Assessment of Antique & Vintage Lenses in UV (& IR)

[bvf]

Lenses used in this test:

• Aldis 120/7.7 Uno Anastigmat
  
• Bausch & Lomb 180/8.0 Rapid Rectilinear
  
• Canon 200/2.8L EF
  
• EL-Nikkor 80/5.6 (metal version, enlarger)
  
• EL-Nikkor 105/5.6 (metal version, enlarger)
  
• Ensign 75/4.5 Ensar Anastigmat
  
• Leitz Wetzlar 50/4.5 Focotar-2 (enlarger)
  
• Meyer Optic Görlitz 50/2.9 Trioplan
  
• Prinz Galaxy 135/3.5
  
• Steinheil München 50/2.8 Cassar S
  
• Voigtländer 105/6.3 Voigtar Anastigmat
  
• Zeiss Ikon 75/4.5 Nettar
  
• Zeiss Ikon 50/4.5 Novar
  

---

Virtually all of the UV-friendly lenses we use date back to the 1950s-1970s and typically have M42, M39, or Exakta mounts. But how about lenses that are older than that, from the days before AR coatings and interchangeable mounts – how good might they be for UV?

 

As a lockdown project I acquired some of these older lenses to try them out in UV (and also in IR). The criteria for the choice of lenses was:

• Made pre-Second World War, and preferably much earlier. Although plenty of pre-First World War lenses are available they are generally of long focal lengths (up to 180mm) because of the large format films in use. You have to go to the 1930s to find 75mm lenses, and wait to the late 1930s (when 35mm and 16-on-127 cameras become more common) for 50mm lenses.
  
• Multi-element lenses
  
• On or from folding cameras, to make lens removal and adaptation to M42 easy.
  
• Maximum price of £20 (inc. any camera still attached to them) – so no Leitz or even Tessar lenses.
  
• A working T function on the integrated shutters
  
• A working aperture
  

These were the lenses tested:

 

 

A couple of observations:

• The Bausch and Lomb Rapid Rectilinear had apertures marked from 4 to 128. It turns out these markings used the U.S. scale (U.S. = Universal Standard rather than a reference to former British colony). U.S. 4 is equivalent to f/8.
  
• The Meyer Trioplan 50mm f/2.9 illustrates why I sometimes find it hard to understand the modern world. It is a mid-range 1930s triplet and yet the early post-WW2 copies in M42 or Exakta mounts demand ridiculous prices of £150-300 (a new Canon 50mm f/1.8 costs just £130). The only thing the Trioplan has going for it is its “distinctive soap bubble bokeh” when used fully open. So if you want a lens that gives you so-so mages with an annoying, distracting background then this is the lens for you. More amazing still is that Meyer Optik Gorlitz is back in business making their range of 1930s lenses – a previous crowd-funded incarnation offering the Trioplan at $2,000 failed a couple of years ago, and the current owners are selling Trioplans and the other 1930s models at a mere $1,000. I think some people just have too much money
  

UV Transmission

 

The transmission of the lenses through my three UV bandpass filters (380BP20, 345BP25, 315BP20) were compared with the metal El Nikkor 80mm. The El Nik was chosen as the benchmark because it is perfectly usable with the 315BP20, whereas benchmarking against the best transmitters (Focotar-2, Cassar S) would have set the bar unfairly high.

 

It turns out that all of these lenses perform at least adequately at 345nm, and so would work well with general-purpose filters like the Baader U.

 

The Voigtar and Trioplan lenses showed good UV reach down to at least 320nm (the peak transmission of the 315BP20), and the Ensar could be used at a push. These are all air-spaced triplets.

 

Unsurprisingly, the lenses with cemented elements (Rapid Rectilinear, Aldis Uno) performed poorly – too poorly to be used at 320nm.

 

But surprisingly the two worst transmitters at 320nm were air-spaced triplets – the Zeiss Ikon Nettar and Novar. Obviously at least one of the elements is made of a UV-absorbing glass. Fluorescence of the lens seems to be the best predictor of UV transmission – it looks like the rear elements of these lenses is the culprit.

 

 

Resolution

 

The table below compares resolution provided by these lenses in UV, Vis., and IR, and compares them with some other lenses to provide context.

 

In visible light (at f/8 or f/11), the Rapid Rectilinear and Ensar performed very badly. The other lenses hold their own. (I’m sure the poor number for the Trioplan in the centre at f/8 is some kind of an error on my part.)

 

In UV (at f/8 or f/11), the Rapid Rectilinear again performs very poorly. The Aldis Uno is somewhat weak. The Trioplan and Ensar are OK at the centre, but drop off badly towards the corners. The Novar, Nettar, and Voigtar perform reasonable well.

 

In IR (at f/8 or f/11), the Rapid Rectilinear completes its lamentable performance. The Trioplan and Ensar are again let down by poor performance at the edges. The other lenses perform adequately.

 

The characteristics of the Trioplan are similar to those of the Cassar S.

 

 

Other Aberrations

 

The lenses are clear of pincushion and barrel distortion. Astigmatism is also almost completely absent (as it should be as they all claim to be anastigmats). But there is a significant amount of chromatic aberration – especially in the Rapid Rectilinear, Ensar, and Nettar.

 

[dabateman]

Well if the Meyer trioplan wasn't expensive before, expect it to go higher now that it looks like a good UV lens.

 

Adding to your question of why?

I have seen lenses with square and triangle apertures go for stupid amounts of money. You can do the same with a lens baby and cut out cardboard or early sets with various cut outs like hearts. Oh well.

 

If I can make a one of a kind ugly lens, I will make sure to sell it for stupid high prices. Then I could use the money to actually help people.

[Andrea B.]

Bernard, another excellent and useful set of lens data. Thank you!

---

 

If I can make a one of a kind ugly lens, I will make sure to sell it for stupid high prices. Then I could use the money to actually help people.

 

David, here's something I noticed recently while prowling around Ebay. I was looking at the usual 3rd party lenses -- similar to that Soligor 35/3.5 above. If the "knock-off" lens was available for a Leica M mount, then there appeared to be about a $500 - $1000 price added to it as compared to the same thing available for mere Nikon or Canon mounts. Too funny! (I should have downloaded some examples to show everyone. Oh well.) So whatever lens you make up, be sure to make it fit a Leica M mount. Then you can charge the usual arm & leg for it and provide a bigger donation to the charity of your choice. :devil:

 

Old German metal lenses are quite popular these days amongst the collector and enthusiast crowd who don't use them for UV. So we are seeing some unexpected prices.

[Avalon]

Would be interesting to see UV photos made with Meyer trioplan and Voigtar 105 mm 1:6,3 as well their transmission spectrum curves. Voigtländer 105/6.3 Voigtar Anastigmat has no helicoid, can it be easily converted to M42?