The Sonnar-type lens and the Gauss-type lens
The advent of the small interchangeable-lens camera, such as the Leica, sparked fierce competition in terms of performance enhancement between the Sonnar-type lens and the Gauss-type lens, in a struggle for pole position as the standard large-diameter lens. The Sonnar-type lens, which was discussed in Tale 34, can be configured as a three-group, six-element lens, as shown in Figure 1, or as a three-group, seven-element lens, with the rear group comprised of three elements. The basic Gauss-type lens, which was discussed in Tale 2, has a configuration of four groups and six elements, as shown in Figure 2.
Up until the 1940s, the Sonnar-type lens was predominant, due to its simple three-group configuration. The lack of anti-reflective coating technology at that time meant that the more lens boundary surfaces there were that bordered the atmosphere, the lower the transmittance of the lens. In addition, irregular reflection by the outermost surface of the lens led to increased ghosting and reduced contrast. There was the problem that improving the theoretical performance of a lens by increasing the number of elements in its configuration did nothing to prevent a reduction in contrast when images were actually shot. For this reason, the Sonnar-type lens, which made extensive use of cemented lenses and had only six boundary surfaces bordering the atmosphere, enjoyed relative superiority over the Gauss-type lens, which had eight elements.
The Sonnar-type lens still has its merits. First, it is compact. While the total length of a Gauss-type lens is considerably greater than its focal length, due to its so-called symmetrical structure, as an asymmetrical optical system the Sonnar-type lens allows a more compact design. This characteristic certainly contributed to its use in Rangefinder cameras, which were designed to be lightweight and compact.
Another feature of the Sonnar-type lens also deserves mention—its advanced high-order spherical aberration and coma correction capability. The Gauss-type lens enjoyed the advantage that, due to its configuration, spherical aberration and low-order coma are corrected naturally. However, its natural configuration made it poor at high-order coma correction. In contrast, the Sonnar-type lens was able to slickly correct high-order coma employing the principle of "like curing like," in the same manner as the 8.5cm f/1.5 lens for the S Series, as described by Haruo Sato.
You might think from this that the Sonnar-type lens had so much going for it that the Gauss-type lens was a non-starter. This was not the case, however. The Gauss-type lens was better at correcting spherical aberration, astigmatism and chromatic aberration, and it offered better resolution when stopped down, due to the aperture's small focal shift. However, surely a large-diameter lens is supposed to be used at its widest aperture setting, it was thought, and sure enough, with its low boundary surface losses and high contrast at full aperture, the Sonnar-type lens was generally regarded as superior.
This assessment was to gradually change, however, due to the technological revolution of the 1940s and 1950s.
One factor in this was the anti-reflective coating, which was invented in the 1940s. This invention served to eliminate the disadvantage that the Gauss-type lens suffered due to its greater number of atmospheric boundary surfaces. Another factor was the development of high-refraction-index, low-dispersion lantern glass, which began around 1950. This new glass enabled the Gauss-type lens to correct high-order coma. Bolstered by these two developments, from the 1950s onwards the markedly improved Gauss-type lens was to come into its own.
Now we will take a look at the early Nikkor large-diameter lenses. The 5cm f/2, 5cm f/1.5, 5cm f/1.4, 8.5cm f/2, and 8.5cm f/1.5 lenses were all based on the Sonnar-type design. However, although it might seem that Nikon was fully committed to the Sonnar-type lens, the company had in fact been engaged in extensive internal research on the Gauss-type lens from early on. Immediately after the release of the Sonnar-type 5cm f/2 lens, the company began trial manufacture of Gauss-type lenses in parallel with its efforts to improve its existing products (although it did not put these Gauss-type lenses on sale). The company's designers were thus alert to both the limitations of the Sonnar-type lens and the potential of the Gauss-type lens.
It was the Leica-mount 5cm f/2 lens that provided the impetus for a switchover to the Gauss-type lens. As mentioned in Tale 34, the lens' helicoids conveniently allowed the lens displacement to be increased to 45cm. However, this led to complaints that this capability did not lend itself to the reproduction of documents.
The same thing happened with the development of the 8.5cm f/1.5 lens. When Nikon test-manufactured lenses configured with four elements in the rear group, the results were terrible (despite the fact that in theory they should have been excellent) and this configuration was ultimately rejected. This was caused by the fact that the performance of a Sonnar-type lens deteriorates significantly at close range, because it is an asymmetrical optical system.
There were also enormous difficulties with the design of the 5cm f/1.4 lens. Every effort had been made to maintain brightness down to f/1.4, and it seemed that any further improvements would prove impossible. Since this was an optical system that employed the principle of "like curing like," it was thought that an over-zealous attempt to achieve a small improvement could instead end up ruining the lens' performance. It would seem that the subsequent selection of a Gauss-type configuration in the development of the 5cm f/1.1 lens and the use of a Gauss-type design for the final edition of the 5cm f/1.4 lens were to a significant degree informed by these experiences.
The Nikkor-S Auto 5.8cm f/1.4
With the emergence of the single-lens reflex camera, the rivalry between the Sonnar-type lens and the Gauss-type lens was finally resolved. As an asymmetrical optical system, the Sonnar-type lens simply could not ensure the back focus necessary for the operation of a quick-return mirror, and thus had to cede the mantle of standard large-diameter lens to the Gauss-type lens. Thus, as the single-lens reflex camera grew in popularity, the Gauss-type lens became the standard large-diameter lens, while the Sonnar-type lens found its niche as a large-diameter telephoto lens, a role for which its short overall length proved advantageous.
Moving on to the Nikkor-S Auto 5.8cm f/1.4, this was the first f/1.4 lens for use with the Nikon F Series. Since the Nikon F Series was intended to supersede the Rangefinder S Series, it was absolutely imperative that Nikon develop a standard lens with good brightness characteristics at an aperture setting of f/1.4.
In charge of the design of this lens was Saburo Murakami. It seems likely that this was the last lens that he ever worked on, since after its development he was promoted to assistant director of the design division.
The design of this lens began at around the time the design of the 5cm f/2 was being completed. The 5cm f/2 was a modified Gauss-type lens that featured a weak concave lens at the front, and which was able to ensure back focus as required by F-Series cameras. However, it was no simple matter to make the transition to a lens that was one stop brighter. In general, the brighter the lens, the thicker it is. Thus its overall length is increased and the back focus is shortened. In the end the designers decided to make the focal length 8mm longer, so as to ensure the required back focus.
Figure 3 shows this lens in a cross-sectional view, with its six-group, seven-element configuration. This configuration is equivalent to a regular Gauss-type lens with a convex lens added at the front.
Although the third and fourth lenses appear to be cemented together, they are in fact slightly separated from one another, with a tiny air space in between them.
The focal length of 5.8cm represented a trade-off between the decline in performance caused by ensuring the necessary back focus and the need to set a focal length that still fell within the range that would allow the lens to be termed a standard lens. For a short telephoto lens, however, this was a fairly workable focal length. The Noct-Nikkor lens, which was released later with a focal length of 58mm, emulated the 5.8cm f/1.4 in this regard.
In terms of the balance between aberrations, this lens was typical of the early Gauss-type lenses. The addition of a convex lens to the front group was superbly effective in correcting spherical aberration. Longitudinal (axial) chromatic aberration and lateral chromatic aberration were both minimized and the astigmatism correction was excellent. However, since field curvature persisted in small measure and the coma correction was imperfect, there was considerable flaring at full aperture, and when the lens was stopped down, its resolution was somewhat inferior to that of the latest lenses. These two problems were probably the result of excessive efforts at the design stage to increase back focus, as required by a single-lens reflex camera. Flawless field curvature correction and coma correction were subsequently achieved on the lens that succeeded this model—the Nikkor-S Auto 50mm f/1.4. However, I would like to save that story for another day.
Although barrel distortion persisted to a small extent, this was not really noticeable unless linear subjects were shot.
Results using the lens
Exposure mode: Aperture-priority auto,
Exposure mode: Aperture-priority auto,
Finally, we will take a look at the photographic results that this lens produces.
At full aperture the center of the image produced by this lens is extremely sharp; however there is blurring from the center of the picture outwards due to coma flare, and (except in the very center of the picture) there is an overall gauzy effect. The coma flare decreases as the lens is stopped down, and generally disappears at f/2.8- f/4. Due to the slight residual field curvature, however, the image never acquires a solid tone but retains a soft appearance overall. When the lens is stopped down to f/4, the images produced are in marked contrast to images shot using the latest AF Nikkor 50mm f/1.4, which exhibit a uniformly solid appearance that extends to the peripheral areas of the image.
The first sample image is a photograph of an azalea, taken with the lens slightly stopped down. The soft appearance produced using this lens is probably shown to best effect when an image is shot at a range of less than 1 meter. At full aperture, a subject in front or behind the paper-thin focal plane will be blurred and will merge into the background. This is an effect that is probably difficult to achieve with any other lens. To avoid excessive blurring, this sample photograph was shot one stop down.
The second sample image was shot with the lens stopped down slightly more, at f/4. Because the focal length of this lens is slightly greater than focal lengths of ordinary standard lenses, at any given stop setting there is considerable blurring with little perspective distortion. Hence the lens is well suited to mid-to-long range photography.
As can be seen from the samples, a fine blurring effect is achieved. One unfortunate aspect of the lens, however, is that since the aperture is hexagonal, stopping down from f/2 to f/2.8 produces shadowing at the corners of the image.
First released in March 1960, this lens was sold until the release of its successor, the Nikkor-S Auto 50mm f/1.4, in March 1962. Once the lens had outlived its usefulness, production ceased. The lens was thus in production for a mere two years. It was its successor, the Auto 50mm f/1.4, that firmly established the position of pre-eminence that the Nikon F Series and Nikkor lenses now enjoy. However, the tale of the development of the Auto 50mm f/1.4 is one that we shall leave for another time.