[MadMathMind]

I started this thread separate from the main one to make it easier to find this information.

Who has the lens and what has you experience been with autofocus? I was shocked to read in the indepth review that the site was unable to get two different copies to focus over five different camera bodies. I must be super lucky because I haven't had a shred of an issue. I don't understand how this can happen, where one copy can work perfectly and others not at all--not poorly, but total failure.

I am running mine on a K-5 IIs with firmware v1.05. That's two versions behind. I had no issues so I did not update.

[Digitalis]

The firmware i'm running on my cameras:

Pentax K10D v 1.30
Pentax K7 v 1.10
Pentax K5IIS v 1.07
Pentax K3 1.10

However I don't think this is a software problem, it may very well be a hardware issue - or a combination of both (I know, that is a cheery thought).

[Class A]

I was shocked to read in the indepth review that the site was unable to get two different copies to focus over five different camera bodies.

Before we get shocked or concerned, we should learn about the measures that were taken to remedy the AF issues observed.

Should it turn out that the procedures carried out to determine the correct AF adjustment values were not adequate, the mystery may disappear.

It may also be the case that not all options to resolve the mystery were explored.

BTW, although it would be unfortunate, it is entirely possible that both copies tested had individual faults that should not be construed as a weakness of the lens design. DPReview looked at three different copies of the DA* 55/1.4 and found none that was entirely without decentering issues. A forum user looked at five (5) copies of the DA* 55/1.4 and could not find one that did not focus erratically. Do both these true stories imply that the DA* 55/1.4 lens design prohibits reliable AF and good IQ? Certainly not. They reflect on QC issues at Pentax and potentially other factors, but since there are copies of the DA* 55/1.4 that do not exhibit these issues, it would be adequate to not mark down IQ & AF in a review (as opposed to commenting on the difficulties encountered obtaining a good copy).

[Class A]

The AF issues observed during the in-depth review can only be blamed on the lens, if they do not occur with other, comparable lenses in a similar fashion.

This excerpt from DPReview's review of the DA* 55/1.4 suggest that one should consider the implementation of Pentax AF as a culprit:

"In normal use, we found autofocus consistency to be lower than in our studio tests (few perfect black and white targets exist in nature), with an higher proportion of shots than we'd expect unusable due to mis-focusing. "

I'd be the first to consider that DPReview are not ideally suited to evaluate lenses, but one needs to entertain the thought that perhaps the large size of Pentax AF areas and the implementation of the phase alignment algorithm may cause some surprising results with lenses that have sufficiently shallow DOF and/or are wide enough to let one observe these properties.

I say, innocent until proven guilty.

AFAIC, no proof has been provided that the Sigma 18-35/1.8 has a systematic problem. There is some anecdotal evidence floating around, but in no way I see it fit to convict the lens. It appears everything observed so far, may perhaps be explained differently, pointing to another culprit.

[UncleVanya]

Non-pentax versions of the lens have demonstrated focus problems.

[luftfluss]

I say, innocent until proven guilty.

That's the polar opposite of Internet Justice

I know one Canonite who used the Sigma briefly, complained about AF and returned it. He did not have the USB dock to fine-tune it.

This excerpt from DPReview's review of the DA* 55/1.4 suggest that one should consider the implementation of Pentax AF as a culprit:

That was with the K-7, so that doesn't apply much here.

Maybe its just Sigma's QC is not great. I've had 2 really bad Pentax lenses that I had to exchange for other copies.

[Digitalis]

Maybe its just Sigma's QC is not great.

...it wouldn't be the first time Sigma's QC has been brought into question.

[Class A]

...it wouldn't be the first time Sigma's QC has been brought into question.

Luckily Pentax QC is so great that we'd never see three bad lens copies at Dpreview (that were specifically ordered for reviewing purposes), the 16-50 developing a reputation for decentering, petitions for an option to revive Sudden Death Motor lenses, or the FA 31/1.8 performing like a shard you found in the backyard.

OK, enough with the fun.

At least you often get three or up to six years warranty with Sigma glass, compared to the one year from Pentax.

P.S.: @Digitalis, have you noticed how Roger Cicala from LensRentals has changed his mind about Sigma?

He clearly adopted another viewpoint when Sigma raised their game.

[Digitalis]

At least you often get three or up to six years warranty with Sigma glass

I voided my warranty to repair my new 18-35mm f/1.8 when I spotted a fault with it.

@Digitalis, have you noticed how Roger Cicala from LensRentals has changed his mind about Sigma?

Why should I care what someone else thinks of Sigma? I have my opinion and I stand by it, sigma has a lot of work do do before I change my tune.

[DSims]

Why should I care what someone else thinks of Sigma? I have my opinion and I stand by it, sigma has a lot of work do do before I change my tune.

I just don't like Sigma's rendering that much, so I don't feel satisfied with the photos enough to really like them.

But I was very impressed with the quality of the Sigma 8-16 I had (nice build quality; infinity mark was spot-on (important in an UWA); impressive sharpness). I have to give Sigma credit - I wouldn't be afraid to buy any recent models because of QC concerns.

[robjmitchell]

It doesn't surprise me that this lens has problems with AF. AF systems are good to f2.8, With my 43ltd i find I can only get the AF accurate at mid/long distances or short but not both. Take that issue into a 1.8 zoom and your going to have a problem. Should work perfectly on a K-01 though!

[Class A]

Why should I care what someone else thinks of Sigma?

You certainly don't need to care at all.

Roger Cicala is just someone who sees many lenses a day, observes the faults they develop, repairs them, and gains experience with service centres. I certainly would value the opinion of someone with these kind of privileges higher than that of an individual who makes singular experiences that may or may not have any relevance for anyone else.

If you wanted to paint a picture of Sigma that is based on a large number of cases and is formed with the benefit of being able to compare to Canikon, etc. then it would make sense for you to read what Roger Cicala has to say. Of course, you can just go ahead and speak about your own singular experiences which is totally fine by me, as long as they are not delivered with great authority and claims to generality.

[Digitalis]

Roger Cicala is just someone who sees many lenses a day, observes the faults they develop, repairs them, and gains experience with service centres. I certainly would value the opinion of someone with these kind of privileges higher than that of an individual who makes singular experiences that may or may not have any relevance for anyone else.

Experience counts for a lot, I'm not denying that. However the Global vision lens line up has only been around for a few years and that really isn't enough time to figure out accurate reliability figures for these lenses. I'm cautiously optimistic when buying sigma lenses* But sigma has a long way to go to improve up the image they cane in the photographic community.

I also have a considerable number of lenses, ones that I work with commercially and personally. Some of the off brand lenses I work with are made by sigma, some are made by Tokina and Tamron. I see many of the brand name lenses have fewer problems than their third party counterparts, and in a professional environment I value reliability, optical quality, and a decent repair/maintenance network - cost is of least concern. My customers have to pay a bit more if I have to buy a lens for a project, I have to recoup my expenses somehow.


* FYI - My mother (who is an extremely good photographer herself) she doesn't like sigma lenses at all, she remembers the disasters they came out with in the 70's and 80's. In fact once she told me to never buy a sigma lens ever again.

[JimmyDranox]

I've made a lot of tests with my cameras and lenses, and I've found one rule.

There is no perfect AF system. Regardless of the lens used, or the camera. At least with the cameras I have had, K-r, K-5 and now K-5IIs. Even on the same target, with the same lights, sometime, with no apparent reason, the focus point was missed.

Yes, some lenses can be more prone to errors than others. For example, I know that my Sigma 50mm F1.4 can miss the focus more than other lenses, but also, this lens can focus in such low light where no other lens in my collection can, with the same camera.

So, the advantages that a lens like Sigma 18-35mm F1,8 brings, comes with some drawbacks. From my point of view is normal. Nothing is perfect.

[rawr]

Perhaps related to the Sigma 18-35, my Sigma 35 f1.4 Art also has a problem with focus consistency, esp when shot wide-open, and on the K-3.

It's constantly wanting to tweak and adjust the focus every shot, even on a flat static target, like it's detecting miniscule variations in subject distance or something. So if I shoot 10 shots of exactly the same flat, static subject under exactly the same conditions, 3 out of 10 shots will be out of focus due to the lens refusing to accept that it has found focus. Kind of like a lens that has had too much caffeine - edgy, unsettled and hyper-sensitive. Maybe it's the same syndrome being reported with the 18-35.

Other HSM Sigmas I have (albeit only f2.8 and f4 lenses) just get the focus, seem happy and confident with what they found, and focus the same way time after time. So I'm puzzled about why the Sigma 35 is behaving differently than it's [slower] siblings.