Voigtländer 21mm f/1.8 Ultron

Price: 
3
Build: 
4
Image: 
4

Last updated February 5, 2014

Important specifications

Focal length 21 mm
M8 equivalent 28 mm
Aperture range 1.8 - 22 (1/2 steps)
Aperture blades 10
Focusing range 0.5 m – infinity
No. of elements/groups 13/11 (1 asph)
Angular field, diag./horiz./vert. 90°/80°/58°
Filter M 58 x 0.75
Dimensions (with caps) ø 69 mm, length 92 mm
Weight 412 g
Produced 2013-

Introduction

A full review of the Voigtländer 21mm f/1.8 Ultron lens, including specifications, overview, sharpness and vignetting tests as well as sample images and links for further research.

The Voigtländer 21mm f/1.8 Ultron was announced at Photokina in September 2012 and became available surprisingly quickly in late January 2013. It's the latest in a string of fast lenses and it's in good company; the 28mm f/2 Ultron, 35mm f/1.2 Nokton, 50mm f/1.1 Nokton and finally, the 75mm f/1.8 Heliar. Voigtländer enjoys building lenses as a niche within a niche. That is, taking lens design - typically speed or ultra-wide focal lengths - just a little bit further and not necessarily competing head-to-head with the Leica and Zeiss offerings.

While popular, the 21mm focal length isn't necessarily the most popular with the rangefinder crowd (where 35mma and 50mm tends to reign) - it certainly is among street and reportage shooters. However it's the specs and price point of the new Ultron that has gotten everyone's attention. Especially if you consider crop sensor bodies such as the M8, R-D1 and mirrorless cameras. At f/1.8, the nearest competitor is Leica's own Summilux-M 21mm f/1.4 ASPH lens at some five times the price or the Super-Elmar-M 21mm f/3.4 ASPH at two stops slower. Even amongst Voigtländer's own lenses, the only other competition is the 21mm f/4 Color Skopar. Zeiss on the other hand, has two lenses that fall in between. The Biogon T* 2,8/21 ZM is a stellar performer and reasonably fast but unfortunately, the C Biogon T* 4,5/21 ZM doesn't work well with the M9 (it requires using CornerFix during post-processing as an intermediary step). Then there's the huge inventory of older 21mm lenses.

But this one, new with warranty - opens up to f/1.8 and costs just under $1,250 USD!

Physical Aspects

At first glance, the lens can be a bit misleading as to its size and weight. One reason is the integrated petal hood - and it is not removable. We prefer the option of being able to remove the hood as sometimes you just don't need one and want a smaller package. Sometimes you damage the hood and want to replace it. Or maybe you want a deeper hood because you're shooting with a crop sensor camera. The hood isn't so bad though - it's a petal hood and therefore isn't as bulky as a full round (or rectangular) hood. If you were to compare the lenses side-by-side, it would be quite similar to the 35mm f/1.2 Nokton (for example this Mark I on the right):

So how about the weight, then? While it might be similar in size to the 35mm f/1.2 Nokton, it does not share its weight - despite having 13 elements (in 11 groups, 1 element being aspherical). It weighs nearly 3oz. (78g) less... Even less if using the optional hood on the Nokton. Furthermore, the distribution of the weight of the lens is such that it feels evenly balanced on the camera as opposed to more front-heavy with the Nokton. This doesn't sound like much but it's noticeable. Handling the lens itself, you get the sense it's built well, solid metal and glass and similar to other Voigtländer lenses... The location of the aperture and focus rings, the combined knurling and scalloping of the latter, the smooth black paint. The markings are engraved and paint-filled in white and red and of a similar font. If you've got another Voigtländer lens handy, you can get a sense of what to expect.


Two minor things to note here. First, the aperture ring feels great. It's looking like Voigtländer has finally got it down. The detents are very positive if perhaps a bit "loud." There's no mistaking them, either tactilely or audibly - which is useful in helping you keep track when changing the aperture sight-unseen. There's no "klack" when hitting the wide open end as with some older lenses (which oddly never happened on the opposite end). Overall, the action is clean, positive and more refined feeling. Secondly, while it's nicely damped and smooth, you will likely notice a sudden change in resistance when the focusing cam disengages at the .7m mark - it feels a bit lighter than the rest of the range. This is because M bodies won't focus closer than .7m - only on the Zeiss Ikon, Voigtländer Bessa and Epson R-D1 bodies will focus down to .5m with the lens. In actual use you won't really notice it though. Since you can't focus closer than .7m, why turn it any further...

Let's take a moment to expand on the minimum focusing distance of .5m. The difference in depth of field between .7m and .5m is about half - from 11.7cm to 5.8cm (4.6" to 2.3"). So what do you do if you want to take advantage of this great opportunity to both focus closer and/or blow out the background just a bit more? Make a mark with a Sharpie or piece of tape on your camera strap at .5m (19.7"). Even at minimum focusing distance wide open, you still have a "fudge factor" (depth of field) of just over two inches!

The biggest complaint we have with the lens is not even with the lens itself... But rather the lens cap! If you've used Zeiss ZM lenses before, you know just how ridiculously fiddly the caps can be when trying to get them back on the lens with the hood mounted (not an option with the Ultron). While the cap itself feels a little more substantial - the handling is easily as bad as any of the ZMs.

Optical Qualities

The whole reason for owning this lens is the f/1.8 maximum aperture. Only one lens is faster - the aforementioned Leica Summilux-M 21mm f/1.4 ASPH... And at five times the price. If you don't plan to shoot at this aperture often (if not most of the time) then there are many alternatives out there that might be a better fit for you. So wide open performance is rather important. How does the Ultron fair?

Wide open, central sharpness is excellent and quite good up until the edges and corners where it lags behind a bit. Slight field curvature there and not totally unexpected - but even worst case scenario, the corners are surprisingly good. In the situations you'd be shooting a 21mm (e.g. street or reportage) or wide open at f/1.8 (poor lighting or for isolation) this shouldn't present much of a problem. More often than not, these areas won't be in focus anyway. Stopping down even slightly rapidly improves the sharpness across the frame until about f/4 or f/5.6 where it's fairly even and peaks. Stopped down as you might for landscape shooting, around f/8 to f/11 - the sharpness is quite even from corner to corner and presents a very nice image, though not what some would call "clinically sharp." It lacks that "bite" or micro-contrast of the Zeiss ZMs or newer Leica lenses that make you say "wow."

Here's a 100% crop (full image in inset) which, granted - was shot at f/8 - but gives you an indication of what the lens is capable of:


Here's a 100% crop (full image in inset) which is wide open at f/1.8 to give you some perspective:


The images from the Ultron are sharp; don't misinterpret. They're just not overly contrasty and in your face; it's slightly subdued. We noticed this when shooting most of the sample images - which were primarily landscape shots, stopped down. They also happened to be after a major blizzard with the sun poking out and lighting a scene that was 90% snow! Because of this however, the Ultron was able to capture very subtle nuances of the light playing across the finely textured snow. Highlights were nicely controlled and kept from blowing with minimal effort (essentially a +2/3 exposure compensation dialed in for the snow and nothing else!). The rendering was quite nice and ultimately, the sharpness is plenty for all but the pixel peepers. For what it's worth, the Summilux-M 21mm f/1.4 ASPH won't offer you much better. You'd have to shoot either the Biogon T* 2.8/21 ZM or Super-Elmar-M 21mm f/3.4 ASPH. Increasing the image contrast with a slight S-curve and applying a conservative Smart Sharpen (e.g. 30%, 0.3px) will get you in the ballpark. The Ultron is still sharp enough to cause aliasing and moire in very fine details - checking "Defringe: All Edges" (in ACR for example) will help clear that up at a slight cost in contrast.

From wide open to f/5.6 and even f/8 there can be some minor purple-fringing. This shouldn't come as a surprise with such a fast lens; even Leica's best lenses exhibit purple fringing. But by f/8 we expected it to be cleaner (if not gone). This happens in the extreme contrast transition points - the sensor blooms trying to accommodate the intense light. A typical cause is an overexposed background (e.g. sky) behind other fine detail, such as tree branches. You can limit this (if not prevent) in one of three ways; underexpose and/or stop down a little - or buy a slower lens. There are also methods to deal with this in post.

Longitudinal chromatic aberration (LoCA) and spherochromatism is also present. Even at f/8, it is evident (especially) in the corners as longitudinal chromatic aberration and wider open in the bokeh as spherochromatism. Granted, ours presented some extreme tests in multiple "danger zones" of tree branches against a bright sky, snow backgrounds and other harsh, contrasty transitions - though nothing was overexposed. Many of Leica's newest lenses are APO (apochromatic) and go a long way to correct for this. Thankfully it cleans up during import super easy in most cases. In fact, ticking the "Remove Chromatic Aberration" checkbox during import (via ACR) removed all but the worst of it.

The two automatic corrections applied during import in ACR cleaned up most of the aliasing and aberrations - but it did not completely remove the purple fringing. Here is a 100% crop comparison of the very top left corner of an image (which was the worst example that we could find), with the left side as-shot and the right corrected:


Oddly enough, when not shooting snow scenes - these aberrations never reared their ugly heads! Even in typical, yet high contrast scenes - practically nothing! If you're the type to shoot tree branches against an overexposed sky, then yes - you will be disappointed, and probably so with most lenses, especially faster ones. In our extensive testing of more "typical" shooting it's a non-issue. In those rare cases where some LoCA was visible, the aforementioned automatic corrections removed just about all of it. We wouldn't really consider this much of a problem.

Distortion is extremely well controlled. In testing, it was fairly hard to discern and more often than not, as with any wide angle lens (let alone ultra-wides like the Ultron) camera misalignment was more a cause for issue than actual optical distortion. With a 21mm lens, unless the camera is fairly close to level in multiple axis, you will have lines that lean or converge. With the Ultron, you can be sure that those lines are fairly straight at least. If anything, the lens has a very slight barrel distortion - but you'd have to look rather closely, dragging guides in to compare against. Compared to some other (even relatively recent) lenses from Voigtländer - this is most impressive. In extensive testing, distortion has proven to be mostly a non-issue across the board.

Flare is handled surprisingly well with the Ultron. Despite pushing the lens, shooting both into the sun and having the sun just outside of the frame - we managed only a few shots where it was even present; and not all that obvious either. In most cases, you get a lovely wash of light emanating from the light source into and around objects in the scene closest to the light source. Contrast drops slightly in this region but remains high in the rest of the scene. At worst, perhaps a tiny greenish blob or magenta blotch - easily cloned or spot-healed out. The worst case scenario we noticed was a bright ring around/near the light source... But this took shooting directly into the unobstructed, glaring sun or having it right on the border of the frame... Even then, it was no guarantee of producing flare! With 13 elements and such a wide angle of view, this is important - and impressive. Sun stars (when stopped down) are nicely defined and look really great, with ten points. You can see them in some of the sample photos below.

Finally, there's the matter of bokeh - something that's hard to quantify, and with a 21mm lens you'd think - hard to come by as well. Normally, yes - but wide open at f/1.8 and with a subject that's close (especially .5m close!) the depth of field is surprisingly thin (5.8cm or 2.3"). And with the Ultron, it looks nice. It's got some of that 35mm f/1.2 Nokton mojo happening. It's generally rather smooth and abstract. While there's a little bit of nisen-bokeh (double line) going on, it's rather minimal. It's also rather predictable from shot to shot. Which is unlike the 50mm f/1.1 Nokton - being neither predictable nor smooth most of the time. It can be, but more often than not it's quite nervous. The Ultron should please the bokeh aficionados out there and if this is part of the reason you're looking at a fast ultra-wide, you should be happy.

Coding for Digital Ms

When used with film or uncoded on a digital body - the lens vignettes quite a bit wide open, which clears up as you stop down - but it stops improving after f/4. This is to be expected, especially on such a fast and wide lens. Turns out, if you're shooting this lens on a digital M that you have two very interesting choices available. You can code the lens as an Elmarit-M 21mm f/2.8 (11134) and the vignetting will clear quickly (better than f/8 uncoded by f/2.8!) - or code the lens as a Summicron-M 28mm f/2 ASPH (11604) and allow some of that natural vignetting through (more so towards wide open until f/2.8). The downside of the latter of course, is that EXIF will indicate a 28mm rather than a 21mm lens, which would be even more confusing if you also own a 28mm - especially the Summicron (as we do).


We found that giving his lens the 6-bit code of the Leica Elmarit 21mm f/2.8 works "best" (most even correction) on both the M8 and M9 bodies. The code for this is 011000 and can be either marked with a template and a Sharpie marker or machined into the flange and properly painted for a more permanent solution via a third-party. If you go the hand-coding route, you'll be pleased to find that there's a groove machined into the lens mounting flange to better hold your coding without wearing off. There are no screws near the coding area. In fact, our hand coding worked on the first attempt! It's nice to see third-party manufacturers paying attention to these two critical areas for use with digital Leica M bodies.

One thing we noticed after more extensive testing was that aperture settings below f/2.8 were not recorded into EXIF - because of the Leica Elmarit 21mm f/2.8 coding! We're going to test the corrections for the Summilux-M 21mm f/1.4 and if that works okay, change the coding of the lens. This should solve that problem.

Update: We re-coded the lens as a Leia Summilux-M 21mm f/1.4 ASPH successfully - which works very nicely. No odd colorizations, vignetting well-controlled… And apertures properly noted in EXIF. We'll add some test photos (especially vignetting correction) shortly.

Viewfinders

People have a love-hate relationship with external finders. Some people just outright prefer not to use them, which is understandable. They add bulk to the camera, they're another item to worry about losing and they require some additional effort to use, switching between focusing and composing. Such is the nature of rangefinder cameras. So what to do? If you use a particular focal length enough, you can get a feel for the framing and just sort of zen the composition, perhaps straightening and cropping in post or darkroom as-needed. But what about ease of use or if framing needs to be more precise? At 21mm, you're pretty far outside of what the camera's viewfinder edges offer - and while you can stretch your eyeballs into the corners to get somewhat close, it's far from ideal. You can get away with not requiring a viewfinder for 24-25mm lenses, but at 21mm you might want to start seriously considering one.

There are quite a few options available when it comes to 21mm viewfinders. Leica, Zeiss and Voigtländer (among others) all offer one (or more). Leica and Zeiss however, tend to be rather expensive - and while it's hard to beat the Zeiss viewfinders optically (they're very bright and have minimal distortion) they're also rather large and boxy. If you don't use the 21mm focal length all that often it's hard to justify an expensive viewfinder. All of these reasons is why we prefer (and recommend) the Voigtländer options - of which there are several. You can go strictly with an (older) 21mm-only viewfinder or go with the newer 21/25mm combo. One option is a plastic finder for $189 USD and the other metal for $209 USD:


Unless cost is really important, we'd suggest going for the newer metal version. For starters, it's all metal, including the foot. This is important because if it's going to break, the foot is the weakest link. It's also quite a bit smaller. While it is said to be optically a little bit brighter and less distorted than the plastic version, the difference is very slight. Both have a bit of barrel distortion. What you see is an outer 21mm set of framelines containing a smaller set for 25mm. There's also a parallax correction mark for when you're focusing up close rather than infinity.

Conclusion

Variety is the spice of life, as they say. While there are certainly enough 21mm options available for M cameras out there, if you compare apples to apples - the Ultron to the Summilux - the Ultron represents exceptional value. It offers rather similar performance (if not better actually) in terms of sharpness, aberrations, depth of field/bokeh for a lot, lot less money. You pay for this speed with trade-offs in imaging performance, the size and bulk and of course weight of the lens. Such is the nature of fast lenses in photography; as with racing - speed kills. Nevertheless, the Ultron holds its own well in this essentially direct comparison.

If ultimate imaging performance is what you're after (with the implied trade-off of lens speed) you may want to consider either the Biogon T* 2.8/21 ZM or Super-Elmar-M 21mm f/3.4 ASPH instead. The Super-Elmar-M currently represents the pinnacle of performance in 21mm lenses for the Leica M platform - but does so at the clear expense of speed. While f/3.4 isn't necessarily slow - it's not necessarily fast, either. The ZM lens is both faster and (again much) cheaper for a similar performance level. These lenses are also much smaller and lighter than the Ultron and share a more common filter size of 46mm.

If you needed one lens that could handle anything you throw at it at a reasonable price point and with imaging characteristics that would make the majority of shooters happy, the Voigtländer 21mm f/1.8 Ultron should place relatively high up on your list of contenders. While it doesn't have the ultimate imaging performance, it's no slouch either - and it open up to f/1.8... A very exclusive club at this focal length with really only one (very expensive) alternative. Especially worth noting if you plan to use this lens on the M8 where it makes for a 28mm equivalent, or on a small-format camera (via adapter) where it's closer to a 31/42mm focal length. We predict this lens will become a real classic - up there with the Nokton 35mm f/1.2 lens, as it's got that same mojo going on. This lens is a real sleeper, and well worth the asking price. It's a keeper!

One line summary: if you want speed, get the Ultron. If you want the ultimate image quality get the SEM. And if you want something in between, get the ZM.

Vignetting

Vignetting was tested using the M9 at ISO 160, firmware version 1.196:

Lens detection set to OFF

f/1.8

f/2.0

f/2.8

f/4.0

f/5.6

f/8.0
 
Lens coded as an Elmarit-M 21mm f/2,8 (11134) and lens detection set to AUTO

f/1.8

f/2.0

f/2.8

f/4.0

f/5.6

f/8.0
 
Lens coded as an Summicron-M 28mm f/2 ASPH (11604) and lens detection set to MANUAL

f/1.8

f/2.0

f/2.8

f/4.0

f/5.6

f/8.0

Sharpness

This test is preliminary, and only tests the center performance - but it should give you an idea of what to expect. It was conducted by shooting a test target at .7m (27.6"). Images were shot at the various apertures with an M9, mounted on a tripod. No alterations were done to the image except for auto level and they are 100% crops:


f/1.8

f/2

f/2.8

f/4

f/5.6

f/8

Sample Images

We will be adding additional images to illustrate bokeh and dark scene situations as soon as we can!

Flickr Pool Images

Further Research

Sample images thread in forum  
   
21mm f/1,8 Ultron Voigtländer
Voigtlander Ultron 21mm f1.8 Aspherical Flickr
Fast M-series 21mm lens shootout review Ron Scheffler