Can you 3D print parts for a lens rehousing? It turns out that yes, yes you can

Adding the need for new parts to the cost of vintage glass can make everything prohibitively expensive, but 3D printing offers the ability to make some decent savings on the process.

As eBay gets down to the bottom of the barrel on old lenses, some of the more ambitious sellers have begun asking six weeks’ wages for a lump of glass and corrosion that’s been lost in a rainwater cistern since the 70s. 

Even once we’ve emptied out all the partly-composted leaf mulch, desperate measures may be required to make things usable. Sometimes that means some mount adaptors made of hammered bacofoil, because who needs the distance marks on the barrel to mean anything. Often it’ll mean a 15mm lightweight rod system, because there’s no comfort in watching a clip-on mattebox containing a £400 polariser dangle precariously from a mid-century expression of Soviet engineering principles. The list goes on.

This is what it takes to do it properly - an engineer's bench at True Lens Services in Leicester, UK

It’s not difficult to avoid these problems. You take a collection of fungus-enriched optical parts to True Lens Services or Zero Optik, hand over a major solid organ and your firstborn child, and walk away several seasons later with a piece of engineering of such staggering beauty and precision that it floods the room with golden light every time you open the case. Of course, the underlying lens might still be a lump of volcanic obsidian hammered into shape by orangutans with mallets, but at least it’ll have sufficiently predictable focus behaviour to be consistently fuzzy.

Sadly, those most at risk from back-alley lens dealers are also likely to have already sold at least one of their kidneys on the internet. The gold standard approach of a proper rehousing is often be well out of reach. It doesn’t make much economic sense, anyway, for most of the proletariat glassware that’s still available at student-friendly prices. It’s always been expensive because of the enormous amount of mechanical-pencil work that’s intrinsic to rehousing lenses.

Even the most basic optical designs, with one assembly of which moves relative to the sensor, demands a lot of precision. More advanced designs might have different elements moving in different directions simultaneously; others shift weights around to counterbalance other moving parts, so as not to unbalance a gimbal or Steadicam. Even once all of the required clockwork has been designed to suit a particular type of historic lens, the workshop time required to crank out those parts is another source of expense. It’d be nice if someone had designed some sort of science-fiction contraption capable of creating three-dimensional parts according to a computer drawing. 

But of course, you can’t 3D print lens rehousing parts, especially on a home-user 3D printer. Can you?

Yes, yes you can

Well, for some value of “rehousing,” it seems that you can, because Helios-44 and Mir-1V lenses, among a few others, have started popping up on ebay in housings that look to have been run up on a filament deposition modeller. It’s hard to figure out if there’s a company name to ascribe to this, but the eBay account claims to be based in Kiev and also offers things like elliptical aperture masks to simulate anamorphic bokeh.

This is not a review, but the immediate reaction of the purist is to scoff at this. 3D printing is a technology at about the same stage that colour inkjet printing was in about 1993, when “colour” might mean vague, rainbow-hued blotches. In a similar way, FDM prints of cylindrical objects generally look as if they were turned up on a wood lathe with a butter knife. Some people would say that means the finish on the housings is about as good as the absolute performance of a Helios-44 in the first place.

A full set of Nikon Series E stills lenses (and a Tokina) with simple modifications for motion picture work - 3D printed gears, step rings for consistent front diameters, and mount adaptors

That said, your narrator owns a set of Nikon’s very basic Series E lenses with 3D printed focus gears (shown in the accompanying images), and can confirm that the only real problem is that modern remote focus devices are sometimes sufficiently powerful to turn Nikon mounts into ugly metallic origami when calibrating their end stops.

It looks like these lenses have been not so much re-housed as additionally-housed, dropping the complete lens into a new shell. Modern conveniences like consistent focus and iris gear positioning rely on some quite large chunks of metal (or, er, some sort of easily-melted plastic). In practice, each of a range of rehoused lenses often ends up being slightly longer than the longest donor lens while being slightly wider than the widest donor lens, which is one reason why movie lenses tend to be huge. Stills lenses are generally pretty compact to begin with, so this over-housing approach has some space to play with.

3D printing: getting better all the time

Of course, the underlying lens might still have all manner of wobbly bits which this kind of effort won’t fix. Bayonet style still photo mounts will always struggle to hold precision back focus because that wasn’t a requirement in their original application. The M42 mount often found on old lenses is actually pretty sturdy, being nothing more than a 42 by 1mm metric thread. Even M42 can struggle with consistent back focus, depending on how long it’s been since the lens last met a kindly technician with a cleaning cloth and some fresh lube.

Possibly the better approach to this sort of thing is the selection of parts sold by companies such as Simmod, which offers cine conversion kits including mount adapters, focus gears in a wide variety of diameters, and front ends which provide a standard outside diameter for accessories. These are properly machined focus gears and metal fronts, although that approach doesn’t give us the consistent gear positioning. Neither approach solves any problem which might be intrinsic to the donor lens’s mechanics.

The fit and finish on 3D printed lens gears is nothing like a properly machined part, but it's not as if lenses are turning at thousands of revolutions per minute

The best thing about 3D printing is where it might go in the future. It’s a hugely popular technology and it seems likely to enjoy a lot of development. When it becomes really, really good – and particularly when precision metal printing becomes more accessible – this sort of thing might end up being even more useful. Inkjet printing was a joke for years, but it is among the very best technologies now available for printing photographs. If the same advancement happens to 3D printing, well, the average metalwork shop of 2050 might look very different, and we might all be able to keep a full set of internal organs.