bokesan
Senior Member
In my ongoing quest to educate people who actually seem to buy and use lenses to take photos, instead of just reading reviews - or, even more shocking, draw conclusions about lens quality from actual use instead of test shots, I present a short macro lens quiz containing only charts. No photos here, I promise!
So: Can you tell a lens by its LoCA?
Here are the results of some LoCA and focus shift testing of 9 macro lenses at 1:2.
Here are the lenses compared:
I set the lenses to 1:2 magnification and, using a focusing rail, focus wide open on a backlit razor blade that is not exactly in the plane of focus but tilted a bit. In this test, the razor's left side is in front of the focus plane, and the right side is behind the focus plane (about 1mm at the extreme right). I take a photo at each aperture without refocusing between shots.
For each photo, the edge is divided into segments, and each segment's MTF is computed separately for each raw channel.
The x-axis is the horizontal position. The Z 7 used for this test has a horizontal resolution of 8252 pixels, so the x-axis runs from 0 - 8191.
The y-axis is MTF50 in cycles per pixel. The value should not be taken too seriously - it is lower than it could be because of poor target quality and low contrast. My test setup does not deliver reproducible results between different tests, but only within groups of lenses tested in one session (as is the case here). And the method itself has some problems: it is sagittal MTF only and as the point of best focus moves away from the center, the lens MTF decreases and field curvature could affect the relative position. Also, sample variation does affect the MTF, and I have only one copy of each lens. The important part is the relative x position of the peaks.
Usually (if I focused correctly), the peak of the green channel should be near the center. If the lens has no LoCA, all three peaks should be at the same position. The more LoCA a lens has, the farther apart the peaks are.
To make this a quiz, I'll only show one graph from each lens (the one with the highest peak value in the green channel), labeled A-K. I will reveal which lens is which letter in a few days. Feel free to deduce or guess in the meantime. I'll also reserve the focus shift graphs for later, as they would make this too easy.
Here we go:
Final note: this is supposed to be a little fun diversion for people who can't, for whatever reason, go out and take photos right now. It is not intended as a serious test and you should certainly not base any equipment buying decisions on it. Also, please don't be offended if your favorite lens looks bad here. It might be my sample. Or my incompetence. Or both
So: Can you tell a lens by its LoCA?
Here are the results of some LoCA and focus shift testing of 9 macro lenses at 1:2.
Here are the lenses compared:
- Carl Zeiss Makro-Planar 2/50 T* ZF.2
- Carl Zeiss Macro-Planar 2/100 T* ZF.2
- Micro-Nikkor PC-E 2.8/45
- Micro-Nikkor 2.8/60 AF-S G
- Micro-Nikkor 4/105 AI-S
- Micro-Nikkor 4/200 AF-D
- Sigma 2.8/70 EX DG Macro
- Sigma 2.8/105 EX DG Macro OS HSM
- Cosina 2.5/125 APO Macro (aka Voigtländer APO-Lanthar)
I set the lenses to 1:2 magnification and, using a focusing rail, focus wide open on a backlit razor blade that is not exactly in the plane of focus but tilted a bit. In this test, the razor's left side is in front of the focus plane, and the right side is behind the focus plane (about 1mm at the extreme right). I take a photo at each aperture without refocusing between shots.
For each photo, the edge is divided into segments, and each segment's MTF is computed separately for each raw channel.
The x-axis is the horizontal position. The Z 7 used for this test has a horizontal resolution of 8252 pixels, so the x-axis runs from 0 - 8191.
The y-axis is MTF50 in cycles per pixel. The value should not be taken too seriously - it is lower than it could be because of poor target quality and low contrast. My test setup does not deliver reproducible results between different tests, but only within groups of lenses tested in one session (as is the case here). And the method itself has some problems: it is sagittal MTF only and as the point of best focus moves away from the center, the lens MTF decreases and field curvature could affect the relative position. Also, sample variation does affect the MTF, and I have only one copy of each lens. The important part is the relative x position of the peaks.
Usually (if I focused correctly), the peak of the green channel should be near the center. If the lens has no LoCA, all three peaks should be at the same position. The more LoCA a lens has, the farther apart the peaks are.
To make this a quiz, I'll only show one graph from each lens (the one with the highest peak value in the green channel), labeled A-K. I will reveal which lens is which letter in a few days. Feel free to deduce or guess in the meantime. I'll also reserve the focus shift graphs for later, as they would make this too easy.
Here we go:
Final note: this is supposed to be a little fun diversion for people who can't, for whatever reason, go out and take photos right now. It is not intended as a serious test and you should certainly not base any equipment buying decisions on it. Also, please don't be offended if your favorite lens looks bad here. It might be my sample. Or my incompetence. Or both