Anybody use the AF micro-adjust feature?

[GeorgeHudetz]

It's something I noticed when I was browsing the S1RII manual. Seems like all the bodies with phase detect AF all have it. Apparently, it's for AFC, as they tell you to use AFC when dialing in a lens.

I assume it's intended to make provisions for a non-Panasonic lens to focus more accurately when using AFC. Or perhaps any lens that isn't spot on. I know the old DSLR bodies had something similar, but I think it was for both AFC and AFS in those days.

Anyway, anybody try it?

 

[Pete_W]

Hi George,

I didn't know this existed. Just looked it up in the S5M2 manual and found the page below. Is the S1RM2 the same?

 

[GeorgeHudetz]

Yes, it's the same. They don't say exactly when you should use it, but I suppose if you point the camera at something not moving (subject detection off) and then grab 6 frames or so in one of the drives modes, and most are out of focus, you have a problem? Or maybe AFC but in single-shot mode? Not sure. I'll probably experiment on the 100-400. See if I can make it worse, LOL.

 

[CharlesH]

I know the old DSLR bodies had something similar, but I think it was for both AFC and AFS in those days.

Canon DSLRs had it. It would drive me crazy trying to calibrate long Canon lenses. You are trying to compensate for minor differences between the focal plane distance and the autofocus sensor distance, and in my experience this adjustment was required for telephoto lenses. This was across several Canon cameras I owned. I often resorted to focusing with 'Live View', which is much the same as focusing with a mirrorless camera. This is one reason I eventually switched to mirrorless cameras, where the focus is on the main sensor and the AF micro adjustment isn't needed. I know Lumix cameras have this function, but I have never seen a reason to use it, and I don't think it has the same function as with DSLR cameras. I'm looking forward to what you find.

 

[Andreas]

It's something I noticed when I was browsing the S1RII manual. Seems like all the bodies with phase detect AF all have it. Apparently, it's for AFC, as they tell you to use AFC when dialing in a lens.

I assume it's intended to make provisions for a non-Panasonic lens to focus more accurately when using AFC. Or perhaps any lens that isn't spot on. I know the old DSLR bodies had something similar, but I think it was for both AFC and AFS in those days.

Anyway, anybody try it?

As I understand things, this is an adjustment to the PDAF system which is only used for AFC. CDAF is still used with AFS, and so adjustments to the PDAF system will have no effect on AFS. With DSLRs, phase detect AF is used in all modes, so the adjustments will affect both.

I found micro-adjustments to be extremely valuable with my Pentax DSLRs. I haven't been tempted to try it with my Sigma 100-400 on my S1Rii because the AF at 400mm--where I tested it extensively--seems spot on. This seems true both in terms of the resulting images (when focus bracketing manually with tiny adjustments), and when testing it in the viewfinder with fully magnified manual focus on a tripod. The AF choice always seems the sharpest. It seems to nail the focus in both AFC and AFS, and the AFC and AFS focus results are consistent.

 

[Babylonia]

AF micro-adjust feature, used by a mirror-less camera still can be handy when taking pictures of models.
As an example, AF taking the focus point to "eye-lashes", versus the iris of the eye.

Or AF using full wide opening of a lens, but during the exposure the aperture is closed to the "working" aperture setting.
Lenses can have "focus" shift defects, by spherical aberration (optic imperfectness) of lenses.
-

 

[Andreas]

I found an interesting study conducted by Reikan Technology Ltd, the UK company that makes FoCal (a program that assists in the calibration of PDAF focus systems). The study addresses the issue as to whether PDAF calibration has benefits for mirrorless cameras, with their on-sensor PDAF systems. DSLRs use a separate phase-detect autofocus sensor--as the mirror blocks the image sensor--so the potential for miscalibration should be larger for DSLRs. This is essentially what they find. Mirrorless systems with on-sensor PDAF are far less likely to require adjustment than DSLRs, but even with mirrorless systems there is a significant number of cases where a camera-lens combination will benefit from calibration. They tested the Nikon Z system, and compared it to Nikon F DSLRs. Here is a brief overview of the results:

DSLRs (F-mount): Just over 96% of tested lenses benefit from autofocus calibration, and the vast majority of these require far more than one step of calibration. This is consistent with my experiences with Pentax DSLRs, where I calibrated everything!

Mirrorless (Z-mount): most lenses did NOT benefit noticably from calibration, although a substantial minority required significant calibration for best performance. Here's the breakdown:
- just over 30% required ZERO units of adjustment
- roughly 67% required ONE unit of adjustment or less "which is not going to make a significant visual difference"
- roughly 10% required at least four units of adjustment "which would definitely produce a visible difference to image sharpness"

They also found that F-mount AF lenses adapted to Z-mount are more likely to receive significant visual benefit from calibration (25%) than native Z-mount lenses (10%). This may explain why Sony makes calibration available for adapted EF lenses, but not for native E-mount lenses. By the way, Canon also does not make calibration available for their mirrorless lenses.

Here's a link to their results:

Mirrorless AF Calibration - Part 6: The Evidence

Do you need to calibrate mirrorless cameras? Here's the evidence from FoCal IQ that you really should, and an explanation of why.

blog.reikanfocal.com

Given that they sell calibration software they may be a bit biased, but I found their results interesting and largely believable. It's good that Panasonic has made calibration available to users, unlike Sony and Canon. It might not matter with a majority of lenses, but it's great to have this capability for those cases where it does matter.

 

[GeorgeHudetz]

^great post; thank you. It makes total sense that adapted lenses would be more likely to need microadjustment; the mechanical tolerances of the adapters alone could easily cause variation.

I did a quick test of my Sigma 150-600 and didn't see any noticeable problems with AFC when compared to a single AFS shot. I'll slowly check other lenses to see if any require adjustment & report back if I find anything. It's good to know the capability is there; kudos to Panasonic for making it available.

 

[CharlesH]

It's good that Panasonic has made calibration available to users, unlike Sony and Canon.

Very interesting Andreas, thanks. I'm going to try micro-adjustments for my 70-300, which others say is very sharp, but mine is not.

 

[Babylonia]

It makes total sense that adapted lenses would be more likely to need microadjustment; the mechanical tolerances of the adapters alone could easily cause variation.

No, it is not about variation of adapters, that micro adjustment is necessary for mirror-less camera's.
As AF always do have some extra "over" distance spare beyond infinity to overcome variation of adapters / and "system".
AF is done by pixels of the image sensor itself.

It is that older DSLR camera lenses itself are not corrected as well as today new corrected lenses in general intend specially for mirror-less camera's.
As conditions and tolerances in general by the far more high resolution sensors of today, you do need comparable far more high optically quality corrected lenses.
Still there can be more "tiny" optical tolerances. Having some focus shift, aperture at "wide open" AF, versus closing down to "work-aperture" value for exposure.

Can be different to, which AF-mode is used, and which high demand repeating shutter exposures are needed for e.g. sport imaging.
Mechanical, versus electronic shutter. The approach of one camera brand versus other camera brand. So several options can be involved.
Where aperture openings are "closed" to the set aperture value itself, also during AF.
Only by this last condition a calibration is not needed.
As it is the only optical condition, where the condition for AF is identical to the taking of the image itself.

 

[CharlesH]

Very interesting Andreas, thanks. I'm going to try micro-adjustments for my 70-300, which others say is very sharp, but mine is not.

About my 70-300, which I have complained about for poor sharpness. I set it up today to try AF micro-adjustments at 300mm to see if I could improve it. The answer is NO. I saw a very slight effect at +10, but still a very non-sharp lens.

Then I used the Filter capability of Lightroom and went back and looked at all the previous photos with that lens, with both the S5II and S5IIx. SURPRISE. The lens was very sharp when it was new, in April 2023. But on a trip to Canada late that year it had a sharpness issue. That was a pretty rugged trip with lots of travel and many flight segments. My assumption is I somehow damaged the lens on the trip. At the time I think I wrote off the bad pictures with that lens as my errors. Looking now, it was the lens.

And the S1RM2 with more pixels makes this condition worse. I will send the 70-300 off for repair. I apologize for calling out the 70-300 as a non-sharp lens.

 

[Pete_W]

Then I used the Filter capability of Lightroom and went back and looked at all the previous photos with that lens, with both the S5II and S5IIx. SURPRISE. The lens was very sharp when it was new, in April 2023. But on a trip to Canada late that year it had a sharpness issue. That was a pretty rugged trip with lots of travel and many flight segments. My assumption is I somehow damaged the lens on the trip. At the time I think I wrote off the bad pictures with that lens as my errors. Looking now, it was the lens.

Good bit of detective work there Charles!

 

[RuleOfThirds]

In the DSLR it was also much more tricky because the af sensors are above the mirror, so also tiny adjustments to the mirror would be necessary. My Nikon D7000 had that calibration system also. It is quite tricky because you had to focus straight on a ruler which was positioned in a 45 degree angle. Not all focus points at that time where of cross type needed for f2.8 or smaller.

With mirrorless the af detectors are on the image sensor.So no mirror involved.

The horror… after an accident with my right hand (multiple shattered finger bones, not a simple fracture) I was unable to handle the weight of the Nikon anymore and had to go to a lighter system and therefore switched to m43. I didn’t miss all that AF trial and error. Never made so much test shots for AF as that dslr period.

After ten years I was back being able to use it. First went Fuji X-S10, then heavier lenses and X-T4, and after that went well I finally could go back to the kind of weight I had with Nikon. Although my current system might still be a bit lighter.