When I became interested in anamorphic filming, and being me*, it didn't take me long to pull an anamorphic adapter apart to see how it works. And something very significant struck me, something which is the key to understanding and developing a single focus solution.
A typical anamorphic adapter gets LONGER as you focus it towards infinity, whilst the taking lens, to which it is attached, gets SHORTER.
A guide TO BUILDing
SINGLE-FOCUS ANAMORPHIC LENSES
using a new method
This basic industrial looking lens is a robust single focus anamorphic lens that weighs only 1kg, that can be handheld or directly attached to a tripod or gimbal. Flares are organic (orange) and the lens is acceptably sharp everywhere and has full coverage on Super 35. It is a separate rig from the camera and therefore attaches to a body as easily as any other lens. It accepts 77mm filters, and even works quite well with a wide angle adapter attached to the front.
It took hundreds of hours and well more than 50 different iterations to perfect it optically and mechanically (it still remains a little aesthetically challenged) .
(Build components are: an externally and internally modified Nikkor 50mm f1.4 (slightly throttled to f2 to achieve clean oval bokeh fully open), joined via a separate focus helicoid to a stripped back push-pull focus Elmoscope II anamophic adapter internally modified for close focusing to 0.7m, all fixed in perfect alignment via found clamps (old lens tripod collars) to a baseplate (modified steadicam plate) that supports everything, allows relevant adjustment and facilitates single-focusing. The Nikon mount is retained for use on Sony or Fuji with mount adapters/speedboosters, or use with Canon mount adapter on a BMPCC 6K or similar).
The following assumes that the reader has a basic understanding of the use of anamorphic adapters in front of a taking lens in order to establish and affordable way to film anamorphically with footage that later gets digitally stretched back out in post-production. Otherwise, read this.
Big budget productions use real anamorphic lenses. Those of us indie DP's who love the anamorphic look are often relegated to building a rig comprising an old anamorphic adapter paired with our favourite old prime lens. We, collectively, are my intended audience. I'll try to listen to my own advice.
I'm going to show you how to pair the two in a way no-one else has done.
The new method overcomes the rather large disadvantage inherent in anamorphic adapters attached to taking lenses of having to control two focus mechanisms very precisely in order to achieve focus. For static subjects this is a significant irritation, particularly as you generally have to rotate the rings backwards and forwards a bit independently to hit the sweet spot. For moving subjects, moving camera positions or for smooth cinematic focus pulls it becomes an impossibility. The method overcomes this (without the addition of any additional optics (such as the existing concept of creating single-focus using a front-mounted variable dioptre which is relatively expensive, adds front-heavy weight and degrades optical performance. There is a video by the Anamorphic Shop in USA that shows this concept in use and the lack of sharpness is in my opinion at the point of making it unusable for anything more than small indie art projects).
So what was that discovery when I first pulled an anamorphic adapter apart again? It's actually very simple. Here it is:
A typical anamorphic adapter gets LONGER as you focus it towards infinity, whilst the taking lens, to which it is attached, gets SHORTER.
Therefore, and this is the "aha" moment, if you can find a suitable taking lens that, whilst focusing from one point to another, increases and decreases in length by the (reciprocally) same amount as the anamorphic adapter, and you can somehow hold the front of the entire system in one place in order to keep the overall length the same, then the action of focusing the taking lens automatically focuses the adapter, by the action of the front of the taking lens pushing and pulling the adapter into focus. Brilliant.
But is it workable?
In short, yes. But be prepared for a journey of note. Hopefully I can help you through some of the issues I've contended with. And I hope you're also a bush mechanic.
Firstly, it assumes you can make the focusing mechanism on the adapter a simple push pull system.
[The type of adapter required is the typical projector lens for 16mm film, such as a Proskar, a Kowa, an Elmoscope etc. These all get focused traditionally using a twist ring, but (very easily) removing this ring reveals a very simple push pull focussing mechanism underneath. Their construction is very simple. Basically the front barrel holding the front elements slides inside the rear barrel holding the rear elements, bringing the two sets of lens elements closer together to focus the adapter on closer subjects. The closest an adapter can focus is when the mechanism prevents any closer sliding of the two barrels. This is either because the 2 or 3 brass guides have reached the end of their slots, or the barrels physically touch. This is typically 1.5m, or 5 ft. Grinding the barrels in appropriate places makes for a far more useful close focussing distance and this is what I do to my various builds in addition to the matching of focusing distances that forms the core of this webpage.]
Adapters of different brands and designs vary in how much they lengthen during focusing, so each final combination is going to be different, and that's part of the fun. I use a standardized measurement where I look at the difference in overall length between when the anamorphic adapter is focussed to infinity, and when it is focussed to 1.5m (5ft). Since most available adapters have 1.5m as their minimum focus distance this provides the most accurate and most applicable specification as it covers the whole focusing range. Unfortunately it is beyond the internet to know that specification, so you have to do your own measurement and I'm creating, hopefully along with your input, a table of information to help the process. Let's say, for example, the specification for a particular anamorphic adapter is 7mm. We now need to find a taking lens, of suitable focal length that doesn't vignette with the adapter attached, that also changes length by 7mm.
A slight discrepency in the measurement between the lenses can still be a workable solution, but might require stopping down a bit more to achieve accurate focus through the range.
The fundamental parts you require are in simple terms:
(1) An anamorphic adapter. of traditional design, with focus ring removed to make it a simple push pull focus lens.
(2) A taking lens that matches the adapter in terms of focus throw index. The lens must obviously extend when focusing (ie not internal focus) and it is preferable that the filter ring does not rotate.
(3) A method of joining the adapter to the front of the taking lens (as with most existing methods of using anamorphic adapters)
(4) A method of rigidly holding/attaching the front barrel of the anamorphic adapter to a mechanism or brace (see below) to hold it in one place. This will require ingenuity to create a suitable clamp, and has been one of the more difficult processes to perfect. There is no perfect way, as long as it works. Earlier versions of mine had no provision for filters or matboxes but I now incorporate a 77mm thread into that front clamp. The clamp mustn't protrude very far in front of the lens to avoid introducing vignetting.
(5) You need a rigid brace that goes from the clamp that holds the front of the anamorphic adapter, to a position that can be somehow fixed behind the focus mechanism of the taking lens to make them operate as one unit. This rear mounting point could be a specially designed clamp, or a modified point on the taking lens, or the tripod shoe on a mount adapter/speedbooster between the lens set and the camera body if applicable, or the tripod socket on the camera body itself or on a cage. The goal here is to create an overall fixed length, so that the lengthening of the taking lens during focusing shortens the anamorphic adapter and the corresponding shortenening of the taking lens (focussing towards infinity) lengthens the adapter (by pulling the rear elements of the adapter back away from the front elements which are fixed in place). I've evolved a lens rig that is "self contained", ie the brace goes only from the front of the adapter to the rear of the tking lens, so the lens rig can be swopped easily between camera bodies or have teleconverters or speedboosters (yes they work with anamorphic lenses with some limitations) used between it and the camera. So the design of your brace here has those sort of implications.
How to pair lenses with adapters to make use of this method:
Most anamorphic adapters that I have encountered designed for 16mm film seem to have an anamorphic focus throw index of between 5mm and 20mm. In terms of taking lenses, a 5mm focus throw corresponds to a typical 70-80mm lens (some macro lenses, and many zooms, can differ substantially though), which means any lens with a shorter focal length, such as 50mm, will have a focus throw that doesn't sufficiently move the adapter to achieve full range focus (not even close) .
A modified method to allow shorter focal length lenses to be used:
To address this issue with shorter focal length lenses, I have determined an additional method to push and pull the adapter, utilizing an focus thread helicoid pirated from an old lens (typically an old 300mm telephoto lens but could be from just about anything) and one part of this focus helicoid is fixed to the focus ring of the taking lens and the other part to the back of the adapter, to amplify the focus throw (well not really to amplify anything, it just can generate more throw in the first place than the inbuilt focus helicoid in short lenses). A lot of trial and error is required, and primarily the direction of the focus thread/helicoid must match what you are trying to achieve (left hand vs right hand threads, to suit either Canon or Nikon focus ring directions for example). A different specification is now required in order to match the adapter...how many degrees does the helicoid turn in order to push pull the adapter by the right amount? Then a taking lens must be chosen whose focus ring turns through the same desired angle, and in the same direction as the helicoid. This is more hit and miss (as we have now introduced more variables), but opens up a whole world of wide angle options, a luxury for full frame or super 35 sensors but often necessary for smaller sensors.
The concept is deceptively simple, and the resulting rigs look fairly basic, at least to date. They aren't really. They require some sophisticated aspects to be right for them to work.
Here are some significant characteristics that you must aim for in the rig and can be hard to achieve:
(1) INTERNAL OPTICAL ALIGNMENT. This applies to the use of any anamorphic adapter, not just in a single focus rig. NB: Anamorphic adapters often require tuning. The front elements must be in the correct alignment with the rear elements otherwise you cannot achieve sharpness no matter what. You can just about disregard every mention on the internet of which adapter is sharpest as it all depends on how that particular adapter is adjusted. Since most writers don't mention alignment it's safe to assume they don't know it's required/accurate/way off etc. The adjustment is usually (but not always) in the brass guide plates which are revealed when you remove the focus ring and have elongated holes. They hold the front and rear portions of the adapter in alignment, but that alignment is not always correct. Remember we are using 50 year old adapters. Loosen, twist front and back parts of adapter slightly until it's optically tack sharp, and retighten the screws that hold the brass guide plates. Redo a few times if necessary to get it right. Use the focus magnifier on a Sony A7R or equivalent high resolution camera to accurately assess the results of tweaking. So you'll need to have the adapter on a suitable lens and camera to do this process. And once you've tried to build a single focus rig with your adapter you'll have to repeatedly check this setting as you'll have twisted, dropped, forced, and aligned your expensive experiment more times than you'd like. They are fairly robust, but...
(2) PRIMARY FOCUS MATCHING. The bracing method you employ needs the capacity for adjustment so that the front of the anamorphic adapter can be shifted backwards or forwards relative to the taking lens to achieve primary focus matching between the two lenses. Once you have this set, the system (assuming everything else is compatible, aligned and set correctly) will rack focus accurately. In other words, aim the rig at a subject and get it sharp as you would have with a traditional two-focus setup, and then lock the front in place. Now the push pull effect (I should have called this webpage that) will do the rest for you when you focus using the taking lens only.
(3) PHYSICAL ALIGNMENT. Alignment between the two lenses is crucial. Not so much optically but to make focusing free of binding. Pushing and pulling tubes with close tolerances quickly leads to binding and very stiff focusing unless they are aligned. Creative engineering methods like shimming are often required when joining the front and back holding clamps to the brace to make small height changes to get alignment. Trial and error, and some luck. Surfaces should be square...if they aren't you will frustrate yourself with alignment going out as soon as you tighten screws that hold the bits together.
(4) OPTICAL SQUEEZE ALIGNMENT. Anamorphic lenses squeeze the image horizontally. This needs to be exactly horizontally relative to the camera sensor otherwise your backgrounds and flares will look off (because they are off) and your footage will look unprofessional (because it will be).
So there needs to be a method that allows the adapter to be rotated and then fixed in the right orientation relative to the camera body/sensor (most think of this as aligning the adapter to the taking lens but since the taking lens produces the same image even when rotated it's much better to think of alignment with the sensor. This is to set your vertical and horizontal alignment (the same thing). This is typically tested using a flashlight and ensuring that the horizontal flares generated are in fact horizontal and vertical bokeh is in fact vertical.
(5) RIGIDITY. The whole system needs to be rigid. If the front or back fixing methods flex too much then longitudinal alignment varies during the focusing range causing binding, and flex also causes the primary focus matching to alter slightly, meaning that you don't achieve sharp results. Rigidity is also required between the two lenses, and anywhere where clamps/screws etc hold things in place. Electrical tape works wonders for testing possible rigs (you can join filters to lenses, lenses to lenses, focus helicoids to focus rings etc etc) but it has a bit too much stretch for more permanent use. For example, you could join the two lenses together using tape, but stretch in that tape typically manifests as a failure to achieve primary focus matching in one direction pulling focus but not in the other. Unless you understand if things are moving where they shouldn't then you'll be left scratching your head. Lack of rigidity further affects changes in focus smoothness and binding depending on how you are supporting the camera, as the weight of the lens or the camera body respectively causes flexing and a change in alignment. I still battle with that. Employing a top brace in addition to the one underneath would go a long way towards solving that. And that said, it doesn't matter where your brace is located anyway, although it's worth considering how it gets in the way of handling, either by hand or on gimbals or tripods.
*This document is derived entirely from my own efforts and development over the last few years and is first published here in July 2020.
I am an experienced bush mechanic, having fixed buses with a penknife (South Africa), motorbike suspension with epoxy glue (Malawi), leaks in inflatable kayaks with cable ties (Fiordland NZ), and built my own medium format panoramic camera. I was instrumental in inventing a method for core sample orientation in geological drilling rigs.
I now gladly rip lenses apart with gender and orientation neutral abandon and put them together again in creative ways.