Mind your Focus Shift...

BobYIL

Well-known
Local time
6:03 AM
Joined
Sep 26, 2009
Messages
1,252
7406808548_075f135cb5_b.jpg


Focus shift is an error caused by the residual spherical aberrations of a lens, meaning the shift of focus plane with aperture changes.

It mostly is observed when a lens is focused at its widest aperture at certain point/plane and the picture is shot at the working aperture which is different than the widest one. The photographer believes that he has adjusted the focus correctly however the picture may come out focused on somewhere different than he intended for, therefore with less sharpness.

Focus shift can be detrimental for almost all types of cameras:

- Rangefinders: Makes the user believe that the rangefinder needs calibration; sometimes needing shipped to the service together with lens.

- SLRs, DSLRs: Always the max. aperture is used and then stopped down to working aperture either automatically or manually. User thinks that something could be wrong with his focusing screen or lens (and tries to correct it via microfocus adjustment as on some new cameras, however provides no remedy.)

- Mirrorless: Exactly the same as the DSLRs..

Focus shift is more emphasized with digital cameras than film as the final output is generally larger than the prints obtained with film (for example our computer screens.)

Do not presume that your favorite lenses are focus-shift free.. Not only the former version of the Summilux 35 Asph. or the C-Sonnar are known for focus-shift. Lenses depending on their designs may show some amount of focus-shift; be it small to not mind at all or large to be careful about. The above examples are from the well-known Summicron 50/2 final version; from left to right: f2.0, f2.8 and f5.6; on a heavy tripod, focused at hi-magnification at f2.0 via Liveview and stopped down by preserving the intial focus setting. The final pictures enlarged to 1.75m (near to 6') and the center portion cropped. See how sharpness deteriorates slightly as the lens is stopped down; actually I was seeing on the full view that the actual focus plane has shifted far beneath the original one at f2; i.e. the "other" planes were sharper than even the f2 shot here.. :) (Do not mix it with diffraction; with this lens diffraction starts around f8.)

Do not also presume that the Depth of Focus can take care of focus shift, at larger apertures it does NOT. (See the samples above .)


How to take care of focus shift in critical shots?

1. If you have Liveview than try to focus at working aperture with screen magnification feature.

2. If you do not have Liveview, then try to "learn" the focus shift characteristics of your lens through a simple test: Tape a double-page newspaper on a wall and then from an oblique angle of around 45 degrees (not critical) focus on some letters in the mid portion with your camera on a tripod and the lens at max. aperture. Then take a picture each time by stopping your lens one stop until f5.6. This will help you to find out how much your lens shifts with focus but also -if you have a rangefinder- whether your focusing is working correctly on the spot you want it to focus.


Focus-shift can also constitute one of the reasons of what we complain sometimes about sharpness by insisting that we focused correctly. This was also a reason for me to post this as I have never noticed the mentioned focus-shifts with some of my lenses during the long film years. Digital contributes a lot in such technical aspects.

Regards,

Bob
 
Interesting article. Thank you for sharing. I should really do the mentioned oblique test with my lenses to better know them at wide apertures. I guess closing down the aperture it might trigger some diffraction effects, but the focus shift should be compensated due to wider DoF margins, right?
 
Interesting article. Thank you for sharing. I should really do the mentioned oblique test with my lenses to better know them at wide apertures. I guess closing down the aperture it might trigger some diffraction effects, but the focus shift should be compensated due to wider DoF margins, right?

Diffraction with FF-lenses starts either at f8 or f11. That's why I cut the test at f5.6 which is the optimum aperture of this lens.

The compensation of focus shift of some lenses by overlapping DOF was the case with the film and that's why we either could not notice or did not mind it before. Now it's digital time, and note that the APS-C sensor on the Nex-5N (the test camera here) is equivalent of 36MP on FF! :) Any slightest focus movement I can see on the outcome when I enlarge it over 100cm.

FYI, do not regard anymore the DOF indicator lines on your lenses if your final enlargement is larger than 30x40cm. The classical DOF tables or indicator lines on our lenses have been calculated based on "acceptable circle of confusion" figures of between 0.025 to 0.029mm (i.e. film) which are quite rough for today's digital resolution standards. In case of critical situations, use two stops larger; i.e. if you are to shoot f8 then take f4 indicator lines as reference.
 
Back
Top Bottom