This is a continuation of a test of the following lenses on the Sony a7RII:
- Zeiss 85mm f/1.8 Batis.
- Zeiss 85mm f/1.4 Otus.
- Leica 90mm f/2 Apo Summicron-M ASPH.
- AF-S Nikkor 85mm f/1.4 G.
- Sony 90mm f/2.8 FE Macro.
The test starts here.
I’ve been testing the lenses for focus shift, and I was unable to get representative captures for the entire aperture sequence for the Batis and the Summicron using the 200 mm rail and a 3.3 m target distance. But now I know that the Batis focus distance moves closer as you stop the lens down and the Summicron focus distance moves further away as you stop the lens down, so I have a better chance of getting good data. So I repeated the testing.
I mounted a Sony a7RII to the Cognisys computer-driven focusing rail. I set the controller up to use 192mm of travel and make 49 exposures 4mm apart. I set the assembly 3.3 meters from the target, mounted the lens, focused a little toward s the back from mid-rail for the Batis and a little toward s the front from mid-rail for the Summicron using manual focusing, and exposed at 49-shot series from f/1.8 through f/8 for the Batis and f/2 through f/8 for the ‘cron. I developed the sets in DCRAW using AHD, and had it convert the files to 16-bit sRGB TIFFs. I told Imatest to calculate the on-axis MTF50 for luminance for a horizontal edge, and plotted the results in cycles per picture height vs distance from the point furthest away from the target.
As the lens is stopped down, the focus point moves closer to it. Peak sharpness is achieved at f.2,8 through f/4. The total focus shift from f/1.8 to f/8 is about 110 mm, or a bit over four inches in 10 feet. Focusing wide open is definitely not the plan here.
You can read off the depth of field in mm if you know what degradation of the peak MTF50 you’re willing to tolerate. Consider the Batis wide open. If MTF50 = 1000 cy/ph is your criterion, at 3.3 meters your DoF is 56mm. Stated in English units, at 10 feet your DoF is 2 inches. Not a lot.
As the lens is stopped down, the focus point moves away from it. Peak sharpness is achieved at f/4. The total focus shift from f/2 to f/8 is about 80mm, or a bit over three inches in 10 feet. Focusing wide open is not the plan here.
Note that, when mounted on a rangefinder camera with no live view, like an M8, M9, or one of the film M-series Leicas, that there is no way to set the rangefinder to deliver maximum sharpness over the range of apertures from wide open to f/8. A compromise might be to set up the rangefinder for best performance in the middle, at f/4, but that would mean that then lens would be delivering an MTF50 of around 600 when rangefinder focused wide open, instead over over 1000, which could be achieved if the lens were indeed focused on the subject at f/2.
If you’re interested in why focus shift happens, you might want to take a look at this post on spherical aberration.
Erik Kaffehr says
I guess that we, who have been advocating live view, have a point?
It seems that focusing stopped down at highest possible magnification is the best alternative?
Well, your tests indicate that focus shift is an important consideration and that we should consider focusing accuracy before spending indiscriminately on the best lenses.
Alternatively, if we own a very good lens, we should consider how to make best use of it?
Your blog is a great learning experience!
David Braddon-Mitchell says
This makes me wonder what to do for critical focus at F 5.6 or 8 in the field. I wonder if AF on a Sony might be more accurate than MF stopped down, given how hard that can be to judge.
Hard to test though because results would only be valid for the tester, would need a lot of trials to average out, and would vary for type of target..
I hope this is not the same on my 21mm manual lens. It’d be nice if there was a “rate” button, so I could take some photos, write down some notes and use ratings to help organise the data.
Do Sony cameras open up the aperture when they focus? I’m concerned they open up the aperture, focus, close the aperture and then lose focus.
With Setting Effect off, yes. With Setting Effect on, it is unclear and seems to depend on the model of the camera and possibly the firmware revision.
Watching my A7R2, it auto-focus’s at max aperture regardless of the effect setting and after achieving focus it then stops down.
The difference is that when the effect is ‘on’ the aperture stays shut after focussing, and also after when the focus button (back or shutter) is released. If the effect is ‘off’ then the aperture stops down again after focusing, but opens when the focus button is released. You can see the effect in live view with low light, when the aperture is stopped down you see the noise in the image.
My concern is if AF is used to do this test, is this a valid test, because all focussing will have been done at maximum aperture. I can only assume that focussing has been done manually?
All the test results I have posted so far have used manual focusing. I thought I made that clear. I will go make and make sure that it’s clear. Thanks.
[Added: OK, now it should be crystal clear.]
Yes its crystal clear 🙂 I congratulate you on your diligence to do all that manual focussing. 49 shots at each aperture.. Thats a lot of focus twiddling.
No, that’s not the way the test works. You wouldn’t be able to see the variations if I refocused for each picture. The camera is focused once at the beginning of the test, and the whole series at that f-stop and all other f-stops for that lens is made with out changing the focus, just the position of the camera.
On the A7 series this shouldn’t be as critical though because except for low-light, or when you’re shooting “settings effect off” the lens is focusing stopped down, correct?
That was clear until the a7RII. It’s not so clear now. If you look at the diaphragm of an E lens while you’re autofocusing, even with settings effects on, it moves sometimes. But so far, all my testing only applies to manual focus.
Anthony Oxlade says
In your conclusions from ‘another medium tele test’ you state that focus shift is not an issue with the A7RII and Setting Effect ‘on’.
I’m a bit confused by this. From what I have read on this page it seems there is evidence that the A7RII operates in AF near wide open. There is also a thread going on at dpreview where the same observations have been reported.
Any chance you could you clarify?