ImageRe logo

50/1.4 Shootout

Nikon Millennium S3's 50/1.4 versus Summilux ASPH 50/1.4

Two recent fast 50mm lenses for rangefinder cameras have elicited comments: “Oh my, but that's sharp!” from reviewers and users. The first of these is the 50/1.4 Nikkor which came with the special edition Nikon S3 Millennium, and is said to be a reissue of the 1964 50/1.4 Olympic Nikkor, but with updated coatings. As far as I can tell, no lens schematic, detailed description, or MTF graph is available to the public for either the original or the update on this lens. It is known that the original Olympic lens had 7 elements, and both a larger rear element and improved coating over the ordinary S-mount 50/1.4.

The other lens is the 2004 Leica Summilux ASPH 50/1.4, the update on Leitz's venerable 1960's 7-element 50mm Summilux formula. This lens has a floating element group for close-up correction, 1 aspherical surface, and a total of 8 elements. Lens diagrams and MTF curves are available for this lens.

What follows is an attempt to dissect and comment on these two lenses' performance on several common parameters. Both lenses were mounted on an R-D1 (the Nikkor using an Orion adapter) so that comparisons could be made of their digital output. Use of the R-D1 with its 3:2 field cut of course limits light rays from the outer field from hitting the sensor and being used in this comparison, but a good characterization of both lenses is still possible. All raw conversions were done in the latest edition of Epson PhotoRaw, with lighting “as shot” and with all adjustment set to zero. Monochrome images were converted in Convert to B/W Pro, again all with the same parameters. I looked at how both lenses drew and at histograms, and then at bokeh, at close-up performance (at 1mtr), flare resistance, issues of contrast and sharpness, and curvature of field, coma in the field, and vignetting.

Following are several paired shots under actual use, taken at a wedding reception. One shot was outdoors in bright overcast and at f/5.6 or thereabouts, one in bright sunlight, and two were in low-level interior light at f/1.4 (one with strong backlight in the picture). I could see little difference between the shots, except for more obvious flare on the Nikkor image of the man's face in the contre-jour lighting pair. Overall I got the impression that the Nikkor has a tad more sharpness, and the Leica lens very slightly more micro-contrast, but otherwise the lenses performed quite closely here. I am posting only two sets of histograms here as examples, as the actual amount of differential information conveyed by the entire group of histograms in this testing seems small to me. In the first set the Leica was registering slight greater micro contrast or saturation than the Leica lens (this is from the "bokeh" images at f/5.6 below), and the second is the f/1.4 close-up images (also below) which shows little difference at all, but both sets of histograms showed very similar width and thus similar macro-contrast.

Leica Summilux Asperhic 50/1.4 at about f/5.6, bright overcast
Nikkor 50/1.4 at about f/5.6, bright overcast
Leica Summilux at f/1/4 against the light
Nikkor at f/1.4, note the flaring
Leica at f/1.4
Nikkor at f/1.4
Leica at about f/8.0, bright sunlight
Nikkor at about f/8.0 as above
Leica histogram of an outdoor scene at f/5.6
Nikon histogram of the same scene. Note flatter "shoulder"-- this was shows up on the image as slightly lesser saturation or micro contrast. The width (macro contrast) of the two histograms are very close.
Leica histogram of a close-up at f/1.4
Nikon histogram of the same scene, showing little difference, but again very slightly less saturation or micro-contrast than the Leica image.
All other images besides those from the wedding were taken tripod-mounted. Three sets of shots of three bluebird houses illustrate color rendition and bokeh. They were taken at f/1.4, f/2.8, and f/5.6. In these shots, taken in medium overcast, the colors rendered by the Leica lens are more saturated and contrasty. The difference is slight. Out-of-focus areas seem very similar between the two lenses. There is none of the wiry sharpness or double imaging as seen with earlier 50/1.4 Nikkors. I personally find the bokeh effects of both lenses pleasing.
Leica Summilux at f/1.4
Nikkor at f/1.4
Leica at f/2.8
Nikkor at f/2.8
Leica at f/5.6
Nikkor at f/5.6

When I was looking around for close-up opportunities, I came upon the “found” scene on top of my bureau, and merely set up the camera and tripod so that the can labeled “Boot Stretch” was at 1 mtr. Reviewing the shots from f/1.4 on the LCD screen, I was quite surprised to find that the images differed distinctly in the focus on the can of silicone spray at the right edge. The Leica lens seemed to have considerably greater DOF here than the Nikkor. It immediately occurred to me that the floating element group in the Summilux might be playing a part. I reset the silicone can so that it was in the same plane as the Boot Stretch can and then both lenses acted much more alike. The Nikkor appears to have again slightly less contrast, but edge sharpness-wise the two are close. Clearly the floating element group helps some with center sharpness, and increased DOF at the edges. (NB. The Summilux focuses to 0.7mtr, while the Nikon body's (or the Orion's) external helicoid stops at 0.9 mtr.)

Additional information: the original first 6 images images below have been reshot a few days later, hence the somewhat different lighting. I did the reshooting becuase it was pointed out that the Summilux image's focusing was slightly "out" and I wanted to be certain that the obvious differences in DOF were not due to that misfocus. During the reshoot in addition to the images below, I purposely misfocused the Summilux both BEHIND and IN FRONT OF the target. Perusal of these images (not shown) gives more information on the change in DOF for the Summlux with its floating elements. The optical engineers have sacrificed DOF behind the image plane for additional DOF in the foreground.

Leica at f/1.4, far right can not in plane of focus
Same shot with Nikkot at f/1.4
Here are enlargements of the center and edge for the pair above.
Leica center at f/1.4
Nikkor center at f/1.4
Leica edge at f/1.4, can not in plane of focus
Nikon edge at f/1.4, much less DOF
Can on right moved into plane of focus, Leica at f/1.4
Nikon at f/1.4, right can now in plane of focus
Edge view of above image pair, Leica
Edge view, Nikkor
Leica, same shot at f/2.8
Nikkor at f/2.8
Leica center at f/2.8
Nikkor center at f/2.8
Leica edge at f/2.8
Nikon edge at f/2.8
Leica at f/4.0
Nikkor at f/4.0
Leica f/4.0 center
Nikkor f/4.0 center
Leica f/4.0 edge
Nikkor f/4.0 edge
In other test shots looking for flare, a framed image of an eagle against a bright background at f/1.4 shows very little difference, but the Summilux has a bit greater DOF and slightly higher contrast, again probably due to the floating elements. Distance from the target was about 1.5 mtr in these shots. In a tougher test shot of our cat Samantha strongly backlit with focus on her sunlit fur, the Leica lens held contrast in her face better than the Nikkor (because the cat moved her body slightly between shots, the focus on her face is out on the Nikkor, but one can still make comments on the micro-contrast.
Leica at f/1.4, 1.5 mtr. Note greater DOF than the following
Nikkor shot of the same scene, especially visible around the binoculars
Heavily backlit shot with the Leica Summilux at f/1.4
Same shot with the Nikkor at f/1.4
Leica of cat's face
Nikkor of same. Forget the sharpness and look at the micro-contrast
The early non-Olympic Nikkor 50/1.4 had a curved field and lack of sharpness in the corners. This Nikkor does not. In a series of shots of a hillside about ½ mile distant at f/1.4 and f/2.8, one can see that it's field is as flat as that of the Summilux. At f/1.4 once again we see, in shots of the center (a cellphone tower) and of the right edge, that the Leica lens is contrastier, but the Nikkor, this time at infinity, is sharper in both areas. At f/2.8 the centers are closer than at f/1.4, but the Nikkor's edges are better! The lenses' fingerprints in this type of use are slightly different, but still very close.
Leica shot of hillside at f/1.4
Nikon shot at f/1.4
Leica shot at f/1.4. cell tower at center
Nikkor shot of center at f/1.4
Leica right edge at f/1.4
Nikkor of edge. Note less contrast but more apparent detail, despite ? color fringing .
Leica Summilux at f/2.8
Nikkor at f/2.8
Leica center at f/2.8
Nikkor center at f/2.8
Leica right edge @f/2.8
Nikkor edge. Perhaps some color-fringing, but sharper.
I took a shot of a string of Christmas lights at f/1.4, and at f/4.0. Detail of the central bulb is similar at f/1.4, but there is some chromatic aberation / color finging. The edge bulb on the Nikkor shows definite coma as well. At f/4.0 the coma and fringing are gone, but they are replaced, both in the center and at the edges, by the star effect one usually gets from closing the diaphragm way done. F/4.0 is not small enough to be getting obvious diffraction effects, so the star must represent internal reflections from the diaphragm itself. Both lenses have 9-bladed diaphragms; the one in the Nikkor tends to be more circular while that pf the Leica is more polygonal. The star effect suggests that greater effort to reduce reflections ought to have been made on the Nikkor, but only the potential user can say whether such a performance artifact will be obvious in practice for his type of shooting.
This was the set-up to look at coma, here the Leica Summilux shot at f/1.4

Leica center at f/1.4
Nikkor center at f/1.4 - color fringing
Leica edge at f/1.4
Nikkor edge at f/1.4 - coma and color fringing
Leica center at f/2.8
Nikkor center at f/2.8. Color fringing gone, but star pattern present
Leica edge at f/2.8
Nikkor edge at f/2.8. Note star pattern.
Finally, I shot a blank wall at F/1.4, f/2.8 and f/5.6, and processed the raw files without vignetting correction. Data on the Summilux says that it shows about 2 stops of vignetting; on the R-D1, this seems less because of the field cut, but it is still noticeable at f/1.4 (only). The Nikkor vignetted about the same.
Vignetting at f/1.4, Leica on left
at f/2.8
at f/5.6

My conclusions should be reasonably obvious from the observations above. Both lenses performed admirably and were quite close in their drawing. The Leica Summilux Aspheric was slightly contrastier, and showed slightly better close-up rendering. The Nikkor was marginally sharper, even in the field. Vignetting was similar. Some coma and color fringing is seen with the Nikkor at f/1.4 and internal reflections with point light sources when the aperture is closed down a bit, slightly marring its performance. Both lenses gave smooth, saturated detail with similar bokeh, but the Summilux showed a bit greater saturation and/or micro-contrast.

Given these results, the Nikkor would appear to be the better choice for digital shooting, due to its lower contrast. The caveat is of course that there is no easily-available digital body to use it on, save the purchase of the expensive and rare Orion adapter or one of its clones. It could also be used with adapters on a Nikon digital body, but unfortunately not focusing to infinity. However, the Nikon rangefinder user has a stellar performer here.

The Leica 50/1.4 is a superlative lens. Its cost is high, about $2700.00 currently from B&H. For $400.00 less, however, one can get the Nikkor on closeout, complete with S3 Millennium body. You make the choice!