Article Index


Resolution

Comparing lenses side by side is tough since any one metric in isolation would not give you the true feeling for the lens' capabilities. Resolution, while certainly an important factor to many users, is one such element - look at it in isolation and you can miss on a great lens, which might not be as sharp as you hoped it to be but still might have an overall pleasing draw. Now, I am not really going to compare Nokton 35/1.4 to the top of the line Summilux 35/1.4 ASPH - don't really have the 'Lux, nor I really think this would be a fair comparison. But I can draw parallels between Nokton and one of the three other 35-ish lenses I currently own, which all kind of fall within the same price bracket - Nokton 35/1.2, C Summicron 40/2 and Biogon 35/2. Only Nokton 35/1.2 gives us f/1.4 and frankly speaking I can't really see any significant difference between these two lenses at f/1.4 - they both have somewhat soft-ish center and are very soft in periphery. And at f/2 both Zeiss as well as Leica pretty much beat both Noktons to pulp across the entire picture frame. I will do a little bit more comprehensive comparison later and in the meantime, let's take a look at what Imatest tells us.

The chart below captures MTF50 results, as measured by Imatest using standard imaging target. Like in all other lens tests, I used focus bracketing to capture series of shots at varying focusing distances (1m to 10m) and selected the series with highest MTF50 numbers. Because Nokton shows focus shift at f/2.8-f/4, the chart was adjusted to accommodate for that. Nokton also exibits mild degree on field curvature (the chart is not adjusted to accommodate field curvature), but more on this a bit later.

According to the chart, Voigtlander Nokton 35/1.4 is a fairly soft lens at wider apertures - center image quality seems a little bit softer, but borders is where the problem seems to be, with both f/1.4 and f/2 clocking pretty dismal resolution per Imatest. There is a noticeable boost at f/2.8 both in the center as well as around corners and from f/4 on the lens seems to deliver fairly decent performance across the frame. Huh, well, softer corners at f/1.4 are expected, the main question is how soft they really are in real life - is this going to be a problem in real life or is this just a MTF50 numbers game?

 

Leica M8 (10Mp)

 

Before we look at the real life examples, let's first take a quick look at the crops from the imaging target, comparing borders at f/1.4 and f/8. Image quality at f/1.4 indeed noticeably softer with less overall definition compared to f/8.

 

Leica M8 (10Mp)

Image borders @ f/1.4 vs f/8 (Leica M8)

 

The crops below compare image quality in real-life like situation. All shots were captured at infinity. Eyeballing the crops confirms the trend set up in the MTF results - both center, and especially borders, are softer at f/1.4, however, while center quality is probably adequate to many users, corners are noticeably smudged and will probably annoy some users. Keep in mind that these are infinity results - with border quality deteriorating more at closer distances might cause further grumbling. And as mentioned earlier, Nokton 35/1.4 exhibits mild levels of field curvature which actually adds to the overall feel of poor image corners at wider apertures. How much all this going to be a problem, depends on your style of shooting though. Personally, while I shoot predominantly wide open with 35-50mm lenses, I tend to move closer to the subject where the DOF effect masks poor border quality and so softer corners don't bother me as much as the somewhat softer center.

 

Center

U.R..Corner

L.L..Corner

 

When comparing to my other 35mm lenses, most notably Biogon 35/2, I feel that Voigtlander Nokton is more subtle in how it renders wide open, both in terms of the resolution, as well as overall color/contrast. If Biogon is soft around corners at f/2 and reasonably sharp in the center, Nokton is soft throughout the frame (more so around corners). Biogon continues to dominate, particularly around corners all the way through f/4 - after that my eyes can't really tell the difference between the two. On the other hand, Nokton 35/1.2 ASPH almost mirrors its smaller version very closely, with slightly better center resolution in f/1.4-f/2 range and similarly soft corners. From f/2.8, both lenses seem to be equal in resolution. So if one is after sharpness with wide open aperture, Nokton 35/1.4 is not going to pass the bar, no matter from which angle you look at it.

Before wrapping up the resolution section, let's take a look at the simplified graph below which captures resolution dependency of Nokton 35/1.4 to the focusing distance to the target. Each point on the curve is measured as an average of center and border Imatest MTF50 across all tested apertures and at specified focusing distance. Notice the significant drop off in the resolution at closer distances, which recovers fairly slowly, primarily due to the lower resolution in the periphery.

 

 

Color & Rendering

ISO 640, 1/60, f/2, 35mm

Like with most other fast wide angle lenses out there, the rendering of Nokton 35/1.4 varies fairly significantly,depending for the most part on the aperture level used. At wider apertures, f/1.4 through f/2, the lens shows moderately low levels of contrast, which fall off further towards the periphery. This is not necessarily a bad thing, and is certainly quite common in fast primes - I think I've seen only one or two super fast f/1.4 lenses that preserve same level of contrast throughout the aperture range. If you're shooting portraits, then lower contrast should not bother you much as it actually gives you more flexibility in dealing with shadows and mid-tones. But expect some subdued colors and somewhat flat look here, requiring at minimum some post-processing to improve overall saturation and contrast. Stopping the lens down to f/2.8 or lower apertures solves the contrast issue - there is a noticeably boost at f/2.8 and if you like high contrast images, that's the minimum aperture you should be using. Colors also get more vivid and crisp, but shadow detail goes away a little bit. I typically try over-exposing the image by ~0.5EV (by the way, Nokton seems to consistently underexpose by 0.5EV to 1EV at wider apertures) and then re-leveling it in post-processing to recover some lost shadows. Sometimes it works and sometimes it does not but when it does, you should expect a nice, evenly distributed luminocity and moderately high contrast in the subject.

Overall, I find Nokton's color reproduction decent but not exceptional in any particular regard - the gamut seems balanced across all three channels albeit a bit compressed, even when shooting in Adobe RGB space. With shots made at wider apertures, I find myself resetting black point frequently, which brings global contrast to satisfactory levels at the cost of some quality degradation in shadows and mid-tones. You can obviously be more elaborate with post-processing to bring up as much detail as possible from Nokton files, I personally find somewhat lower contrast and softness wide open work out fairly well for portrature type photography, but all this obviously is subjective. The picture above is the typical case - I have re-leveled the black point to compress shadows a little bit, which boosted contrast and then added a little bit of saturation to reds to make the subject stand out a bit better. I should have applied a mask, but I was lazy and so you can see a little bit of reddish tint in the white wall in the background.

While discussing color rendering of Nokton, I should also briefly touch-base on chromatic aberration, which by the way is present with this lens. Per Imatest results, shown below, Voigtlander 35mm f/1.4 shows moderate levels of chromatic aberration, averaging ~0.7px around corners across all aperture levels. This is fairly benign when compared to other wide angle lenses, but notice consistency throughout the aperture range, which is rather unusual. In practice, I do see frail signs of CA around borders in high contrast transitions, but I really have to blow the image to 100% magnification to notice the fringing. Nokton also shows minor levels of longitudinal CA in the periphery, but those are also fairly negligible and should not bother majority of users. Also, with a fairly consistent halation (bleeding of color around bright objects) the lens creates an interesting style pattern that might appeal to those who like older lens designs from 50s and 60s, like Summarit and Summitar.

 

 

DOF & Bokeh

ISO 640, 1/60, f/1.4, 35mm

One of the main uses for a fast f/1.4 prime like Nokton, is obviously night-time, or I should say, low-light, photography. Up until recently, the only 35/1.4 game in town was Leica's Summilux. There were plenty of f/2 lenses out there, but that one extra f-stop seemed to matter quite a lot to the users, particularly to those who owned the kludgy, first generation Leica M8s. With M9's improved (compared to M8/M8.2) high ISO handling, this might no longer be the case, but there is still something magical about the f/1.4 aperture - mention to anyone that you own an f/1.4 lens and you are likely to receive wishful looks in return.

Nokton fares quite well here - I mentioned previously that I prefer using older, lower contrast lenses indoors or basically in dim, but evenly illuminated environments. Nokton, which was designed to copy the look of older primes, is certainly more versatile than say Summitar or screw-mount Summarit, and I would not hesitate using it in bright light, but the lens does have a very interesting (in my opinion) way of rendering images when used at wide apertures in low light. It feels subtle, non-intrusive, dreamy if you will (although I don't like using this word since it has been overused too much) - muted colors, smooth gradation transitions, low contrast, subjects literally blending into the shadows. If you like lenses with 'softer touch', this is probably one of the better choices among modern lenses.

While rangefinder lenses clearly have a number of advantages over the retrofocus designs, one of their main weaknesses is the MFD.. With minimum focusing distance of 70cm (typical for 35mm wide angles), Voigtlander Nokton is a far cry compared to top of the class 35mm SLR lenses like Zeiss Distagon 35/1.4, which has MFD of 30cm. This naturally puts some limitations to the DOF of Nokton, which, according to online calculators is 2.47cm at f/1.4 (compared to 0.55cm for Zeiss 35/1.4 at f/1.4). Nokton's sweet spot is somewhere between the MFD and 1.5m, beyond which the lens carries through too much detail because of the expanded DOF, even if you shoot at f/1.4. Personally, I tend to prefer this focusing range anyway and so it does not affect my style in any way.

If you take a look at the series of pictures below, you would understand what I mean by the 'sweet spot'. With wide open aperture and 'up close and personal', Nokton 35/1.4 renders background in smooth, wide brush strokes, blurring the detail for everything but the largest objects. Yet, as you move a little bit away, the lens brings almost as much detail into the frame as when shot at small apertures. After using the lens for a couple of weeks, I started finding myself shooting predominantly at f/1.4 and f/2, where the lens gives a little bit more character and almost completely avoiding small apertures, where I find Nokton practically indistinguishable from many other 35mm lenses.

 

DOF @ 80cm

ISO 160, 1/8000, f/1.4, 35mm

ISO 160, 1/250, f/8, 35mm

DOF @ 2m

ISO 160, 1/350, f/8, 35mm

ISO 160, 1/800, f/1.4, 35mm

ISO 160, 1/350, f/8, 35mm

 

With overall subtleness at wider apertures, Nokton's bokeh feel seems pretty neutral to me. I can't really put my finger on anything that would make me claim that I dislike it. Caveate emptor obviously here is on the dependency on the focusing distance and aperture used. But when used wide open or at f/2 and close to the MFD, Nokton clearly benefits from lower contrast levels and lower resolution - OOF highlights are evenly lit and edges are either smooth or neutral (you can see occasional harsh highlight here and there, but they are superficial at worst). There is occasional slight doubling of countour that you can observe here or there, but the artifact is mild. Low resolution does impact the feel though - subjects don't really stand out in any particular way, at least not until you do some post-processing. Nothing extraordinary really, but not bad either...

 

ISO 320, 1/500, f/1.4, 35mm