Review: Sigma 50mm f/1.4 DG HSM Art Lens

[Artifex]

"Interesting is that, according to the owner's manual, "You can [snip] adjust the amount of focus ring rotation to operate Full-time MF function." via the dock.".
I'm not sure I understand well this sentence (english is not my native language). Does this means that it would be possible to have, for instance, a 270° focus throw instead of 90° in manual focus? Help would be appreciated, thanks!

 

[traveller]

This is a real bummer, as I rate AF consistency as one of the most important aspects of this type of lens. I've found that for the most part, missing focus is far more destructive to resolution than a few hundred line widths per pixel height difference in MTF50. Of course, this new Sigma 50mm and the Zeiss Otus are exceptional in that they are clearly far superior at large apertures to the conventional double-gauss designs.

The situation might not be so bad except that my camera's focusing screen (5D3) is pretty useless at showing depth of field at large apertures, so it's easy not notice the fact that the focus is off; this also makes accurate manual focus almost impossible below f/2.8.

 

[sagittariansrock]

The full text of the review indicates a 40% AF miss rate in formal testing, and includes statements like, "...the longer I focus tested this lens, the less sure I was about its focus accuracy," and, "Sometimes, most images are properly focused and when my shots counted, this lens delivered. But sometimes, more images are out of focus than I am comfortable with." To me, that does not equate to, "...occasional AF inconsistency." Which of those statements made it into the concluding paragraph of the review, which is the part most likely to be picked up and quoted, as it was in this post by CRguy?

+1.
The observations in the body of the review seem to have been toned down a lot in the summary.
AF inconsistency is a biggie. I will hope that it is correctable via firmware.

By the way, one advantage of reviewing pre-production or early production samples is that Sigma gets the opportunity to fix things before shipping out mass orders to the customers. Probably the only advantage IMO.

Let's hope for the best and keep fingers crossed.

 

[sagittariansrock]

"Interesting is that, according to the owner's manual, "You can [snip] adjust the amount of focus ring rotation to operate Full-time MF function." via the dock.".
I'm not sure I understand well this sentence (english is not my native language). Does this means that it would be possible to have, for instance, a 270° focus throw instead of 90° in manual focus? Help would be appreciated, thanks!

Ok, I think I get it now.
Here's what it says in the manual: "It is possible to adjust the timing to operate Full-time MF function." And it provides the following image.
I think it means you can change the amount by which the focusing ring has to be turned to trigger the MF override. So, let's say your fingers accidentally touch the focusing ring, that will not change the focus.

 

[Maximilian]

Allthough I was and still am interested in buying a decent - or lets call it high value for money - 50 mm lens right now I must say, I am lucky not beeing in any hurry. Because this review made me patient.

first:

Of those 10 shots, 4 are sufficiently OOF as to be unusable (3, 4, 6, 10). A 60% hit rate with a static subject and a tripod-mounted camera, particularly one with an excellent AF system, does not inspire confidence.

This was, what I was most afraid of: AF-issues with this 3rd party lens. So let's see, what comes with retail product tests.

second:

Also, this is a departure from the norm for Bryan's lens tests (and one, frankly, with which I'm not too pleased):
[quote author=Bryan @ TDP]
My evaluation lens was a short term loan from Sigma, as they offered the production-grade lens before it was commercially available.

Any time a manufacturer supplies a product to a well-known reviewer, a big unanswered question is whether the provided copy is truly representative of units purchased retail. ...
[/quote]
neuroanatomist explained it very well, what might be the difference between selected test lenses and retail products.
But also a really big "plus" to Bryan Carnathan that he was telling this important fact so plainly.

So I am again lucky not beeing in any hurry and maybe see if ther is and what might be the answer comming from Canon.

 

[Viggo]

But according to a very lengthy thread about the 35 Art I was specifically told there are no AF issues with the GV Sigmas :

First off I don't think it's as simple as, oh I'll wait two weeks for Sigma to fix the AF with a firmware, they have already tested it on the 5d3 and 1dx extensively, I guarantee that, and as it is now, it's the best they can do with the option they have, reverse engineering. It's of course very sad and to a lot of people, me included, it's the biggest deal breaker of them all.

And if you buy an MF lens you know you have to turn the ring in order for focus to be achieved, but seeing time and time again your AF lens not delivering is the worst thing I can imagine (in photography).

 

[sagittariansrock]

Is there a lens out there that never has occasional AF inconsistency?

Does a 40% miss rate really constitute occasional inconsistency? I think not...

Compare this to Zeiss missing 100% of AF shots

This comparison is as pointless as comparing a Audi with (automatic) transmission problems to a stick-shift Ferrari.

 

[neuroanatomist]

Is there a lens out there that never has occasional AF inconsistency?

Does a 40% miss rate really constitute occasional inconsistency? I think not...

Compare this to Zeiss missing 100% of AF shots

This comparison is as pointless as comparing a Audi with (automatic) transmission problems to a stick-shift Ferrari.

I don't think Audi makes cars that 'occasionally' mis-shift, causing the car to lurch and hot coffee to spill all over the driver. Just sayin'…

 

[thepancakeman]

Is there a lens out there that never has occasional AF inconsistency?

Does a 40% miss rate really constitute occasional inconsistency? I think not...

Compare this to Zeiss missing 100% of AF shots

This comparison is as pointless as comparing a Audi with (automatic) transmission problems to a stick-shift Ferrari.

Only if the Audi has comparable performance to the Ferrari and works fine in manual mode.

 

[mackguyver]

Is there a lens out there that never has occasional AF inconsistency?

Does a 40% miss rate really constitute occasional inconsistency? I think not...

Compare this to Zeiss missing 100% of AF shots

This comparison is as pointless as comparing a Audi with (automatic) transmission problems to a stick-shift Ferrari.

Only if the Audi has comparable performance to the Ferrari and works fine in manual mode.

I think the Zeiss is actually closest to a Tesla single-speed, high-torque transmission, but powered by human electricity . The Sigma would be like a new driver missing the occasional shift :-[. And the 50L, like a sweet dual-clutch automatic 8).

 

[YuengLinger]

Just waiting for the apologists to scold those of us who like reliable AF, or tell us we aren't true photographers.

A whole bunch of photographers paid a whole bunch more money for a 5D3 than for a 6D precisely because they wanted better AF. Unless this latest review is an aberration, or Sigma addresses AF problems without forcing us to buy a USB lens tweaker thingy, I can live with my ef 50mm 1.4 for now.

All that said, I'm surprised nobody has been discussing how odd it seems that AI Servo was apparently ok, but One Shot on a tripod wasn't...Any conjecture?


(Btw, have the 35 A, like it very much, and have had only one fairly dim-light AF struggling event, one that couldn't be reproduced. Like it came and went.)

 

[sagittariansrock]

Is there a lens out there that never has occasional AF inconsistency?

Does a 40% miss rate really constitute occasional inconsistency? I think not...

Compare this to Zeiss missing 100% of AF shots

This comparison is as pointless as comparing a Audi with (automatic) transmission problems to a stick-shift Ferrari.

I don't think Audi makes cars that 'occasionally' mis-shift, causing the car to lurch and hot coffee to spill all over the driver. Just sayin'…

I had a hard time finding a car manufacturer that makes a faulty automatic transmission in an otherwise excellent car. I don't think there are any, actually.
Hence the hypothetical example.

Pancakeman, I am sure you won't find much difference in a regular city commute between a Ferrari and an Audi, and while the Sigma works 'fine' I think the Otus can boast a much better manual focusing mechanism (throw, smoothness, accuracy of distance scale, etc.).

 

[mackguyver]

Just waiting for the apologists to scold those of us who like reliable AF, or tell us we aren't true photographers.

A whole bunch of photographers paid a whole bunch more money for a 5D3 than for a 6D precisely because they wanted better AF. Unless this latest review is an aberration, or Sigma addresses AF problems without forcing us to buy a USB lens tweaker thingy, I can live with my ef 50mm 1.4 for now.

Btw, have the 35 A, and have had only one fairly dim-light AF struggling event, one that couldn't be reproduced. Like it came and went.

All that said, I'm surprised nobody has been discussing how odd it seems that AI Servo was apparently ok, but One Shot on a tripod wasn't...Any conjecture?

No scolding, but it's likely because Canon doesn't license their AF algorithms and Sigma (and the others) have to reverse-engineer them. Also, there are 2 other reviews (Phoblograper & LensTip) that have mentioned this same issue and given that one in in Poland, that has to be from at least 2 different lenses. As to why this would be different than the 35 Art, I'm not sure. There is a little less DOF, but beyond that. they should be very similar.

 

[sagittariansrock]

All that said, I'm surprised nobody has been discussing how odd it seems that AI Servo was apparently ok, but One Shot on a tripod wasn't...Any conjecture?

I did comment on that and hypothesized it is an incompatibility with Canon's algorithm- therefore not necessarily Sigma's fault, just the trade-off with third party lenses, as Mackguyver mentions.

BTW, Mackguyver, I don't agree with your analogy. Zeiss is highly dependent on human factor (although the long throw is forgiving), while the Tesla is not (talking about acceleration, not going off the road). The Sigma AF is not controllable by the photographer, while the new driver can get better at shifting.
I have no experience with either the 50L or with a sweet dual-clutch (which isn't technically 'automatic' by the way). My car has one of the worst dual-clutch transmissions ever made in an otherwise excellent car (I suppose I should have compared that to the Sigma...).

 

[YuengLinger]

sagittarriansrock said, "I did comment on that and hypothesized it is an incompatibility with Canon's algorithm- therefore not necessarily Sigma's fault, just the trade-off with third party lenses, as Mackguyver mentions."


Missed that--But why would you say it isn't Sigma's fault? Who is making a lens first released in a CANON mount? If they CAN'T get it right, why put it on the market and lose all the respect they've been earning with the newer products?

If it isn't Sigma's fault, then who is to blame? Canon for being crafty?

 

[brad-man]

Just waiting for the apologists to scold those of us who like reliable AF, or tell us we aren't true photographers.

A whole bunch of photographers paid a whole bunch more money for a 5D3 than for a 6D precisely because they wanted better AF. Unless this latest review is an aberration, or Sigma addresses AF problems without forcing us to buy a USB lens tweaker thingy, I can live with my ef 50mm 1.4 for now.

Btw, have the 35 A, and have had only one fairly dim-light AF struggling event, one that couldn't be reproduced. Like it came and went.

All that said, I'm surprised nobody has been discussing how odd it seems that AI Servo was apparently ok, but One Shot on a tripod wasn't...Any conjecture?

No scolding, but it's likely because Canon doesn't license their AF algorithms and Sigma (and the others) have to reverse-engineer them. Also, there are 2 other reviews (Phoblograper & LensTip) that have mentioned this same issue and given that one in in Poland, that has to be from at least 2 different lenses. As to why this would be different than the 35 Art, I'm not sure. There is a little less DOF, but beyond that. they should be very similar.

This is what I find curious about Brian's results. I have the 35, and although it (or I) occasionally misses focus, it is nowhere near 40%. I have not done formal testing, but I would place the number of AF misses around 5-8% under normal shooting conditions (whatever they are).

 

[sagittariansrock]

sagittarriansrock said, "I did comment on that and hypothesized it is an incompatibility with Canon's algorithm- therefore not necessarily Sigma's fault, just the trade-off with third party lenses, as Mackguyver mentions."


Missed that--But why would you say it isn't Sigma's fault? Who is making a lens first released in a CANON mount? If they CAN'T get it right, why put it on the market and lose all the respect they've been earning with the newer products?

If it isn't Sigma's fault, then who is to blame? Canon for being crafty?

It might be their fault. But not necessarily so. It might be an unavoidable issue with any and all third party fast lenses with AF. And how many of those do we have?
Tamron and Tokina don't make fast lenses. Zeiss and Samyang don't make AF lenses. And all of Sigma's earlier (Local Vision) lenses have had AF inconsistency issues.
So Sigma might have be at the receiving end of a raw deal. Let's see if they can fix it via firmware. I am hopeful, but not confident.

 

[YuengLinger]

Well, the early adopters will boldly tread, thank you. If AF truly is a problem, and it can be fixed by firmware, then it's a shame if early adopters find themselves forced to buy a USB dock. Otoh, shipping a lens in for a firmware fix has to cost something, and I don't think even Canon pays for shipping to them for a warranty repair.

I've micro-adjusted my Sigma 35mm 1.4 A on my 5D3 with good, consistent results. Knowing that works, the idea of having to buy into the USB dock is just one more stumbling block, imo.

 

[jebrady03]

Several USERS have reported AF inconsistencies with the Sigma 35 Art, I'm one of 'em. In poor lighting I end up with about a 40% hit rate, even using AI Servo and BBF on my 70D. Once the 35mm IS USM hits the refurb store, I'm selling mine. I also won't be purchasing the 50 Art.
Im really disappointed... :-/

Edit: I should note that I never go looking for problems but decided to see if my 35 Art would benefit from MFA 2 nights ago. It needed -5, so that might explain some of the difference between my 60% miss rate and Bryans 40% miss rate.

Edit 2: just checked, it was -3, not -5

 

[JustMeOregon]

Something doesn't add-up... Maybe I haven't had enough coffee this morning, maybe I've had too much, or maybe I'm hoping that my faith (and pre-order) in Sigma's new-found quality is not misplaced... But something just feels wrong here.

Just as most everyone else, I've read every other review out-there on the Sigma 50A, and the only other one that mentioned anything about any AF issues is the Lenstip review. There, the missed focus problems were limited to a single camera-body type -- the 1Ds MkIII. When Lenstip tested the 50A on a 50D and a 5D MkIII the result was "pretty much predictable, with the number of misses reaching 6-7%." Of course any missed focus is not a good thing, but Lenstip's failure rate of 6-7% (that is described as "only a good result") is pretty much what I feel I get from my Sigma 35A that I'm very happy with.

But, there is a HUGE difference between a 6-7% failure rate and a 40% failure rate!

As much as I respect (& rely on) Bryan's reviews at The Digital Picture, something just doesn't feel right here.