Sigma 35mm f1.4 sporadic AF Poor

I am looking for advice from anyone who may have experienced this sporadic AF issue with the Sigma 35mm 1.4. I got it yesterday and extensively tested the Autofocus and am really disappointed with how it just decides whether it is going to focus where you want it to. I am not paying £600 for a lens that I cannot depend on to nail focus on my 5D MKIII.

I carried out a MA test and dialled in +10 as it was front focussing slightly at close up distances. Once I thought that I had sorted this I got up this morning and immediately set up a test with some of the kids play cups in the garden. The focussing worked at f1.4 for the first 3 cups that were evenly spaced from very close back to approx 2.5 metres. HOWEVER on taking the shot of the one that was focussed some 3.5m away it was way way off. I have attached the image showing the huge back focussing issue at distance. I was using the centre focus point solely for this experiment. I am about to go outside and try the other focus points however yesterday the other focus points were working fine.

Question. Is this was the Sigma Dock is supposed to fix Or is this way off and should I return it to Amazon and try another ??

Any views and guidance greatly appreciated.....

The dock will help mainly at short distances, how did you preform the MA?

+10 seems a lot, what adjustments did your other lenses require?

My S35 required +5 MA compared to other lenses, with the doc it is not 0 MA. plus more linear at different distances. But straight out of the box, it was still better than any other lens I have.

I also fine my 5D3 MA is not really that great. +10 could be less shift than +8 etc...
I recently adjusted my 135L to my 5D3 and 6D, on the 5D I had adjusted ok, but the next day I was mis focusing, so I reverted it back to zero, which now seems to hit focus ok. At +10, which is where I first adjusted it, it would hit focus at some distances but not others.

Performed my MA with camera on tripod at 45 deg to kitchen table using a chart I had downloaded. Haven't needed to do the other lenses although should really try the canon 85mm 1.8 which I am not all that happy with. I think that I may be expecting too much from the Sigma as everyone has raved about it.

Just tried outer autofocus points and they all work well with quick focus. I would say that upto approx 2.8 metres everything is OK. Beyond that not a hope !!!

Yesterday, prior to carrying out the MA I took the lens to the woods to test the AF. I have attached an image that I took straight out of the box with no MA. I was focussing on the flower I positioned on the nib to the right on the tree. The line of autofocus was way way off and too far forward (hope you can clearly see).

Please take a look at the image attached.

Comments ??

You can return it and try another.

The dock can adjust MA over 4 distances, so yes, it could potentially help, but I would just try a new one since the chance of getting a faulty lens is quite low.

I think most owners of this lens requires + MA value, so that's not uncommon and to be expected imo.

Mine is exactly the same, perfect for close to normal distance with +6, as soon as I try focusing something further wide open it back focus.

However I have no problem with that. Why? Because I use the lens mainly for portrait, meaning I always use it under close to normal distance. When I need to focus something far or close to infinity, I almost never needed f1.4 anyway.

So yes, just for the sake of testing, mine will have the same problem, but it didn't bother me at all under real life shooting situation.

kin2son wrote in post #16133947
I think most owners of this lens requires + MA value, so that's not uncommon and to be expected imo.

Mine is exactly the same, perfect for close to normal distance with +6, as soon as I try focusing something further wife open it back focus.

However I have no problem with that. Why? Because I use the lens mainly for portrait, meaning I always use it under close to normal distance. When I need to focus something far or close to infinitely, I almost never needed f1.4 anyway.

So yes, just for the sake of testing, mine will have the same problem, but it didn't bother me at all under real life shooting situation.

Well thanks Kin2son that is really reassuring and kinda the comment that I was seeking from someone who uses one and may have gone through the same as me. To be honest I will only use it for portraits also and rarely at f1.4 so I will test it this weekend on the kids and the family party I am going to and report back. I am only an amateur and do not get paid to take photos hence it takes a while to save the money for these things and one shouldn't have to take so much time testing and tweaking something that costs £600.

A few points:

You indicated you bought from Amazon, so if you aren't comfortable with it, Amazon has the best return process IMO. Just get a 2nd copy. Also, the USB dock does indeed help. There is also nothing wrong with a +10 MA setting, or even +20, that's why the settings are there! What creates problems is when the MA is different at different focus distances and different lighting conditions.

My S35 was pretty good, but after I got the USB dock and installed the firmware, all inconsistency with my AF disappeared and it now does great at all distances and lighting with a MA of +10 on my 5D3.

GuitarDTO wrote in post #16133971
A few points:

You indicated you bought from Amazon, so if you aren't comfortable with it, Amazon has the best return process IMO. Just get a 2nd copy. Also, the USB dock does indeed help. There is also nothing wrong with a +10 MA setting, or even +20, that's why the settings are there! What creates problems is when the MA is different at different focus distances and different lighting conditions.

My S35 was pretty good, but after I got the USB dock and installed the firmware, all inconsistency with my AF disappeared and it now does great at all distances and lighting with a MA of +10 on my 5D3.

The day is getting better already with these great comments of confidence. You know the problem with re ordering another one through Amazon is the worry that the second will be worse than the first so I think that I will order the dock then and give that a try.

Nice one people....better add the Sigma to my signature.....kiss of death before the weekend kid portrait test hey !!!

Wish me luck.

Jasonfire124 wrote in post #16133997
The day is getting better already with these great comments of confidence. You know the problem with re ordering another one through Amazon is the worry that the second will be worse than the first so I think that I will order the dock then and give that a try.

Nice one people....better add the Sigma to my signature.....kiss of death before the weekend kid portrait test hey !!!

Wish me luck.

I bought mine from Amazon as well. If I didn't have the dock I would have 100% sent it back. After getting the dock and adjusting all adjustment zones I can say that I couldn't be happier with the lens. It is dead sharp even wide open.

Did you perform your MA in artificial light? Or in natural light?

Myself and quite a few others noticed that the Sigma is not fond of artificial light. The two copies I used were perfect, near to far, outdoors. No MA needed. Indoors, it would front focus heavily, requiring +10MA. Go outside, everything backfocuses now.

Not sure why... People say it is normal, yet my Canon 35L, 85L II, 85 1.8, 135L, etc... All perform the same inside or out. Returned both and have yet to try again.

Yup sounds like the same ole same ole from Sigma. I have tried probably a dozen sigma primes and all but one had the strange front focus sometomes, back focus other times issues. Since the direction and amount the focus is not constant MA is useless.

I had the same problem with mine. It back focused on longer targets. The dock fixed this. My settings are +3 +3 +4 -10.

Has anyone contacted Sigma directly about this?

It sounds as if the problem is artificial light (indoor) vs natural light (outdoor). With the preponderance of flourescent bulbs these days (CFLs), could that be where the fault occurs? It's not uncommon for CFLs to wreak havoc with exposure, maybe it's causing focus problems, too.

If there is a difference depending upon the type of light, I can see where if you did your testing and adjusted MA indoors under artificial light, then went outdoors (where it was fine before adjustment), it would now be way OOF.

By the way, most MA testing procedures call for the target to be set up 50X the lens focal length distance from the film plane. With a 35mm lens, that would put the target just short of 6 feet away (35 x 50 = 1750 mm. 1750 / 25.4 = 68.898 inches), wouldn't it?

I'm curious about these things because I've been seriously considering this lens.

If it allows fine tuning at different focus distances, it sounds as if the dock might be the solution... though it adds extra cost to owning and using the lens.

Remember focus + recompose with aperture set wide can cause fair bit of the subject you originally focused on to be oof.

amfoto1 wrote in post #16134921
Has anyone contacted Sigma directly about this?

It sounds as if the problem is artificial light (indoor) vs natural light (outdoor). With the preponderance of flourescent bulbs these days (CFLs), could that be where the fault occurs? It's not uncommon for CFLs to wreak havoc with exposure, maybe it's causing focus problems, too.

If there is a difference depending upon the type of light, I can see where if you did your testing and adjusted MA indoors under artificial light, then went outdoors (where it was fine before adjustment), it would now be way OOF.

By the way, most MA testing procedures call for the target to be set up 50X the lens focal length distance from the film plane. With a 35mm lens, that would put the target just short of 6 feet away (35 x 50 = 1750 mm. 1750 / 25.4 = 68.898 inches), wouldn't it?

I'm curious about these things because I've been seriously considering this lens.

If it allows fine tuning at different focus distances, it sounds as if the dock might be the solution... though it adds extra cost to owning and using the lens.

It is not just CFL's though... For example, at Knotts Berry Farm I had the Sig 35 with my 5D3 taking night shots... There was a mix of LED, (Halogen?), Neon and tungsten bulbs illuminating the midways in various parts... 80% of the time, it required +10 to be sharp wide open... Sometimes, 0 worked better (consistently, not just random shot tests), but at one of the shows that I would guess was illuminated mostly with LED or stage lighting (not sure what type that is), it was WAY off at 0...

Yet in daylight, +10 causes massive backfocus... 0 was flawless all day at Disney, Universal and Knotts, not a single problem.

Back at our rental home, at night with artificial lighting... 0 no longer worked... I had to set to +10 to be anywhere near sharp.

Its weird because you would think the camera body would have the big influence on all this... Yet I never had this issue with Canon lenses... If anything, it is very slight to the point I never notice. However, the Sigma was extremely noticeable and really became frustrating to use since I would constantly go from daylight to artificial light within the amusement parks.

Not sure what is going on with the lens itself to cause to behave this way, but I sent my first back for the exact same behavior... 0 perfect from MFD to infinity in normal light but required +10 in artificial light... Got my new one and same thing, 0 MFD to infinity, +10 in artificial.

So I had luck getting good copies of this lens that required no MA at all (in normal conditions), but really went crazy in anything but.

I wish they could figure this issue out though, because that lens is stunning. I would LOVE to own it and not have to hassle with MA depending on where I am shooting. I don't care if it required +10 all the time or whatever, but not to change... I would rather take a softer lens (35L) for reliability.