Article Index

Resolution

With somewhat limited number of alternative 40mm lenses in Leica M or LTM mount (there are only two, to the best of my knowledge - above mentioned Minolta Rokkor 40/2 and current Voigtlander Nokton 40/1.4), benchmarking C Summicron against another lens becomes rather tricky. Personally, I would have loved to compare C Summicron to either Voigtlander Nokton 35/1.4 or even Voigtlander Color Skopar 35/2.5, both of which lie within the same price category as the Summicron, unlike the more expensive Zeiss Biogon 35/2 or Leica Summicron 35/2. Unfortuantely, I don't own either of the Voigtlanders and did not manage to secure one in time for this test. Hence I will settle on comparing C Summicron with Biogon 35/2, which is currently my default 'go-to' 35mm lens.

C Summicron clocked fairly good resolution in Imatest tests. The simplified graph below summarizes results, showing the best resolution recorded by Imatest in the 1m to 10m.test range. As can be observed from the chart, overall results are fairly consistent to many other 'moderately' fast lenses, with center resolution remaining decent to very good throughout the entire aperture range, while border quality suffering a little bit at wider apertures. Imatest measures ~15% difference in raw resolving capacity of the lens at f/2, which can result in significant enough visual difference. But, once stopped down to f/4, the lens starts to show fairly consistent performance across the frame, with the sweet spot remaining in the f/5.6-f/11 range, which is quite typical to most lenses anyway. Next, let's take a look at some sample images...

 

Leica M8 (10Mp)

 

The two crops of imaging target compare image quality around borders at f/2 and f/8. There is a visible to the naked eye difference here, with the crop taken at f/2 showing slightly lower resolution and contrast. I would not call quality at f/2 disastrous though - yea, it is softer, but should be acceptable to many users. Quality at f/8 is top notch and so there is nothing to comment about it. Next section will examine what the lens is capable of in a real-life situation.

 

Leica M8 (10Mp)

 

The series of crops below, taken with an APS-H type Leica M8, give us a glimpse of what to expect from C Summicron in normal shooting conditions. The lens was focused at infinity, with the distance to the target exceeding 20m. I used bracketed focusing and then selected series that had best resolution in the center as the main samples. I then also compared various series for variance in corner quality indicative of field curvature - basically, samples showcasing the sharpest center should also have produced sharpest corners among all series. After eyeballing sampels for qutie a while, I sweak I could see some minor differences in series, which lead me to think that C Summicron actually shows minor field curvature. However, for practical purposes, I'd not be bothered by the discrepancy in quality, since it is very minor and should not pose problems to users of the lens.

Going back to the resolution, you can observe from the crops below that center image quality is fairly good at f/2 as well as at f.8 (and in between). There is a little bit more contrast and definition at f/8, but we will discuss that in the next section. Corner crops, on the other hand, do show noticeable difference, with f/2 crops around corners being visibly softer. How bad is it? While obviously not clinically sharp, I think results are of acceptable to many users quality. When  I compare  these samples to images taken with Biogon 35/2, I can see a slight quality difference at f/2, with Biogon images carrying more resolution as well as contrast particularly in corners. By f/4 difference is gone and both lenses seem to perform equally well.

 

Center (Leica M8)

L.L.Corner (Leica M8)

U.R.Corner (Leica M8)

 

On the final note, before we wrap up with the resolution section, take a look at the simplified chart below, which captures the resolution dependency of the lens against focusing distance. The chart averages MTF50 resolution numbers captured by Imatest across all aperture positions and plots it against focusing distance. For example, a plot at 1m is calculated as an AVERAGE (MTF50 (Center,@1m,f/2), MTF50 (Borders,@1m,f/2), MTF50 (Center,@1m,f/4), MTF50 (Borders,@1m,f/4),...). You can observe that the lens seems to perform slightly better at longer distances predominantly because of the drop off in quality around corners at wide apertures.

 

 

 

Color & Rendering

ISO 320, 1/750, f/2.8, 40mm (Leica M8)

Over the years, I have always been surprised by the lack of any serious web coverage for Leica C Summicron 40mm f/2 . There are occasional 'Is this the right lens for me' and 'Which one: 40/2 or ...'  questions on various web forums along with a wide variety of sample images, but nothing comprehensive like a review or heat-to-head comparison. This partially contributed to my hesitations for picking a sample, always thinking that it would be safer to go for one of the 'better covered' lenses.

Assuming you're careful about your exposure (which sometimes becomes quite challenging with M8's hybrid center-weighted exposure metering), you will find that C Summicron carries in a fairly good amount of detail into its files. You can improve channel detail further by shooting in Adobe RGB mode, which covers a larger color gamut than the standard sRGB space, but also requires more tweaking in post-processing. The lens offers a nice balance between amount of contrast and amount of detail in mid-tones and to a slightly lesser degree in shadows. At wider apertures the lens produces lower contrast in the periphery and moderate contrast in the center, with contrast increasing steadily as you stop down the aperture level.

The only challenge with the C Summicron that I experienced was when shooting  in brightly lit environments. Here the lens consistently caused overexposure by ~1EV. It is not quite clear to me whether this is the result of the Summicron's single coating or the camera's metering limitations. Regardless of where the problem is, manually dialing down exposure compensation helps channels to more balanced levels, but sometimes at the expense of some detail loss in shadows.

Interestingly, the lens fared quite well in chromatic aberration. The chart below captures results of Imatest tests and as you can observe, C Summicron delivers consistently low levels of CA throughout the entire aperture range. If you examine pictures at high magnification rate, you would notice very minor signs of lateral as well as longitudinal aberration (look towards the edges in OOF areas), but the degree of artifact is so low that you would notice it under normal circumstances. There is a little bit of spherical aberration at wider apertures, which is not too distracting to my eyes either, so overall, I'd argue that the lens handled all kinds of aberration quite well - not too shabby for a 70s lens, with a single coating!

 

 

DOF & Bokeh

ISO 640, 1/45, f/2, 40mm (Leica M8)

With a couple of 35mm lenses in my arsenal, the 40mm C Summicron has never been my primary choice in photography. I opt for this lens only when I need to get an absollutely smallest and lightest footprint, which is pretty rare. My default 35mm rangefinder lens is Zeiss Biogon 35/2, which I find to be a capable, versatile performer, suitable for almost any situation. If I need to be creative with DOF, and I mean 'really creative', I go for the super fast (and super bulky) Voigtlander Nokton 35/1.2. That leaves C Summicron underutilized most of the time, which is a shame since C Summicron's performance in this area is fairly decent. But I am jumping ahead of the review.

As mentioned earlier, like with other older, single coated lenses, I tend to prefer using C Summicron in evenly illuminated environments, including low light, indoor shooting. The sample on the right, which is one of my favorite shots with this lens, was taken with ISO 640 at f/2 and a focusing distance of about 2m. I applied a little bit of post-processing to re-level blacks and correct magenta blacks using curves with a selective mask. Resulting image has a nice, old-style look and feel to it, with an overall pleasant, bit not too blurred background. Some of you might shrug and claim that there's nothing particularly special about how the bokeh is rendered here and you would be right - after testing close to 200 lenses, I feel that relatively few will fall into either 'absolutely marvelous' or 'absolutely worst' categories, with most lenses falling somewhere in the 'decent' category. But keep in mind that with normal and wide-ish lenses, including C Summicron, the focusing distance to the subject plays the major role in bokeh rendering. My personal evaluation criteria for the 35mm, 40mm and even 50mm lenses is all about how the lens draws in the MFD to ~3m focusing range - the typical shooting range for me when I am trying to get some separation between the subject and the background. And I find it to be a complete moot point to discuss the look of bokeh with a lens focused at infinity (super long telephotos are the only exception to the rule here) - take any two randome lenses of similar aperture and focal length, focused at infinity, and chances are that you most likely are not gonna be able to tell any difference between the two.

Fortunately, with the MFD of 80cm, C Summicron produces a reasonably shallow DOF at f/2 - about 3.5cm according to online DOF calculators. Change the focusing distance to 2m and DOF increases to ~22.5cm - huge difference for someone trying to achieve as shallow in-focus zone as possible. The sample below approximates the DOF scale - the lens was focused down to ~80cm with the camera set at about 45 degrees to the table.

 

ISO 320, 1/90, f/2, 40mm (Leica M8)

 

But let's take a look at some other real-life samples sine the ruler shot does not really tell us much beyond the DOF scale. Shots below showcase what you can expect from this lens in real life situation. If youv'e been reading other reviews on this site, you're probably alredy familiar with this setup - the scenery is not particulalry interesting, but is static and does not change over time, thus giving us a decent baseline to compare against. The distance to the farthest building is over 50m, while the distance to the closest one (the palm tree on the right hand side) is about 10m.

As is the case with most lenses, and C Summicron is no exception, three variables, namely aperture level used, focusing distance to the subject and focusing distance to the background elements influence most the look of bokeh. Unfortunately, C Summicron is somewhat challenged in this area - if you're after the smoothest bokeh possible, then your best bet is MFD with wide open aperture (as can be observed from the shots below). Here the lens does a decent job of blurring the background, achieving fairly good sense of separation between the foreground and the background. You can still distinguish outlines of medium and large size objects, but to my eye, bokeh is not too 'rough'. Moreover, as discussed in the previous section, the slightly lower contrast background at f/2 that C Summicron produces actually plays to the benefit here IMO. Stopping down the lens or moving away from the subject brings us into the 'uncharted' zone - DOF increases quite rapidly, thus bringing more and more detail into the frame and making bokeh look bland, practically non-existent. Examine the rendering in the shot taken at the MFD with the aperture stopped down to f/8 and compare it to the one taken at 2m, but with wide open aperture - do you see any difference? I sure don't...


DOF @ 85cm

DOF @ 2m

 

 

If you examine images at higher magnification, you would notice minor color shift in the OOF areas - highlights sometimes carry pinkish and cyanish outlines, hinting on the presence of  chromatic aberration.  There is also a hint of spherical aberration in OOF areas at wider apertures, although it seems fairly contained. OOF highlights have neutral look to them for the most part, although you can see occasional edge highlighting here and there. But either of these artifacts is not very pronounced and does not effect overall feel of image quality in my opinion.


ISO 320, 1/125, f/2, 40mm (Leica M8)