Woe and Pathos in the Sigma 50 Art?

[[email protected]]

(Please excuse any unintentional etiquette errors: I posted this in another topic link and I shall remove that one.)

Colleagues:
I bought a Sigma 50 Art. I am not at all happy with the focus: it is gorgeously, excruciatingly sharp ... about 1/3 of the time. My standards for "sharp" are pretty high, and my reference lenses are the new 70-200 and 24-70. I am a thoroughly experienced photographer with decades of work behind me.
I am utterly baffled; I never, ever have this problem with other lenses; I have never, for instance, used microfocus adjustment on my 5D3 or 1Dx for my reference lenses. Here are a couple of photos from the other day, and I was focusing on the inner corner of the left eye (the 'medial canthus.') By the way, re. the girl with horns, do not adjust your set, you are not hallucinating: the dots on the cornea are contact lenses ... I will leave these up for a few days. Your thoughts deeply appreciated.
https://www.flickr.com/photos/johnrudoff/sets/72157648392663316/

 

[tayassu]

Looking at you pictures, I know what you mean... I have the same problem with my Tamron 24-70.
I would definitely AFMA the lens, maybe even with the USB dock. If that does not work for you and there are still some inconsistencies, I would send the camera and lens combos to Sigma for professional adjusting (helped with my Tamron a little, but then noticed the camera is the problem )
BUT I have also noticed it is not only the microadjustment of the lens but also the focus technique. I normally do spontaneous portraits/people shots with AI Servo (respectively don't hold the AF-Off button ->BBF), but with the Tamron I'm actually on the safer side with holding the button down (respectively shooting One Shot) and let go whenever the person moves... That improved my hitrate dramatically, but I'm not a pro as you are.
I hope that helps a little to solve your problem!

 

[Sporgon]

Looks like you must use the Sigma dock to calibrate these very fast lenses reliably - which is why I don't use them personally.

 

[CarlMillerPhoto]

I think is the perfect example of why the Sigma AF issues are so bloated and exaggerated. Those shots are absolutely fine and usable. It's posts like these (though usually without sample photos) that infect others with paranoia, causing them to be hypersensitive to their Sigma AF. Then they go out and do all these repeated, silly tests looking and looking to find that issue and what do you know - they find one. Hmm, think their drive and determination to find a problem might have resulted in a false-positive? Every lens has AF that varies a bit, but the problem is most don't scrutinize their other glass to the same degree, leading them to the erroneous conclusion that it's a uniquely Sigma or third-party lens issue, when in reality most of their glass would perform similarly under the same inspection. While of course there are very valid issues with the occasional lens, the Sigma "AF epidemic" is hardly what online camera communities make it out to be.

Did I mention those sample shots are fine and usable? Maybe a lil AFMA and you're good. Seriously.

 

[Shanly]

Having to AFMA a lens does not mean that there is a problem with the lens or with the camera for that matter. I have 6 lenses, 4 of them are Canon L's. I have had to AFMA every single one of them. They each have their own unique adjustment. I am very happy with all of them and they focus well.

I don't have the Sigma 50mm ART, but I do have the 35mm ART. I have the USB dock but only use that to upgrade the firmware. I AFMA'd the lens and find that it focuses consistently well and that any problems focusing at f/1.4 appear to be my problem. It is a wonderful lens although it took some time to get used to not zooming!

BTW, I use a LensAlign target to perform AFMA. Works well for me.

I also have very occasional focus misses on my L lenses, but, again, that's seems to be my fault due to the very thin depth of field.

I do wonder if sometimes people (not necessarily the originator of this thread) are quick to jump on non-Canon lenses for focusing issues that are no more frequent than Canon. YMMV.

To be fair, I don't use servo mode with my 35mm.

 

[KBStudio]

Not finding any problems with my copy of the Sigma 50 Art. I have been it using both my 5D MII and a Sony A7r. AF on the Sony is impossible due to the Metabones adapter so I use focus peaking and LCD enlargements to focus. But AF is fine on the 5D MII. When the focus is off I am finding it more likely "operator error".

 

[teedidy]

I just finished renting this lens and the usb dock for the past 2 weeks. Everything that is said about this lens is true. It is wonderfully sharp, and without the usb dock it has severe focusing issues. AFMA does not help except for the exact distance you set the AFMA for. Using the usb dock and software you set the focus for 4 different distances. each of the distances had different calibrations. After some meticulous calibrations (roughly 60 minutes tethered and zooming in at 100%), this lens worked great! Testing for constancy I was able to find the following miss rates:

Canon 50 f1.8 @1.8 = 3/60 miss rate
Canon 24-70 @ f2.8/50mm = 1/60 miss rate
Sigma 50 art @ 1.4 = 3/60 miss rate (closer to 2.5 as one was usable, but not tack sharp so I included it)

In summary, unless someone spends the same required attention to detail I refuse to consider their input on the quality and consistency of this lens' focusing ability. This is an awesome lens, but only with the addition of usb dock.

If requested I can upload some of my testing shots and discus my methodology.

 

[Arctic Photo]

I just finished renting this lens and the usb dock for the past 2 weeks. Everything that is said about this lens is true. It is wonderfully sharp, and without the usb dock it has severe focusing issues. AFMA does not help except for the exact distance you set the AFMA for. Using the usb dock and software you set the focus for 4 different distances. each of the distances had different calibrations. After some meticulous calibrations (roughly 60 minutes tethered and zooming in at 100%), this lens worked great! Testing for constancy I was able to find the following miss rates:

Canon 50 f1.8 @1.8 = 3/60 miss rate
Canon 24-70 @ f2.8/50mm = 1/60 miss rate
Sigma 50 art @ 1.4 = 3/60 miss rate (closer to 2.5 as one was usable, but not tack sharp so I included it)

In summary, unless someone spends the same required attention to detail I refuse to consider their input on the quality and consistency of this lens' focusing ability. This is an awesome lens, but only with the addition of usb dock.

If requested I can upload some of my testing shots and discus my methodology.

Hi, very interesting input. I am getting this lens in a couple of months (when I get paid for an article with pictures I just sold). I'd definitely like to know more on how you went about this.

 

[YuengLinger]

I think is the perfect example of why the Sigma AF issues are so bloated and exaggerated. Those shots are absolutely fine and usable. It's posts like these (though usually without sample photos) that infect others with paranoia, causing them to be hypersensitive to their Sigma AF. Then they go out and do all these repeated, silly tests looking and looking to find that issue and what do you know - they find one. Hmm, think their drive and determination to find a problem might have resulted in a false-positive? Every lens has AF that varies a bit, but the problem is most don't scrutinize their other glass to the same degree, leading them to the erroneous conclusion that it's a uniquely Sigma or third-party lens issue, when in reality most of their glass would perform similarly under the same inspection. While of course there are very valid issues with the occasional lens, the Sigma "AF epidemic" is hardly what online camera communities make it out to be.

Did I mention those sample shots are fine and usable? Maybe a lil AFMA and you're good. Seriously.

I had a bad one. Outer AF useless on two 5D3's, center point erratic. Otoh, my 35 Art was great after typical AFMA in-camera.

Sent it back for a refund, but won't try again til firmware update.

Why bash customers who have had bad luck?

 

[[email protected]]

I am very grateful for your input. PLEASE keep it coming. I have the USB dock and will get to work on it; will also try lensAlign. I admit to a high degree of extreme paranoia about sharpness. This is my problem, not Sigma's. But I just finished a 2-hour theater shoot with the 70-200 and 24-70 and I saw lots of liquid and fine red capillaries on eyeballs, handheld, wide-open (the aperture, not the eyeball); and so my paranoia is bone deep. I'll give it a try though. Thanks again, keep it coming.

 

[Mt Spokane Photography]

I've found that most, but not all lenses need AFMA. Generally, you can improve focus accuracy, but some times its already perfect. Reikan Focal does a super job for me, with my 5D MK III, I can calibrate zooms in two places (Wide and Tele), and all my zooms benefit, none are perfect at the ends of the zoom range.

Cameras also have variability. Very tiny distances are involved, so there is a tolerance, and sometimes a body is at one end of its tolerance and the lens at the other so you can end up with two in-spec components needing a 10 AFMA.

 

[Hjalmarg1]

I am very grateful for your input. PLEASE keep it coming. I have the USB dock and will get to work on it; will also try lensAlign. I admit to a high degree of extreme paranoia about sharpness. This is my problem, not Sigma's. But I just finished a 2-hour theater shoot with the 70-200 and 24-70 and I saw lots of liquid and fine red capillaries on eyeballs, handheld, wide-open (the aperture, not the eyeball); and so my paranoia is bone deep. I'll give it a try though. Thanks again, keep it coming.

Sigma 50(A) is being considered the second sharpest "standard" lens so, when properly calibrated should satisfy your sharpness requirements.
By the way, I have the 100mm f2.8L IS and 35mm f2 IS and both required a very little AFMA on my 7D and 5D3. It wasn't much but there was needed to get perfect AF.

 

[YuengLinger]

Please pay attention: The AF problems popping up on the Sigma 50mm ART are not AFMA related. It is some kind of firmware issue or other problem that causes certain cameras, such as the 5D3, to not be able to use the outer cross-type points at all, or the inner points reliably.

Apparently massive numbers of owners are NOT having this problem, but I had it, as have others seeking answers in this forum.

So, take a deep breath, and stifle that urge to defend a brand without hearing what the problem is. And I will repeat: The 35 ART is one of my favorite lenses ever, as is my Sigma 15mm 2.8 fisheye!

 

[infared]

My first copy was BAD. Totally psycho. All over the place with the auto focus. I spent the better part of a day trying to calibrate it with the the dock, then thought, "why am I doing someone else's job???". LOL!

The lens was so incredible when actually focused, that I HAD to give it another shot. (B&H is very understanding of this issue.) I sent the first one back and tried another. Glad I did. Out of the box it was a completely different experience. Very slight tweaking on the dock. I left the focus adjustments on my 5DIII on zero.

The whole experience is a bit ridiculous for a $950 lens..but it does WOW me.

 

[razeac]

I have one. Maybe one of the first who acquired the lens here. I do not have the USB dock. While it's true that it has issues on its autofocusing, I have learned to compensate for it. Don't use it on sports - it's just not meant for it. I use live view very often. And one shot focus only - and spamming that button before I press the shutter. This way I get more accurate focus most of the time. This lens rocks, the images are really sharp. Keep it and learn its ways.

 

[Hector1970]

Maybe I need to get my eyes checked. The last time they were checked the results were excellent but maybe they are starting to go wonky.
Even full size the picture look sharp. I'm amazed as to how sharp people expect a lens to be.
These photographs are shooting a swaying object focused on an eye which much be moving back and forth a few mm/cm all the time.
There is a very good article in here somewhere about lens and camera tolerances.
Sometimes you have a camera +3 and and lens +3 so it's out +6
other times you have a camera -3 and and a lens +3 and you have a perfect combination.
People blame the lens when there is nothing wrong with the lens (as in its within acceptable tolerances). Its the combination of lens and camera that's at fault.
Maybe I should be more demanding of my lenses. The Canon 50 1.4 wouldn't be as sharp as that.
I'm not sure more sharpness would add in any way to the photos.
But as you say you are a bit of fanatic about sharpness.
I thought I was a bit too but it actually looks like I actually expect too little.
Do the AFMA, hopefully you'll find sharpness and happiness and contentment with what looks like a pretty exciting lens. I should consider it myself.

 

[Eldar]

I am very happy for all of you with working copies of this lens. Optically it is phenomenal and it is a steal, when it works. I see many of you doubting some of the results others have had with this lens and write it off on impatience, incompetence or what have you.

I have been doing this for some time and I am a fairly patient guy. I run all my lenses trought AFMA, primarily using FoCal, but also with a manual LensAlign rig. Especially the very fast lenses (f1.2, f1.4 type) benefit from that. Some require a bit more than the others, but the main thing is to be consistent.

I have done my best with two copies of the 50 Art and I don't need anyone to tell me I don't know what I'm doing. I know what to expect from a lens, I know how to focus, I know how to do AFMA and I am able to judge if an image is in or out of focus.

The issue with these lenses are not AFMA. The issue is incosisntency. The first I got was so bad I was unable to establish a conclusion to the AFMA measurement, both through FoCal and manually. It was all over the place and totally unpredictable. I bought the dock and did my best with that. After a few weeks of trying, I returned the lens and got a new copy. This one was better, but still very inconsistent. After a couple of weeks trying, I returned it and got my money back. I'm not talking 3% off or 6% off, but rather 50-60% off and sufficiently off to be useless.

The problem here is that Sigma obviously have a very poor quality regime for these lenses. The fact that some of you apparently have lenses you can use, show that it is possible to make it work. I have bought my last Sigma though.

 

[flylife]

I have one too and at first, after reading "horror stories" on here, was pretty sceptical about the focus. When I got it I did an AFMA correction (nothing unusual there, all my lenses need it) and got shooting real stuff. it seemed all over the place, but MAN it was sharp when it was on. I was getting quite frustrated, so thought I'd compare my other lenses.
Canon 85 1.8, less consistent.
Canon 50 1.4, WAAAY less consistent
135 f2, better
Canon 70-200 f4IS, less consistent
Canon 100 macro, about the same.

What was really telling from this though was that the Sigma is so much sharper wide open than all of these lenses that its easier to see that it's out of focus. With the Canon 50 1.4 for instance you only get a vague notion of where the focus point is and if it's wrong. When I really look at the shots from that lens, it's less consistent than the Sig, but just harder to tell: they're all so soft that NOTHING looks really sharp, so i assume that it's in focus. i think they're ALL inconsistent but we just can't usually see it that well.

I think it's a victim of its own awesomeness!

I got a dock, calibrated on that, and it's been great in the real world. No-one has complained because the front of their eyelashes are out but their pupil is sharp.

Now I'm wishing Canon had a dock!!!

 

[Eldar]

I have one too and at first, after reading "horror stories" on here, was pretty sceptical about the focus. When I got it I did an AFMA correction (nothing unusual there, all my lenses need it) and got shooting real stuff. it seemed all over the place, but MAN it was sharp when it was on. I was getting quite frustrated, so thought I'd compare my other lenses.
Canon 85 1.8, less consistent.
Canon 50 1.4, WAAAY less consistent
135 f2, better
Canon 70-200 f4IS, less consistent
Canon 100 macro, about the same.

What was really telling from this though was that the Sigma is so much sharper wide open than all of these lenses that its easier to see that it's out of focus. With the Canon 50 1.4 for instance you only get a vague notion of where the focus point is and if it's wrong. When I really look at the shots from that lens, it's less consistent than the Sig, but just harder to tell: they're all so soft that NOTHING looks really sharp, so i assume that it's in focus. i think they're ALL inconsistent but we just can't usually see it that well.

I think it's a victim of its own awesomeness!

I got a dock, calibrated on that, and it's been great in the real world. No-one has complained because the front of their eyelashes are out but their pupil is sharp.

Now I'm wishing Canon had a dock!!!

I could compare my copies to the a number of other high quality and fast lenses, including L-series and Zeiss. The inconsistency of the 50 Art has nothing to do with it being sharper. On a regular LensAlign test, which is a very visual and simple process, I could have AFMA variations from -15 to +15 in a controlled 10 shot series. If your lens is as good as you say, compared to your other lenses, you apparently have one of the better ones.

 

[YuengLinger]

My first copy was BAD. Totally psycho. All over the place with the auto focus. I spent the better part of a day trying to calibrate it with the the dock, then thought, "why am I doing someone else's job???". LOL!

The lens was so incredible when actually focused, that I HAD to give it another shot. (B&H is very understanding of this issue.) I sent the first one back and tried another. Glad I did. Out of the box it was a completely different experience. Very slight tweaking on the dock. I left the focus adjustments on my 5DIII on zero.

The whole experience is a bit ridiculous for a $950 lens..but it does WOW me.

Glad to hear this additional case of success on a second try. I will definitely buy a new one when the firmware update is announced.