Obi-Wan's Normal Lens Tests
I set out to do resolution tests on all of the normal lenses to which I had access, meaning anything that covered the 50-55mm focal length. The collection includes 14 lenses. 11 owned, 3 borrowed. 11 manual, 3 autofocus. 10 primes, 4 zooms. 5 Pentax, 3 Nikon, 2 Sigma, and 1 each Olympus, Yashica, Tokina, Canon.
For the test, I used an ISO 12233 test chart created by Stephen H. Westin from Cornell University. I printed the PDF on 11x17" paper using a Konica Minolta BuzHub 501 printer. For shooting, I taped the chart to a cardboard box and stood it upright on a card table. The lenses were mounted on a Canon 7D (1.6x crop) body on a tripod. All of these tests were performed in one 3-hour session. The camera was set to ISO 100 and 1/180s. I used a cable release with mirror lock-up, although that probably wasn't necessary since the chart was light by a flash with no ambient room light.
The camera was positioned to fill the frame with the black border around the chart. It's curious to note the varying distances from the sensor plane to the chart, indicating some variation in focal length of these lenses. Also interesting was that when positioned 43" from the chart (the average for the 50mm primes) and zoomed to what the barrel markings said was 50mm, all four zoom lenses gave a noticeably wider field of view. I kept the camera 43" from the chart for thoses tests, and zoomed the lenses to fill the frame as before. The barrel markings generally said I was at 55-60mm.
To make things easy on me, the chart was lit with a single, on-camera Canon 430EX flash in eTTL mode. The brighter reflection near the top center of each chart is a reflection from the flash. The camera's exposure meter doesn't always use the correct exposure on these (generally un-chipped) manual-focus lenses, so the exposure compensation had to be adjusted somewhat during shooting in order to get half way consistent exposures across all images. The amount of compensation is indicated below each image. Of course, exposure (set at +1/2 stop) was dead on for all the autofocus lenses. I debated whether to run an automatic contrast stretch for each of these images in post, but then decided that would eliminate any info about the lens' native contrast. The camera was forced to daylight white balance for consistent interpretation across all tests, so color casts you see in these images are representative of the lens.
All lenses were focused manually on the center square using LiveView at 10x magnification. The lone exception was the newer Sigma 17-70/2.8-4 (last on the list), which generally autofocuses accurately enough with my 7D that I decided to trust it.
I'm sure there are more scientific ways to test lenses, but this is what I had available. I would have liked to use some sort of subject that showed colors with objects of varying distances to demonstrate bokeh, but that must be a test for another day. For resolution tests, I think this is reasonably accurate.
Nikkor-S.C 55mm f/1.2 (1974, 45")
Nikkor-S 50mm f/1.4 NKJ (1969, 43")
Micro-Nikkor 55mm f/2.8 (1990, 47")
Olympus F.Zuiko 50mm f/1.8 (1980, 43")
Asahi/Pentax Super-Takumar 55mm f/1.8 (pre-1971, 47")
Asahi/Pentax Super-Takumar 50mm f/1.4 (1967, 42")
Asahi/Pentax SMC Takumar 50mm f/1.4 (my copy, 1974, 42")
Asahi/Pentax SMC Takumar 50mm f/1.4 (borrowed copy, 42")
Asahi/Pentax SMC Pentax-M 50mm f/2 (borrowed copy, 43")
Yashica ML 50mm f/2 (broken aperture, only shoots wide open, 43")
Tokina AT-X SD 35-200mm f/3.5-4.5, at 50mm (46")
Canon 28-80mm f/3.5-5.6 (1997), manually focused at 50mm (43")
Sigma 17-70mm f/2.8-4.5 (2007), manually focused at 50mm (43")
Sigma 17-70mm f/2.8-4 OS HSM (2011), autofocused at 50mm, with UV filter (43")
Sigma 17-70mm f/2.8-4 OS HSM (2011), autofocused at 50mm, w/o UV filter (43")
I haven't taken the time to analyze most of these images yet to see which lenses are the real winners, but below are a few things I have seen. If you come to any of your own conclusions, please email me. If I agree, I'll publish your findings here.
- The Nikkor-S.C 55/1.2 is somewhat softer than the Nikkor NKJ 50/1.4 and considerably softer than the Super-Tak 50/1.4. The 55/1.2 at f/4 is on par with the Tak at f/1.4.
- The Yashica ML 50/2, my newest acquisition of the bunch, seems to be slightly sharper than the Super-Tak 50/1.4 at f/1.4, but slightly softer than the Tak at f/2. The Zuiko 50/1.8 seems to be slightly sharper yet at f/1.8 than the Tak at f/2. I've long suspected that the Oly was one of my sharpest normal lenses. If I'm able to repair the aperture mechanism of the Yashica, I'll do a full test at every aperture.
Do you like this site?
Help me keep it going by throwing a few pennies my direction.
last updated 3 May 2012 Obi-Wan (firstname.lastname@example.org)