Does Canon 6D MFA work for 35/1.4 ART?

Recently have become interested in the 35/14 ART lens, mainly based on several terrific samples. Great reviews also but SOME reports of inconsistent AF. There is also a Sigma AF dock, but it's another 60$.

So the question: will MFA work in case there is a front/back focussing issue?

Kind regards,

Bas

Aoi Usagi wrote:

Bas Hamstra wrote:

Recently have become interested in the 35/14 ART lens, mainly based on several terrific samples. Great reviews also but SOME reports of inconsistent AF. There is also a Sigma AF dock, but it's another 60$.

So the question: will MFA work in case there is a front/back focussing issue?

Kind regards,

Bas

AFMA technically works with any lens. However, it may not eliminate all focus inaccuracies with third party lenses. I use a Sigma 30mm ART, and had a slight misfocus issue with a 70D, and AFMA did fix the issue most of the time, but not all. Sigma uses reverse engineering to make their lenses, and use AF algorithms from other Canon lenses, which may cause issues.

Thanks. I am slightly worried about reports of inconsistent focussing with this lens, some say to be totally frustated with it (a minority) and most reviews say it's dead on accurate.

Is there a quick and dirty test that I can do, at the person I buy it from, without big screen computer evaluation? Aim at the middle of a ruler carefully avoiding recompose and enlarge on screen? That's how I would normally do it. Would that reveil any issues? Maybe both near and far?

Bas

Unfortunately Sigma + Canon DSLR can be problematic.

I had a Sigma 17-50mm f2.8 zoom and tried it on both 600D and 70D. No AFMA on 600D, so I assumed that the 70D would be better - it was .... sort of, sometimes.

I spent many hours trying to fine tune it in using AFMA (on 70D) at both wide & tele, and got it to be generally quite good. But sometimes (erratically) the lens would just mis-focus - often at the wide end and f2.8, but not always.

When it worked, it was a nice sharp lens, but I just couldn't rely on it because it cost me a number of irreplaceable shots, so I gave up on Sigma and have stuck with Canon AF lenses. YMMV.

Andy01 wrote:

Unfortunately Sigma + Canon DSLR can be problematic.

I had a Sigma 17-50mm f2.8 zoom and tried it on both 600D and 70D. No AFMA on 600D, so I assumed that the 70D would be better - it was .... sort of, sometimes.

I spent many hours trying to fine tune it in using AFMA (on 70D) at both wide & tele, and got it to be generally quite good. But sometimes (erratically) the lens would just mis-focus - often at the wide end and f2.8, but not always.

When it worked, it was a nice sharp lens, but I just couldn't rely on it because it cost me a number of irreplaceable shots, so I gave up on Sigma and have stuck with Canon AF lenses. YMMV.

Hmm not what I was hoping to hear of course...the sample are SO wonderful! Really on another planet compared to the 50/1.8. But for this type of lens AF is critical and 500 bucks (used price) is not nothing for a risky investment.

(I had the Sigma 18-125 once and until it died after a couple of years it outperformed the kitlens in sharpness and AF)

Bas

My experience with Sigma - prior to going mirrorless - was not inconsistency as such, but rather the need for different focus adjustment at different subject distances. This is exactly what Sigma's AF tuning software and the lens dock were intended to mitigate; in an ideal world you would set distance-relative adjustments via Sigma's own system and leave the in-camera MFA at 0. However neither MFA nor Sigma's own AF adjustments will work if a lens does not focus (or mis-focus) consistently, and I believe the people who reported this very problem.

If I were buying a used Siggie Art for a conventional DSLR I'd probably take 3-5 wide-open photos of the same subject at key distances (close, a few feet away, and far enough away to count as inifinity), refocusing for each shot. If most photos look sharp enough, or are just off focus in the same direction, that's fine. If there's variation between the distance groups but each group appears consistent within itself, then assume you'll need the dock to get things right. If the focus is all over the place with no obvious pattern, well skip that lens for sure.

Hope that helps.

GammyKnee wrote:

My experience with Sigma - prior to going mirrorless - was not inconsistency as such, but rather the need for different focus adjustment at different subject distances. This is exactly what Sigma's AF tuning software and the lens dock were intended to mitigate; in an ideal world you would set distance-relative adjustments via Sigma's own system and leave the in-camera MFA at 0. However neither MFA nor Sigma's own AF adjustments will work if a lens does not focus (or mis-focus) consistently, and I believe the people who reported this very problem.

If I were buying a used Siggie Art for a conventional DSLR I'd probably take 3-5 wide-open photos of the same subject at key distances (close, a few feet away, and far enough away to count as inifinity), refocusing for each shot. If most photos look sharp enough, or are just off focus in the same direction, that's fine. If there's variation between the distance groups but each group appears consistent within itself, then assume you'll need the dock to get things right. If the focus is all over the place with no obvious pattern, well skip that lens for sure.

Hope that helps.

Yes, that's good info, thanks! Though I'm a bit less sure about the purchase than a week ago now

Bas

Not that all lenses have an issue, but my Sigma 35mm Art has issues on both a 5D4 and R5. Consider the Tamron 35mm 1.4, which is excellent also and it works great. I never used it on my 5D4, but it's my most-often used lens on my R5 next to the RF 85mm 1.2.

To answer your question, MFA works with any lens that has AF.

I used to own a 6D for many years and also a Sigma 35/1.4 art lens. The focusing with this combo was always hit and miss. I have the dock but the problem is the weak AF of the 6D. No MFA can fix this. In 2019, I moved to the R and all focusing misses were gone. The AF on the R right on the sensor is so much better that I rarely have a missed focus.

I had the Sigma 50 Art lens and the dock. It’s focus accuracy was unpredictable and could not be compensated for as it kept moving. It drove me insane. I returned it after a week of trying.

I did the tuning methodically with a tripod, focus target, cable release on 2 Canon 5Dmk3 bodies.

It would work for maybe half a dozen shots and then totally miss the exact same static well lit target, each time I tried defocusing first, tuned different ranges, nothing was stable.

All ten of my Canon lenses are mfa tuned to perfection and only need checking every 6 months in case they move a point in either direction.

The skeptic in me now thinks that as a Canon body can detect which lens is mounted and provide a Canon profile, perhaps it can also detect 3rd party glass and throw a random focussing spanner into its focussing algorithm.

It’s the inconsistency that is infuriating, one shot is “ oh that’s amazing” the exact same next shot is…”Why on earth did it do that?.?! nothing changed between shots.

I eventually bought the 35 F1.4L mk2….which is astonishing and my same 5D mk3 bodies focus accurately every time wide open.

Good luck

The skeptic in me now thinks that as a Canon body can detect which lens is mounted and provide a Canon profile, perhaps it can also detect 3rd party glass and throw a random focussing spanner into its focussing algorithm.

Ok I thought about this. And I definetely think you are on to something. Software wise it would be trivial to do. And the way you are describing it has exactly the DNA of a software glitch. And there are LOTS of people reporting the same inconsistency problem.

Canon got away with it, because people attributed this behaviour to clunky reverse engineered AF protocols. And BOTH explanations might be right, suppose Canon has set a sort of randomizer "trap" in their AF protocol and only for proven Canon gear the trap is avoided. You could call this "incomplete reverse engineering" if you didn't know about the trap. Could they actually be sued for dirty tricks like this or are they in their rights?

The journalist in me has awakened: what if some clever Magic Lantern programmers put one and the same sigma 35/1.4 on a 6D body and somehow software-wise FAKE to be a Canon lens, I mean these people are hackers. And suppose problems are completely gone. And with the fake ID removed the problems return. Now THAT would be a big scandal wouldn't it? Thousands of Canon customers completely sabotaged for not being 100.00% loyal.

Kind regards,

Bas Hamstra

rmexpress22 wrote:

Not that all lenses have an issue, but my Sigma 35mm Art has issues on both a 5D4 and R5.

Can I ask what issues you have with the 35 Art and R5? My 50 Art performs beautifully on my R6, and 35 Art may be in my future - if I can count on it to behave as well as the 50.

GammyKnee wrote:

rmexpress22 wrote:

Not that all lenses have an issue, but my Sigma 35mm Art has issues on both a 5D4 and R5.

Can I ask what issues you have with the 35 Art and R5? My 50 Art performs beautifully on my R6, and 35 Art may be in my future - if I can count on it to behave as well as the 50.

It may be luck of the draw. I don't recall if I ever sent it to Sigma, but I have the dock for it with its latest firmware, etc.

The problem with my copy is that if I MFA it using the center AF points it will focus perfectly except with the outer AF points on a 5D4. If I MFA for the outer AF points, then all the other points will be out of focus. It's a very consistent shift so it's not use error. I shot regularly with a Sigma 85mm 1.4, wide open, and didn't have issues on a 6D or 5D4.

On the R5, it's a similar thing except I do the MFA in-lens with the dock. I believe I had the same issue on a M6 II, but that was with one of those focal reducers and while I'm sure that played a part, it's also consistent with the issues on the 5D4 and R5.

I'm not saying all copiers are this bad. But if I recall, the 35mm was one of the earlier Art lenses. I also have a Sigma 30mm 1.4 and although it wasn't impossible like the 35mm Art, it was inconsistent on a 6D. I haven't used it on the 5D4, or R5.