Article Index


Field Tests

Handling: When I first got this lens for testing, Cosina was offering only offering the Color Skopar ASPH 20mm f/3.5 in Nikon and Pentax mounts. Recently the company expanded this selection to Canon - an absolutely fantastic news for Canon camera users, as there is now no need to worry about getting the right adapter for the lens and using stop down metering! The Canon version of Color Skopar also sports an electronic aperture control, so there is no dedicated aperture ring to turn around. Like with Canon versions of Carl Zeiss lenses, which by the way are also manufactured by Cosina, I have always been split on whether to get myself a Nikon mount or native Canon mount lenses (Voigtlander and Zeiss I mean). If you standardize on your body, Canon or Nikon, your choice is easy - get the lens for the mount you use. But I have both Canon as well as Nikon cameras and use them interchangeably, and so after pondering this issue for a while, I decided to go for interchangeability and stick with Nikon mount, so I can use the lens on either body, even if that means going back to the adapter and stop down metering.

Anyhow, like the other Voigtlander Nikon mount lenses (58mm Nokton and 40mm Ultron), the 20mm Color Skopar ASPH is a regular Ai-S type, which means that it is pretty much compatible with most Nikon cameras since mid 70s. However, unlike completely manual Ai-S lenses of the past, Voigtlander incorporated a chip into the lens, which allows it to carry some information down into the camera and enables matrix metering (if your camera has one). Also, on native camera mount, the lens will allow you to stop it down by 1/3 if you use the electronic controls from your camera, but only by 1 full f-stop if you use its aperture ring.

The biggest surprise is how tiny this lens is. Obviously the fact that it is a manual lens and does not have any 'excessive' electronics helps to keep the size in check, but the optical design itself is the major contributing factor here. Plus the fact that the lens is fairly slow. I need to remind you here that the only two other lenses of similar focal length that can give Color Skopar the run for its money, err its size, are the older Olympus OM 21mm primes and the limited production Pentax SMC DA 21mm f/3.2 AL. Plus the rangefinder lenses, but those don't count, since they employ different design concepts. Now, if you are using a hulk of a camera like Nikon D3 or D3x, then the size of the lens is not going to be playing to your advantage - on my D3/D3x I always end up 'searching' for the lens and its focusing ring with my fingers all the time while looking through the viewfinder (the focusing ring is really thin on this lens) and often end up with my fingers on the front elements.

Because of the small size of the lens, mounting an adapter on the lens, or even mounting the lens itself onto a Nikon body gets a little bit tricky as well. The barrel has very little surface area, effectively a really thin strip of metal that has engraved DOF scale imprinted on it, that does not rotate. When you twist the lens to lock it into position or unscrew it from the camera, it's easy to twist the focusing or aperture rings, or both if you don't apply a kung fu grip on this thin strip. Twist the focusing ring too much and too hard and you will risk unscrewing the barrel and opening up the helicoid. This is probably the most annoying part of using the lens for me, but I obviously realize that Cosina did not have much choice here when trying to make the lens as small as possible.

One other issue to keep in mind is that as with other manual focus lenses, the camera does not provide focus confirmation. That is unless you use one of the cameras that have a rangefinder, in which case you can simply watch it turn into the green dot when the lens is focused on the subject. This is one main difference between the Nikon and Canon mount versions - the Canon version of the lens actually triggers all AF points, so you get both an blinkning red dot as well as sound beep when one of the AF points gets activated.

 

Resolution: The very  first Cosina's announcement of the newly designed Voigtlander Color Skopar 20mm f/3.5 SLII put the dozen or so most active online photography forums on fire. The level of excitement among the photogs was understandable - Cosina was known to often deliver very well designed lenses (APOLanthar 125mm f/2.5 Macro is the best example of that), and considering the scarcity of good quality, reasonably priced ultra wide angles was just adding fuel to the fire. Prior to Color Skopar, you had to either go 'high-end' with Carl Zeiss Distagon T* 21mm f/2.8 or its wider and slightly slower version Distagon T* 18mm f/3.5, or opting for a used and often discontinued gear from Leica, Olympus, Nikon or Pentax (of course this also assumes that you'd be willing to either surgically alter your lens or rely on an adapter). At about the third of the price of Distagon T* 21mm f/2.8, Voigtlander Color Skopar was the holiest of holy grails for many users, including myself. Well, after testing the lens on my recent trip to South East Asia, as well as after running it through my standard process, I have rather mixed feelings about its performance.

Voigtlander Color Skopar ASPH 20mm f/3.5 SLII is certainly not the sharpest lens on the block. It does not come anywhere close to the stellar performance of Carl Zeiss Distagon T* 21mm f/2.8 (or its older Contax version as a matter of fact). I doubt it even can match the old Olympus 21mm OM duo, the OM 21mm f/2 and the OM 21mm f/3.5. If I had to characterise the resolution capabilities of Color Skopar, I'd say it closely resembles that of a 20+ year old Nikkor 20mm f/3.5 Ai-S (and its modern AF-D  version, with one obvious difference - it's slower). This is not to say that the overall image quality Voigtlander Color Skopar produces is bad - center image quality remains absolutely fine at all aperture settings and regardless of the type of the camera, APS-C or  full-frame. Border image quality however, is where the weakness is. At f/3.5 and all the way to f/5.6, images look fairly soft around corners, even to the naked, untrained eye. This is where the camera type, or I should say the sensor type, matters - the high-resolution, full frame sensors of Nikon D3 and particularly Nikon D3x (Canon 5D and Canon 5DMk2 respectively for Canon mount) put significantly more stress on the lens and the images taken with these cameras look softer then those taken with lower resolution, APS-C body like Canon XTi. On a 10Mp APS-C body, all you need to do is stop down to f/5.6 to get fairly acceptable (after minor sharpening during post-processing) results. But on a full frame bodies you need to stop all the way to f/8 to start getting decent results.

All this is not that unusual for ultra wide angle lenses, so why mixed feelings here? After all, if you plan to use the lens mostly for landscapes, just stop it down to f/8 or f/11 and your problems are solved! I think I am just feeling some amount of disappointment that a new lens does not delivery better results then a 20+ year old design. Plus, the fact that the Voigtlander costs even more then the modern, auto-focus Nikkor 20mm  f/2.8D. Ohh well, can't get it all at once...

 

DOF & Bokeh: Unless you are planning to use Voigtlander 20mm f/3.5 SLII on a Micro Four Thirds camera with a 2x crop sensor, which turns your 20mm into a 40mm widish lens, the quality of bokeh is probably not going to be of significant importance to you. A combination of wide angle and fairly slow maximum aperture makes primes like Voigtlander 20mm fairly poor performers in this area - under most circumstances, DOF is rarely thin enough to create enough sense of separation between fore- and background, and because of that, the bokeh often feels 'flat'. But one thing at a time - let's take a look at the simulated DOF first.

 

ISO 400, 1/13, f/3.5, 20mm (Nikon D3)

 

The DOF is not particularly thin here - about an inch and a half, but it is really hard to judge the true impact with such a 'limited' shot. Since practically all lenses render background differently at various focusing distances. As a rule of thumb, the closer to the minimum focusing distance the subject is and the farther the background is, the higher is going to be the feel of separation will be. Naturally, aperture level is also a contributing factor in this, and as you can guess, you can compensate one for another, for example, a slower aperture setting can be compensated to some extent by moving closer to  your  subject and/or trying to compose with the background located further away. Anyhow, the two samples below show what you  can expect from the  lens in real life situation - the first set of shots was done at  the minimum focusing distance, while the second set was done at ~1m. No surprises here - the widest aperture setting and minimum focusing distance produce the heaviest blurring of the background. Same aperture setting but at 1m does not really blur much of the background, leaving everything fairly well defined. While you should expect majority of UWA lenses to behave quite similarly, one aspect of Voigtlander Color Skopar ASPH 20mm f/3.5 SLII you be careful about in such situations is the rather excessive color fringing. You can notice a fair amount of it in all shots, but particularly those done at the widest aperture level.

 

DOF@18cm

ISO 200, 1/400, f/3.5, 20mm (Nikon D3)

ISO 200, 1/80, f/8, 20mm

DOF@1m

ISO 200, 1/200, f/3.5, 20mm (Nikon D3)

ISO 200, 1/40, f/8, 20mm (Nikon D3)

 

Lateral chromatic aberration aside, the lens also shows minor levels of halation in OOF areas, as well as double-edging, which is not particularly disturbing. It is hard really to find anything particularly impressive in how the lens renders bokeh, but it is hard to blame the lens here - focal length here is the main factor and the only other lens that renders bokeh even worse then a typical 20mm prime is a 16mm fish-eye.

 

ISO 400, 1/100, f/3.5, 20mm (Nikon D3)

 

 

Flare: Most wide angle lenses suffer from some flare and Voigtlander Color Skopar ASPH 20mm f/3.5 SLII is not an exception here. The lens shows practically no resistance to flare - the two shots below demonstrate the case with the sun positioned near the edge of the picture frame at about 70 degrees on the right had side above the building. Wide open or stopped down, the lens exhibits fairly high degree of flare and you can notice ghosting as well as streaks of color shifts


ISO 200, 1/125, f/3.5, 20mm (Nikon D3)

ISO 200, 1/25, f/8, 20mm (Nikon D3)


 

Vignetting: As one might expect, Color Skopar 20mm shows some light falloff when used at its widest aperture. While the amount of vignetting is not the worst I have seen in an UWA lens, I am still a little bit surprised at its  amount given the relatively slow  maximum aperture of the  lens. On FF cameras, vignetting does darken the corners and this will be particularly visible in light-colored scenery. On an APS-C body vignetting might be visible but in a much more subtle form. From practical standpoint, you either want to stop down the lens to f/5.6 to reduce light falloff to minimal levels or simply use the camera's built-in vignetting correction.

 

 

Color Reproduction: The lens  shows fairly good color reproduction - color palette remains mostly neutral, although the colors seem to be in need of a minor boost in saturation. Amount of contrast in shadows and mid-tones sufferers a little bit at the widest aperture setting, but gradually improves as you stop down the lens. You can get quite nice and vibrant looking images at f/8 or f/11 - the optimal aperture range  fort the Color Skopar. As mentioned, the lens does produce fair bit of aberration, particularly around borders, so you need to be on a lookout for it and possibly deal with it during post-processing.

 

ISO 200, 1/200, f/3.5, 20mm (Nikon D3)