Leica 16-18-21mm f/4 Tri-Elmar - A Preamble on People Photography at Super-Wide Angles
A couple of weeks ago, a reader of the blog asked me if I had ever shot with the Leica 16-18-21mm f/4 Tri-Elmar. Rather dismissively, I said no. You see I was under the opinion that 21mm was already wide enough, and the addition of two wider focal lengths on the same lens was completely unnecessary. The way I saw it, 18mm was too wide for any meaningful use, let alone 16mm. So as you could imagine, I didn't believe the Tri-Elmar made much practical sense.
But then a thought occurred to me. I've been shooting at super-wide angles for as far back as I can remember. Back when my preferred camera was the Nikon F4 and the Nikon D700, one of my favorite lens was the Nikon AF-S 17-35mm f/2.8 ED. And more recently, I've updated my Canon system with the current version Canon 16-35mm f/2.8L USM III. Clearly, I don't have anything against super-wide lenses. In fact, I've even gone as wide as the Canon 12-24mm f/4L and the Nikon AFS 14-24mm f/2.8G. So from that perspective, why have I forsaken the Tri-Elmar?
I suppose the typical answer is it's super tricky to shoot super wide with a rangefinder. It doesn't help that M cameras do not support in-camera optical viewing or frame lines wider than 28mm (unless if the camera is customized, which makes no sense). Admittedly, that limitation can be resolved with an external optical viewfinder. However, it still doesn't take into account full visualization at super wide angles. While the external optical finder can provide a visual indication of what's in-frame, it cannot exactly display the extent of distortion for proper composition.
In other words, you will be framing the composition with some degree of visual impairment, since the optical viewfinder cannot provide accurate preview of what is actually being captured through the lens. As a result, you can never know with certainty if the image capture would end up giving your subject tree trunk legs, tentacle arms, or deformed heads. That is the real-world hazard with using super-wide lenses incorrectly, and why lenses like the Tri-Elmar tend to be a special purpose tool limited to landscape or architectural photography where distortion of the human form is less of an issue.
Even so, difficulty in handling distortion involving the human form shouldn't automatically limit super-wide angle lens from consideration for general use. Nevertheless, that is the prevailing consensus. One only needs to do a simple Google search for sample images taken by the 16-18-21mm Tri-Elmar to anecdotally confirm this observation. Examples of portraitures are conspicuously less represented at super-wide angles than landscapes and architectural image capture. But I can understand this attitude. It is after all why I've never considered it in the past - despite my attitudes towards super-wide angles with DSLRs.
Given the perception of how difficult it is to tame a super-wide angle lens, it may be beneficial to take a step back from diving straight into a proper review of the Tri-Elmar, and have an impromptu discussion on the technique of shooting super-wide. It's not like there is a pressing need to post a review as soon as possible. There are already many reviews of the Tri-Elmar, since it has been in production for over ten years. So in that respect, it might make more sense to demonstrate how the Tri-Elmar can be used for the more conventional task of people photography. In doing so, it should provide some meaningful context that may have eluded previous consideration, thereby illustrating how the Tri-Elmar could be applicable for general use.
Besides, I need to postpone my originally intended Tri-Elmar review until the next post. You see, I used a Sony A9 for this shoot. And from the perspective of a Leica purist speaking apples to apples, it didn't seem entirely appropriate to review a Leica lens with an off-brand sensor. I had picked the A9 over the Leica M10 because of greater ease in compositional accuracy. With the Sony, I have the benefit of accurate electronic viewing, thereby doing away with the inherent limitations of rangefinder composition.
Admittedly, I could've just as easily given the M10 greater compositional accuracy, with live-view shooting or an external electronic viewfinder. Alternatively, if I didn't want to deal with the M10's relatively clumsy electronic viewing solution, I could have just as easily adapted the Tri-Elmar to the Leica SL. However, the A9 has the benefit of a tilting screen, which in my opinion enhances the shooting experience of the Tri-Elmar significantly. Furthermore, the A9 has the added advantage of a rather blunt virtual horizon in live-view, which suited me more than the subtle solution offered by the SL.
Still, apples to apples comparison supersedes ease of compositional accuracy for a proper review. Nevertheless, shooting the Tri-Elmar with the A9 wasn't a complete waste of effort - or at least that is how I justify my goof-up. In familiarizing myself with the lens on the A9, the experience of composing with it unconstrained on a freely tilting LCD screen quickly brought me up to speed in dealing with distortions on a super-wide lens. The resulting effort enabled me to fine-tune my approach in optimizing its usability in people photography.
Not bad for backwards engineering a narrative for this shoot!
In dealing with distortion at super-wide angle, the cardinal rules are fairly simple and straightforward.
The least amount of distortion is in the center point.
The closer you are to a focal plane, the more any object in that focal plane will pucker out. By comparison, the further you are to a focal plane, the less any object on that focal plane will distort.
The closer an object is to the edge of the frame, the more that object will be distorted along the perspective lines, running from the corners to the vanishing point at the center of the frame. In other words, there is less distortion for a standing figure along the middle the vertical axis or a lying figure along the middle of the horizontal axis.
Minimizing distortion is easiest when the camera is completely leveled to the horizon line and parallel to the subject's main focal plane.
Shifting the camera out of the leveled position can cause dramatic distortion
Once you understand these cardinal rules of shooting a super-wide lens, the fun bit is how to approach the composition with the human form in frame. Again, the methodology is simple and straightforward.
Minimize distortion on your subject's face
Use distortion to enhance your subject's silhouette
Use distortion to create dramatic effect with the environmental documentation (like what's happening in the background).
Reduce compositional tension by placing the intended subject on the closest focal plane relative to other prominent objects. In other words, refrain from placing other object (like another person) on a closer focal plane, especially near the edge of the frame.
The rules and approach listed is a guide to minimize distortion and facilitate what a super-wide lens is intended to do in composition. It's a tool designed to document a subject closer up with as wide an angle of coverage as possible. This enables an environmental documentation of a subject in front of an expansive background, adding to the visual narrative. That is essentially the defining characteristic of a super-wide angle lens.
Beyond that, a super-wide lens is also extremely versatile - once you realize how it can be used to manipulate distortions. When handled with intent, a super-wide lens can distort reality, which in turn can capture a unique perspective in documentation. Optically, buildings can be made taller, streets and hallways can be made deeper, and the ground can look further away. But the singular most unique aspect of a super-wide lens is how it can visually enhance the subjects silhouette. The trick is to combine both the dramatic documentary perspective of the background while enhancing the subject's appearance.
Ultimately, this is why a super-wide shouldn't be excluded for consideration for people photography. It's far more useful from both a practical and creative perspective than a normal or telephoto lens. For this reason, and because of the obvious reason that I goofed up with my camera choice, I will proceed with reviewing the already seasoned Leica 16-18-21mm f/4 Tri-Elmar for my next post - which I am working on concurrently. In other words, I've already gone through the sample images.
All images shot on a Sony A9. All images have been optimized in Lightroom. All images shot at 16mm unless stated in the caption. All body images have not been cropped.
Special acknowledgement goes out to John who took the time to assist me with getting this Tri-Elmar, while still on holiday in Colorado. I'm sorry to learn you will no longer be assisting me any further, especially since I am a creature of habit. I hope you good fortune with whatever you do next.