找了一篇原文 当时出那个ASPHERICAL S.S.C. 是为了对抗NIKON NOCT 58MM F1.2 和 徕卡 NOCTILUX 50/1.0的
CANON FD Lens Reports By Erwin Puts
50 and 55mm lenses
Introduction
Since the introduction of the Canon F1 in 1971/72 the Canon optical designers have created a number of lenses that challenged the conventional wisdom of lens design. And in cameras too. they were innovative: the Canon Pellix may not be a great commercial success, but it showd the forward thinking of the Canon engineers to see the limitations of a design and find solutions.
In optical design they too were very capable: they were very early with aspherical surfaces, floating elements, a new correction philosophy for aberrations (see my Leica R publication on the Leica site), the breachlock bayonet system, and many others.
For this report I tested two Canon 1.8/50mm lenses (FL and FD). two 1.4/50mm lenses (FD and new FD) and the redoubtable 1.2/55 aspherical. To indicate the relative performance (with respect to each other I will give some resolution figures, measured on the optical bench). Resolution figures are dangerous, as you may read too much into them. But it gives some much needed quantified information about these lenses.
Canon 1.8/50mm FL and FD
Both are classical 6 element Double Gauss designs. Both show strong curvature of field, as can be seen from the figures. If you focus on the center of the image, the outer zones are quite soft and if you focus on the outer zones the center becomes soft. This is a classical dilemma for camera designers. Most lenses exhibit this form of aberration. A lens does project a curved image on a flat receiver (the film). So the designer can select a back focus distance where the tip of the curve intersects the front of the film plane: we have excellent center sharpness. Or the designer can select a distance where the outer parts of the curve intersect the film plane, giving a high quality in the outer zones with a weaker center part. You can count on film curvature to correct the center part, but in any case the designer has a choice here. For a normal standard lens it is important to have even coverage, so the second solution would be wise. for a high speed reportage lens, the first option may be advisable. The official approach in lens testing is to focus on the center part and use this position as an analyis base. Here it comes:
At 1.8 and on axis both lenses have a central disk of about 6mm radius (12mm diameter) of high resolution and low contrast. These lenses resolve easily 125 linepairs/mm but the contrast is very low and some flare can be seen on axis.
In the outer zones the resolution drops to about 20 lp/mm with a higher contrast, but now we see soft edges at the black/white borders. The lvel of astigmatism is very low, which is a major feat. There is low vignetting and some very small pincushion distortion.
Stopping down to f/2 and f/2.8 brings a slight improvement in contrast and it is at f/4 that the lenses start to show a punchy performance. At f8 we have excellent performance with high contrast, and good edge contrast (micro contrast) over the whole image field. The FD version shows a slight improvement in the outer zones, but overall both versions perform identical.There is no decentring: always a good sign of outstanding workmanship in manufacturing and quality control.
As the FL version is steel and glass, where the FD version already uses plastics, my preference would be for the FL version, but generally both are fine performers. There is some tendency to flare and overall contrast is low, giving the pictures a flat and dull appearance when uses wide open. Stopped down the perfomance is very commendable.
Canon FD 1.4/50 SSC and New FD 1.4/50mm
Both are seven element Double Gauss designs. The fingerprint is similar to the 1.8 designs. No astigmatism, strong curvature of field, some slight pincushion distortion, low level of vignetting (causing the corners to be very weak in performance).
Wide open at 1.4 the SSC version has excellent quality in the center area (6mm central disc), but again contrast is low and resolution high with more than 125 lp/mm. Going to the corners there is a rapid drop to about 30 lp/mm, where the lines have soft edges, giving the pattern a blurred appearance. At f/2 edge contrast improvs and at f/4 we have good overall contrast. Usable resolution in the outer ones has reached 40 lp/mm and at f/8 we have excellent performance with 125lp/mm in the central disc and 77 lp/mm over the rest of the image field, excepting the corners, where we see at best 30 lp/m.
Flare is strong on axis at wider apertures.
The new FD version of the 1.4/50 lens shows the same performance with a slight improvement in the outer zones and overall some higher micro contrast and a more even performance over the whole image area.
The new FD is shorter, more compact and has less weight: to have the same performance as the larger predecessor is qute good and overall the performance is better than most films could deliver in those days. Both are not stellar performers, even if Canon indicates these to be their main lenses as referece for the rest. Mechanical quality is very high and they are better than the 1.8 versions. Here I would choose the new FD version.
The Canon FD 1.2/55mm SSC Aspherical
Here we entering a new game. With this lens Canon used every trick they knew. The lens was in its day (1975) the most expensive standard lens in the world, even more expensive than the Noctilux 1.2/50 from Leitz. There were 8 lens elements, one aspherical surface and a kind of floating construction where the last lens stayed in postion and the rest seven elements moved as agroup when focusing.
At 1.2 the central disc shows a clear but low contrast image where 154 lp/mm can be easily sen. In the outer zones we see the performance drop from 77 to 50 lp/mm and even the corners have 50 lp/mm. Amazing as it seems to be, but this lens is wide open better than the other designs at f/4 and smaller. The snag is that micro contrast is quite low and even the 10 lp/mm pattern has soft edges. In practical shooting you see this as a quite soft image when looking at finer details. You see the detail in the mage, but it lacks the sharp delineation of outlines that give the picture transparancy and sparkle. Coma is very low (only in the outer zones) and now we see strong astigmatism and less curvature of field. Again there are no decentring problems and this lens must be at the top of mechanical construction. Even close up pictures (at 50 to 70cm) are quite crisp wide open.
Vignetting is quite low and so is flare. But straight into the sun, we see some secondary reflections, which is unavoidable, but even inside the flare patches we still see details of the image, a sign of very good reduction of the internal flare at the micro level.
At f/2 we see some crispening of the edges of detail and outlines and at f/2.8 we have a very high quality. There is some softness over the whole image, but that is gone at f/4 where we have an overall performance of 80 lp/mm over the whole image field (center is still at 154 lp/mm) with good contrast. From f/4 to f/11 we have the same level of performance.
Compared to the Noctilux 1.0/50m that was designed in the same period, we have to say that he decision of Mandler to drop the aspherical surfaces and go fore a straight design with all spherical surfaces was not very smart. The Canon lens outperforms the Noctilux at wider apertures and closer distances. Accepting that the Noct has a half stop more power in light gathering we still have to note that that mere half stop asks for some compromises.
And compared to the Noctilux 1.2/50m with two aspherical surfaces, we would say that again the Canon is the better lens overall. the exception might be the full aperture where the Noct has some advantages in transparancy, but that is not much.
This Canon lens is an amazingly effective lens and a superb design in itself. Of course it lacks the sparkle and transparancy in small details that we can see in better designs, but they are of lower maximum aperture. Some commentators called the Canon FD 1.2 aspherical the best standard lens in the world. It is hard to disagree.
Please note that I am happy to share my testreports and all the information contained in them to all members of the Internet. This report however is the result of considerable research and must be considered as a document protected by copyright law. So if you cite from, or use information contained in these documents, please respect the notion of intellectual property and refer to the original document or ask permission to copy parts of the text.
Erwin Puts
Author: Erwin Puts email: <[email protected]> |
楼主: 彧趣
楼主
