Sigma 135 f1.8 Art; Inconsistent AF

[ScottyP]

My 35 Art workers great from the beginning. I did a simple inclined ruler AF test and put 2-point adjustment but it didn't really need it. I am happy. Maybe with a longer FL a small varianc might seem more relevant?

 

[streestandtheatres]

I have a couple of ART lenses, though not the 135. I have found that FoCal does not work well with them. In both cases I get a bunch of inconsistencies. But in real world they are fine, once the right AFMA is dialled in. That, without FoCal, takes a bit of patience. But once I get the right setting both lenses are super reliable. Without the right setting they both tend to forward or rear focus, and to do so inconsistently.

 

[[email protected]]

I appreciate the distinction between a single point with AFMA vs. a number of distances with a USB dock, but there are two different issues at play here:

• Front- or back-focusing AF can be treated with their dock to good effect, and yes, Sigma gives a more comprehensive tool to dial that in as a function of distance.
  
  
• Inconsistent AF in which the AF does not consistently hit the same target at the same distance is not dock correctable. Again, see my TDP reference -- no dock on the planet will solve that.

The second one above is the bit people are complaining about.
- A
[/quote]

I think you are right to distinguish between those two potential flaws. I wonder, though, if the first could help cause the last. When you have things hinkey with the autofocus, my experience has been that things act much more inconsistently, especially at apertures <2. I speculate that the camera is attempting to determine a binary status (focused adequately or not) on something that looks fudgy due to back- or front-focus.

A couple of my Art lenses - which are rock-solid now for both focus and consistency - required pretty different adjustments at different subject distances. When new out of the box, they'd appear to be hugely inconsistent, which turned out merely to be a factor of the diversity of the distances I was shooting. Not saying this is necessarily the issue with the OP and others chiming in, but hoping that perhaps some of them might find this is a cheap solution.

 

[Eldar]

I appreciate the distinction between a single point with AFMA vs. a number of distances with a USB dock, but there are two different issues at play here:

• Front- or back-focusing AF can be treated with their dock to good effect, and yes, Sigma gives a more comprehensive tool to dial that in as a function of distance.
  
  
• Inconsistent AF in which the AF does not consistently hit the same target at the same distance is not dock correctable. Again, see my TDP reference -- no dock on the planet will solve that.

The second one above is the bit people are complaining about.
- A

I think you are right to distinguish between those two potential flaws. I wonder, though, if the first could help cause the last. When you have things hinkey with the autofocus, my experience has been that things act much more inconsistently, especially at apertures <2. I speculate that the camera is attempting to determine a binary status (focused adequately or not) on something that looks fudgy due to back- or front-focus.

A couple of my Art lenses - which are rock-solid now for both focus and consistency - required pretty different adjustments at different subject distances. When new out of the box, they'd appear to be hugely inconsistent, which turned out merely to be a factor of the diversity of the distances I was shooting. Not saying this is necessarily the issue with the OP and others chiming in, but hoping that perhaps some of them might find this is a cheap solution.
[/quote]
I have not had the chance to return to the dealer, to give them a chance to fix the problem. I doubt that they will though. I have been running both FoCal calibration on multiple cameras (5DSR, 5DIV and 1DX-II) and manual attempts, using a LensAlign rig. I have also tried to do it on different distances from. However, the spread of focus is such that both FoCal and I are unable to determine an AFMA setting. A dock will not change this.

It may well be that I have a poor copy (again), but it behaves exactly as the 35/1.4 and 50/1.4 Art lenses did. So I am not optimistic. However, if they fix it, I will still have a hard time using it for anything critical, because I simply will not trust it. Luckily I have alternatives.

 

[Random Orbits]

I have one Sigma lens, the 20A, and so far I've only used it with the 5DIII. The focus with the center point is OK most of the time but not as accurate as the 24-70L II or the 35L II. My frustration with it comes to using the outer points (I primarily use the left and right banks) on subjects that are close. It consistently front focuses with the outer points but is mostly on in the center. The performance is similar using the left and right outer banks, so it's not a decentering issue. The dock wouldn't help with this. If the subject is farther, then it starts to fall into the DOF and the issue is masked.

 

[SecureGSM]

Eldar,
I have completed calibrating my Sigma 135 F1.8 Art. I used Focal target and Focal in automated mode. camera:6D.
I also took my time to verify that caibration was valid for daylight and LED lighting conditions.
calibrated at 0.9m, 1.3m, 2.5m and 12m (infinitiy). Focal reported focus consistency no less than 98.5% through the test. In order to ensure focus consistency I then ran callibration test at each given distance 4 more times and Focal suggested very similar calibration adjustment values fluctuating around the best AFMA values and within -1/+1 AFMA units around the set value.
Lens is impressively sharp with maximum sharpness achieved at around 10m to target (near infinity) and equal to 2145 Focal units. My sharpest lens is Sigma 85 F1.4 Art with 2185 Focal units achieved at approx. 4m to target.
It is a sharp lens, not as sharp as my copy of Sigma 85 F1.4 Art. They both great lenses but produce very different images of course...