably sharp also at f/5.6.
Aperture operation can be switched between automatic (i.e. fully open by default, and closed to the preset aperture when the camera body presses a pin on the lens mount) and manual (always closed to the preset aperture) with a slider located near the lens mount.
The use of this slider is quite counterintuitive: when you move the slider toward the M or Man(ual) label, the slider covers this label and the lens is actually in Auto position. When you move the slider toward the A or Auto position, the slider covers this label and the lens is in manual aperture mode.
The adapters I use for mounting these lenses on mirrorless cameras have a flange that keeps this pin pressed at all times, so the aperture operation is manual regardless of the position of the manual/auto slider. When an extension ring without aperture pin is used between lens and adapter, however, this slider must be placed in the manual position.
My suggestion for avoiding problems with the aperture slider is to always leave it in manual position (i.e. covering the A/Auto label).
Physical differences among variants
Variants 1-2 are distinguished by a large diameter of the lens barrel (especially its front, which is 60 mm wide) . Variants 3-4 have a narrower front of the barrel (51 mm wide at the front). As a whole, variants 1-2 are significantly wider, longer and heavier than variants 3-4. The latter were probably designed to reduce the size and weight of the original design, as well as to allow the use of smaller filters and lens shades.
Adapting M42 lenses to mirrorless cameras
Since these lenses are designed to cover a 36 by 24 mm film frame, they are suitable for use on full-frame mirrorless cameras. Aperture operation is fully manual on mirrorless cameras. I did not test these lenses on smaller sensors.
Dielectric-coated filters, which include the large majority of filters used in UV imaging, do not perform well with wideangle lenses and often produce images with a different general tint in the center of the image. Mounting the filter at the rear of the lens avoids this problem, and the adapter for mounting M42 lenses on mirrorless cameras provides an ideal place for housing filters up to about 37 mm in diameter. In general, I use 30.5 mm filters (or 1.25" astronomy filters re-mounted in 30.5 mm filter frames) in a 37 to 30.5 mm step-down ring.
Since the rear element in the lenses discussed in this page (and the vast majority of legacy SLR wideangle lenses that use a retrofocus design) is much smaller than a 30.5 mm filter, the latter causes no vignetting.
Another obvious advantage of rear-mounted filters is that they allow the use of wideangle lenses with different sizes of filter mounts and front elements, including sized so large that specialist UV-pass filters of mathing sizes are not available. A further advantage is that many legacy wideangles do vignette unacceptably when using a lens shade stacked at the front of a filter, while rear-mounted filters avoid this problem.
This variant has 58 mm filter mount and 5 aperture blades. It stops down to f/22 in half-stop clicks and focuses down to 40 cm. The clear diameter of the front element is slightly less than 45 mm.
The lens I tested has scalloped focus ring and finely ribbed aperture ring.
This variant has 49 mm filter mount and 5 aperture blades. It stops down to f/16 in half-stop clicks (with only a full-stop click between f/11 and f/16) and focuses down to 40 cm. The clear diameter of the front element is slightly less than 39 mm.
The lens I tested has scalloped focus ring and aperture ring.
The following tests were carried out as an evaluation of the suitability of these lenses for NUV imaging. For this purpose I used a full-spectrum Sony A7 II and an M42 to Sony E adapter modified to carry a 1.25" Baader U UV-pass filter (re-mounted in a standard 30.5 mm filter ring) a short distance behind the rear element of the lens. The camera was set to use a custom white balance as described here.
Images are cropped but not otherwise post-processed.
I modified the M42 to Sony E adapters shown above by attaching with silicone adhesive a 43 mm to 37 mm filter step-down ring at the rear of the M42 ring (which is part of the original adapter). The M42 ring is held in the adapter body by three Allen bolts.
The adapter on the left has been disassembled to show the individual parts. The 30.5 mm filter is mounted on the modified adapter via a 37 mm to 30.5 mm step-down adapter. The two step-down rings keep the filter well clear of the rear of the lens.
The adapter on the right shows a 1.25" Baader U filter in place (re-mounted in a 30.5 mm filter ring).
In spite of using very different optics, the UV transmission and the UV false color of the two lenses is virtually identical. The lenses produce UV-blue and UV-violet, but no UV-yellow. Compared to the Enna Lithagon 28 mm f/3.5, both Super-Takumars absorb one more stop of UV and display a more restricted range of UV false color.
Both Super-Takumars produce sharp UV images and convincingly reproduce the UV signatures typical of many flowers. Considering the fact that these lenses are quite cheap on the second-hand market, either one can be used for UV imaging, and especially UV landscapes, as long as monochromatic or almost monochromatic images are acceptable. The Enna Lithagon 28 mm f/3.5 produces much better UV false color, but is roughly 3-4 times more expensive.
Among legacy wideangles usable for UV imaging, the Enna Lithagon 24 mm f/4 produces roughly the same slight false-color as the Super-Takumars and a significantly wider angle of view. However, the Lithagon 24 mm absorbs one stop of UV radiation more than the Super-Takumars, which may cause problems with NIR leaking through the UV-pass filter. The Lithagon 24 mm is roughly 4-5 time more expensive than these Super-Takumars.
The variant 1 and variant 3 of the Super-Takumar 28 mm f/3.5 are cheap and produce sharp UV images, but transmit only a relatively narrow band of UV between roughly 380 and 400 nm, with one stop of UV transmission less than the Enna Lithagon 28 mm f/3.5, but one stop more than the Enna Lithagon 24 mm f/4.
This site is ad-free. If you see any ads here, they are added by your ISP, or by spyware on your computer, or you are visiting this site through frames of another site.