Schneider V38 mount lenses,
a general discussion

Schneider Kreuznach markets a number of lenses in a proprietary V38 mount. The male V38 mount has a 38 mm diameter flange with a V-shaped groove around the perimeter, while the matching female mount has a cylindrical aperture with three headless set screws around its rim. The set screws engage the groove of the male flange.

Note that there are also wider V-groove Schneider mounts, including V48, V70 and V90, used for larger machine-vision and line-scan lenses.

The V38 mount is designed for semi-permanent lens installations. The multiple set screws used to join V38 components require a small Allen key, and the assembly procedure requires the use of both hands and a good clearance all around the lens barrel. The set screws easily leave permanent indentations on the aluminium groove when tightened (some of my Schneider accessories came with a thin nylon padding on the grooved lip in correspondence to the screws, presumably to provide some protection). Positive characteristics of this system are that it allows the lens to be rotated in any orientation before locking the set screws, and that accidental loosening is unlikely. Presumably, the metal lens barrels are also more durable and better suited to applications in machine vision than the plastic barrels and fiddly aperture rings of enlarger lenses.

The V38 mount is primarily designed for machine vision optics, but Schneider markets also adapters to mount V38 lenses and accessories on M42, M39 and other mounts, like Nikon F. China-based eBay sellers offer comparable (but much cheaper) M42 to V38 adapters.

Schneider lenses in V38 mount have metal barrels, usually with a male V48 mount at either end for easily reversing the lens. The front V38 mount is normally covered by an aluminium ring engraved with the lens model and ending in a 40.5 or 43 mm female filter thread. This ring sometimes also functions as a rather shallow lens shade, and its length differs in each lens model.

In addition to a few high-end lenses specifically designed for machine vision and only marketed in V38 mount, Schneider also offers a few of its Apo Componon, Componon S and Componon enlarger lenses (also marketed in plastic barrels with M39 attachment) in V38 mount. The optics of these lenses are the same in both types of mount.

Although the V38 Apo Componons are not engraved with the HM (High Modulation) acronym present on the corresponding lenses in M39 barrels, there seems to be no difference in optics. This page discusses the three lenses shown above.

All lenses in V38 barrels are assembled with the optics in the same orientation as the corresponding enlarger lenses in M39 barrels, i.e., they must be reversed for use at magnifications above 1x. This implies that the aperture scale becomes oriented the opposite way than the lettering on the lens shade/filter carrier and any V38 system accessories mounted at the rear of the lens (i.e. the aperture scale is upside down when the lens points straight downward).

The diaphragm of the lenses discussed in this page has only 5 blades, and the aperture is distinctly pentagonal. This is the same diaphragm as in the corresponding enlarger lenses.

The above figure shows the lenses in their "normal" orientation, i.e. as supplied by Schneider. The magnification range specified for each V38 model refers to this orientation. Note that the aperture scale is marked in integers starting with 1. The 1 setting corresponds to fully open (e.g. f/4 on the 60 mm, but f/2.8 on the 40 mm and 28 mm), and each unit corresponds to one stop. There are half-stop markings, but no click stops. The main specifications are:

Presumably, Schneider specifies the diameter of the image circle at infinity. Note, in particular, the small size of the image circle of the Componon 28 mm. The Apo Componon 40 mm provides an image circle just sufficient to cover a full frame sensor.

A small thumbscrew can be inserted in one of three threaded holes placed around the aperture ring and used to lock the aperture ring at the chosen setting. Also this screws may mar the underlying metal barrel, unless tightened very lightly. Schneider supplies with each lens also a headless set screw that can be used for the same purpose. This screw differs from those used to mount the lens shade in not having a sharp tip, but it still can mar the metal of the lens barrel.

V38 lenses and accessories are remarkably expensive when purchased new (in my opinion, overpriced for what they provide). They are not rare on the second-hand market, and a handful are often available at any one time on eBay. Dedicated machine-vision lenses in wider V-groove mounts are instead rarely seen on eBay.

The ease with which these lenses can be reversed without requiring additional adapters, the sturdy metal barrels with lens shades and filter mounts, and their compact sizes, compared to the corresponding lenses in plastic barrels with M39 mount, are advantages in photomacrography.

Mounting the lenses

Extension rings are available as accessories of the V38 system (see the above figure). Adapters to other mounts are also available. Some of the accessories shown above may no longer be produced, like the ?? mm V38 to M42 adapter. In addition, the system also includes a few focusing helicoids.

The Makro Unifoc 12 helicoid (above figure) extends by a total of 12 mm and can be locked with a thumbscrew or headless screw. This helicoid has a sufficient friction to resist sagging and gliding (albeit not under continuous vibration), and can also be used, without locking, for frequent changes of extension. It is very well made, with no detectable play and wobble, and unlike most third-party helicoids it has a scale displaying the amount of extension for 0 to 12 mm. Some of the lenses, if directly attached to this helicoid, prevent the helicoid from fully retracting. In these cases, it is best to place an extension ring between helicoid and lens.

The above figure also shows a different, lockable helicoid with a threaded front barrel. The whole lens barrel revolves when this helicoid is adjusted. This helicoid should be locked with its locking screw after setting the desired extension, because nothing otherwise prevents the lens from falling off at the end of the helicoid thread.

In a pinch, it is possible to modify the female end of an M39 extension tube to accept a V38 lens by drilling and tapping three M2.5 holes for set screws around its female mount. It is also possible to mount a V38 lens by using the filter holder/lens shade that accompanies each lens, together with an M43 to M42 reversing ring.

A word of caution about these accessories: the internal surfaces of extension tubes, helicoids and adapers are smooth and painted with a matte black paint. This is obviously insufficient to prevent internal reflections (above figure, left). All three lenses displayed the same problem.

Flocking the interior with Protostar or other flocking solved this problem (above figure, right). Care must be taken to avoid flocking the inner surfaces that mate with lenses or other accessories.

It is remarkable that Schneider did not recognize this problem, or did not regard it as important. Given the high prices charged for these simple accessories, I would have expected better quality control from Schneider.

As a whole, it is a good idea to use extension tubes significantly wider than the Schneider V38. I get good results with tubes with 57 mm threads. Even though their finish may cause some internal reflections, even without added flocking the large internal diameter helps to shield the sensor from off-axis illumination

Using the lenses in practice

For magnifications above 1x, all three lenses must be reversed. Image quality is unacceptably poor if the lenses are used at 1x or higher magnification in a forward orientation.

For my purposes, the M42 adapters are the most useful among those available in the V38 system.

Earlier tests

I am not aware of other test comparing all three of these lenses. CoinImaging.com tested the Componon 28 mm f/4, which may or may not have the same optics as the Componon 28 mm f/2.8 mounted in a barrel with aperture scale that starts at f/4. Several years ago, I made a qualitative test of the Apo Componon 60 mm f/4. There is also a thread on photomacrography.net discussing the same lens. More recently, Robert O'Toole included the Componon 28 mm f/2.8 and the corresponding f/4 lens, with several others, in a major series of tests at a magnification of 4x.

While the above tests and discussions refer to these lenses being used reversed on extensions, I remember reading on photomacrography.net a short report of the Apo Componon 40 mm f/2.8 being used in an infinity system. I can no longer find this.

The Schneider Makro-Symmar 120 mm f/5.9 is another lens, available also in V38 mount (the versions in M39 mount are rated at f/5.6), that has been extensively tested in photomacrography, both as a macro lens and as a tube lens. This page on my site contain a discussion of this lens, tests as a tube lens, and links to other sites describing experiences with the same lens.

Resolution and magnification in theory

The Apo Componon HM lenses discussed in this page use a relatively simple optical scheme with 6 elements in 4 groups, with a cemented doublet at either end. This scheme uses thicker elements of (presumably) modern glass types, and should be moderately more expensive to produce than modifications of the double-Gauss scheme used in the majority of non-apochromatic enlarger lenses.

The Schneider literature on these lenses also shows that they are apochromatic only with respect to transversal chromatic aberration. They behave like ordinary achromatic lenses with respect to axial chromatic aberration (i.e. they correct this aberration only at two wavelengths, not three like apochromatic lenses are expected to do).

Axial chromatic aberration was not of particular importance for enlarger lenses, since they were only supposed to be used in perfect focus on a flat subject, and to produce a flat projected image. This aberration is much more counterproductive in photomacrography, where it becomes immediately visible in out-of-focus portions of even slightly three-dimensional subjects.

To me, this is one more indication that these lenses were designed for photographic enlargement, and subsequently re-purposed by Schneider for machine vision and small-subject imaging without any re-design of the optics.

The reciprocal of the optimal magnification can be used as a starting point for the optimal magnification with the lenses reversed (second column in the following table). However, this does not take diffraction into account.

Diffraction is taken into accunt in the calculation of maximum useful magnification at a given lens aperture. For this test, I used an Olympus E-M1 Mark II with 20 Mpixel sensor, with pixels size of 3.33 μm. The maximum useful magnification at a specific lens aperture is specified in the remaining table columns. This is only a best-case estimate, since there is no guarantee that testing will show an acceptable image quality at the specified magnification and aperture.

It is evident that the small pixel size of a 20 Mpixel Micro 4/3 sensor, used in the calculation of the CoC (Circle of Confusion), forces the lens to be used at actual magnifications significantly lower than its optimal magnification by design. This is equivalent to saying that these lenses are designed for use with sensors with physically larger pixels.

The purpose of this test is to show the limitations of these lenses. If a lower resolution is acceptable than the maximum theoretical resolution capabilities of the sensor, on the other hand, then it is acceptable to stop down the lens. As always, to decide which resolution is acceptable, one must consider the final pixel count of the image.

For example, if it is already known that the image will only be used on a web site that limits image width to 1024 pixels, there is no point in producing a 5184 x 3888 pixel image (20 MPixel) with a resolution of 2.5 pixels per line-pair. A much lower resolution (around 12 pixels per line-pair in a 20 Mpixel image, or 2.5 pixel per line-pair in a 0.7 Mpixel image) will be sufficient for this use. Such an image can be recorded by further stopping down the lens past the point where diffraction causes a visible image blurring when pixel-peeping the original image. Doing so may correct other aberrations to a higher degree. The pixel count can then be reduced in post-processing. Possibly, but less likely, an alternative in these cases can be to shoot at a lower magnification and crop the picture in post-processing.

Lens resolution test

In 2017, I carried out a few test of the Componon and Apo Componon lenses in V38 mount mentioned above, mounted reversed on extension tubes in the photomacrography range (defined as imaging with a non-compound optical system in a magnification range starting at 1x). These Schneider lenses, in practice, become difficult to use above roughly 5x. For some reason I did not publish the half-finished page discusing these results, until I "found" again the page in 2022.

I had no particular need to test these lenses in the macrophotography range (hereby arbitrarily defined as the magnification range between roughly 0.25x and 1x), because this range is already covered by a large number of (often excellent) macro lenses with built-in focusing helicoids. As such, I find that the use of lenses that require an added extension (extension tubes, helicoids, or bellows) is mainly justified for lenses of unusually good performance, e.g. in terms of very high image resolution, extended spectral range, extreme correction of aberrations, or other uncommon but desirable characteristics.

Given the unremarkable results of the test with these lenses on extension, and the fact that I read about the Apo Componon 40 mm bein successfully used reversed onto a tube lens in an infinity-corrected system, I tested all three lenses on 100 mm and 200 mm tube lenses. For this test, I used the Olympus 100 mm f/2.8 and Olympus OM 200 mm f/5 (see here for details of this setup). The approximate magnification with these lenses and these tube lenses is shown in the following table.

As discussed at the above link, the effective aperture of an objective on an infinity system is different (somewhat faster) than with the same objective on empty extension.

In an infinity system, the objective projects an image in focus at infinity, not at a finite distance. This also implies that the working distance is slightly lower when the objective is used on an infinity system.

Ideally, an objective for an infinity-corrected system should be optimized for infinity focus on the image side (or on the subject side for a lens to be reversed and used as an infinity-corrected objective). This is not the case for the lenses discussed in this page. Although they are designed for quite a low magnification (0.05x), this is nowhere near infinity. Therefore, without testing, it is uncertain whether these lenses will perform better on tube lenses versus empty extension. We already know from the tests discussed above that their performance on empty extension is far from exciting.

In the following table, in addition to magnification I am also specifying the horizontal FOV (Field of View), since the magnifications provided by this setup are not easily interpreted.

I do not show the test results with Componon 28 mm on 200 mm tube lens, because the magnification (7.1x) brings the 28 mm clearly outside its optimum range.

The working distance of the Componon 28 mm becomes so short in this setup that the lens shade constrains the subject illumination to a low angle of incidence, literally grazing the subject from the sides. A curious type of rainbow flare appears in these conditions, perhaps caused by the low-angle of illumination and parts of the wafer surface behaving like a diffraction reticle. The rainbow flare increases substantially by stopping down, which suggests it is accompanied by a low utilized lens aperture caused by insufficient diffusion of the illumination (which is another consequence of grazing illumination).

Removing the lens shade and using an illumination with a higher incident angle reduces this type of flare (or rather, blends it with enough diffused light to hide the rainbow effect). This type of flare still becomes visible now and then, on a small portion of the image, with all three of these lenses.

Conclusions

While Schneider Componons and Apo Componons do not rate among the best lenses when reversed on extension in photomacrography, they do tend to perform better when reversed atop a tube lens in an infinity-corrected system. This, however, is not necessarily true of all such lenses. In addition to the total magnification of such a system being sometimes too low to be useful, or too high to allow a sufficient immunity to diffraction, design characteristics of some of these lenses may result in optical aberrations, sensitivity to flare and excessively short working distance. Only tests can tell for sure whether a given combination of lens and tube lens is usable in practice.