I received my Sigma 35/1.4 yesterday and took it out for a spin this morning. It was front-focusing consistently and badly:







So I fired up Reikan FoCal, thinking it wouldn't even be able to calibrate the lens based on how badly it was front-focusing. But not only did it come back with a tiny adjustment in the wrong direction (-1), it was the fastest calibration I've ever done with FoCal. It only took 5 readings: -20, -10, 0, 10, 20 and -1. Then I did a focus consistency test with 10 shots at AFMA -1 and got a 98.7% CoF, which is pretty good from what I understand.

So then I took a few more shots, thinking those original ones could have been an anomaly. They weren't (focus was on the dog's eyes):







(Side note: dogs are great test subjects for focus issues!)

Now I am really confused. How is it possible that the lens would front-focus so consistently in real-world shooting, but score so well in a focus test? If the problem is with the camera, as I suspected, then how could the lens do well in a test?

I should point out that I'm beginning to suspect that the AF is off in my 5D3, not this lens. Why? Because I've had AF issues with many lenses. At first I thought the issues were with the lenses, especially because two of them were Sigmas (50/1.4 and 85/1.4). But I also had issues with a Canon 50/1.4. I sent it to Canon, they said the focus assembly was broken and fixed it, but after getting it back it still doesn't focus accurately.

But again, if the problem is with the camera, then how could the 35/1.4 have performed so well on the FoCal test? In addition, I just bought a 24-70 II and did a test in FoCal, and it returned a zero adjustment on the wide end and a +1 on the tele end. Would that be possible if the camera's AF were off?

How does the 24-70 do in actual use? And as I said in the other thread, I'd suggest trying to establish the correct adjustment for the Sigma 35 by trial and error, never mind the program. Since you're getting consistent front focusing with actual shots, why not try setting the adjustment to +20 and seeing what that comes up with? I'd expect it to be consistently focusing farther from the camera, but I'd definitely want to try that and see what happens. You've obviously got access to a dog, which I understand makes a good focusing test.

Ernie Aubert wrote:
How does the 24-70 do in actual use? And as I said in the other thread, I'd suggest trying to establish the correct adjustment for the Sigma 35 by trial and error, never mind the program. Since you're getting consistent front focusing with actual shots, why not try setting the adjustment to +20 and seeing what that comes up with? I'd expect it to be consistently focusing farther from the camera, but I'd definitely want to try that and see what happens. You've obviously got access to a dog, which I understand makes a good focusing test.

The 24-70 arrived in the same package as the 35/1.4, and I was busy testing the 35 so haven't had a chance to shoot with the 24-70 yet. I'll do that tomorrow. The DOF at 35mm and f/1.4 and a subject distance of 4 feet is about 4 inches. So, to replicate that at f/2.8, I should shoot with the 24-70 at 50mm and the same subject distance (DOF of 3.82 inches).

Hopefully I'll be able to tell if it's the lens or camera that way.

I just took a few shots with the 24-70. It's definitely better than the 35/1.4, but still seems a bit off to me. Here are a few 100% crops, all at f/2.8 between 50mm & 70mm. I've seen samples of this lens online that are much sharper at f/2.8. What do you think?

It does seem there may be a problem with the 35/1.4, since it's much further off than the 24-70. But that doesn't mean there isn't also a problem with the camera's AF.

Did you do your FoCal calibration @the same distance as your subjects shots?

Maybe the lens needs different MA at different distances. ?

I wonder what the new USB widget could do with the lens ?

Just to be on the safe side, I'd go to a Canon lab and have the 5D3 checked.

Happy shooting,
Yakim.

Yes, I'm definitely going to send the 5D3 into the Canon Service Center. It's still under warranty so like you said, I just want to be sure it's not an issue with the camera.

And so the Sigma Stigma begins. It was exactly this sort of nonsense which forced me to sell my Sigma 85/1.4 for the Canon L. I haven't looked back since.

AF woes. Good luck in resolving this.

badlydrawnboy wrote:
Yes, I'm definitely going to send the 5D3 into the Canon Service Center. It's still under warranty so like you said, I just want to be sure it's not an issue with the camera.

Yes, have camera checked under warranty.

I am biased here, but I still think it's a lens problem 'cos don't trust AF on third party lenses.

AF woes are the most painful... If you can live with deeper DOF, MILCs may be the long term solution.

Welcome to my world with the FoCal program!

I find it highly inaccurate anymore. Back when the program was new and only worked with the 5D2, it was highly accurate for me. Been using it since the beta days... LOVED the program back then...

I find it has gotten worse over time, only getting more fancy with its feature-set.

With my 5D3, properly setup at EV +10 or more, outside light, inside light, completely level/measured, 25x or 50x focal length, etc... I get some varying results even though it says its perfect. I have tried different papers, sizes, everything. Nothing helps get me accurate results.

Some examples...

Had my 70-200 II tested over and over at something like +3W and +8T... With a TC, it was something like 0W and +6T. My 24-70 II tested at -2W and +3T... 85 1.8 at something like -7...

Mind you, the body and lenses were sent to Canon for calibration (body, 70-200 II and TC all together)... Afterward, the tests showed slightly less away from 0, but still consistently giving me the same (+) numbers. Real world results were terrible. Set all back to 0 and images are MILES better. Canon properly did the calibration to the body... FoCal was giving me "Excellent" results with very defined bell curves to the MA value it would give me. I should just trust Canon and leave it all alone (which I do now).

Same with the 70-200 + TC... Gave me some +4 and +6 or something. Causing backfocus on real-world shots. Set all to 0 and perfect. Was bummed because many of my aviary shots that weekend were ruined.

My 85 was most crazy though... Tested over and over and got -7 at (3) different distances and types of light. Real world shooting showed very noticeable front focus. Put back to 0 and perfect.

24-70 II was the ONLY lens it got right on the wide end (-2)... The tele end best at 0 and not +3.


Not sure what is up with Focal, but it simply does not work for me anymore. Before people try to claim its the target setup, trust me, its not. I did both 25x and 50x the FL of the lens, measured to make sure everything was level and parallel, very sturdy tripod, remote release, mirror lockup time extended, lighting was even and consistent (with halogen, natural light, shade/direct, etc...). Program spits out excellent test results but simply does not play well with real-world shooting, at all.

Camera was checked/adjusted/calibrated by Canon, as were the lenses. With real-world shooting, I notice NO focus issues. Spot on actually, as long as I don't use the FoCal values...

Again, used to LOVE the program. Back when it first came out, I did my MA via LensAlign and FoCal matched those values almost EXACTLY in a fraction of the time. After the target changes and all the "updates", it simply does not give me good results. I want to love the program like I used to, but it simply does work with my 5D3 and lens collection for whatever reason.

*shrug*


After all the headache with FoCal, my lenses CLEARLY perform best with no MA anymore, granted, the 70-200 and TC combo was calibrated directly so should be perfect. But the 24-70 II and 85 never were. My 5D3 has been at Canon for check/clean twice this past year and have been checked/adjusted as needed.

70-200 II : 0W/0T
70-200 II + 1.4x TC : 0W/0T
24-70 II : -2W/0T
85 1.8: 0


My advice to you would be to step away from Focal and if its front focusing, add (+) values until you get good results. See if that fixes your issue. Ignore what FoCal is telling you, because yes, it lies.

Thanks for sharing your experience. The problem for me is that the amount of front-focus evident on the 35/1.4 could not be corrected even by a +20 adjustment. I noticed the 35/1.4 was front-focusing badly even before I did any MA. I will try setting the 24-70 back to zero on the tele end, and see if that helps. But the fact that I've had focus issues with almost all of my lenses does suggest it might be a problem with the 5D3. I'm going to send it to Canon to double check. Also, although I know Sigma is prone to AF issues, so many people (including Roger from LensRentals) say the 35/1.4 is different in that regard.

Who knows... it's so complicated!

Invertalon wrote:
Again, used to LOVE the program. Back when it first came out, I did my MA via LensAlign and FoCal matched those values almost EXACTLY in a fraction of the time. After the target changes and all the "updates", it simply does not give me good results.

Wondering if you tried LensAlign again. Perhaps automating the process isn't such a good idea after all.

badlydrawnboy wrote:
The problem for me is that the amount of front-focus evident on the 35/1.4 could not be corrected even by a +20 adjustment.

What about trying the new Sigma USB adjustment dock? Wondering how reliable it is.

deepbluejh wrote:
And so the Sigma Stigma begins. It was exactly this sort of nonsense which forced me to sell my Sigma 85/1.4 for the Canon L. I haven't looked back since.

I'm not so sure about that. The OP experienced problems with Canon lenses as well.

badlydrawnboy wrote:
I should point out that I'm beginning to suspect that the AF is off in my 5D3, not this lens. Why? Because I've had AF issues with many lenses. At first I thought the issues were with the lenses, especially because two of them were Sigmas (50/1.4 and 85/1.4). But I also had issues with a Canon 50/1.4. I sent it to Canon, they said the focus assembly was broken and fixed it, but after getting it back it still doesn't focus accurately.

Happy shooting,
Yakim.

I have the Sigma 35 1.4 in Nikon mount. I am not having any focus issues. I did calibrate the lens using Focal, it came out to a minus 4. Focusing is rock solid at any distance including close ups. I have never used a 1.4 lens this good wide open.

Gene

badlydrawnboy wrote:
Thanks for sharing your experience. The problem for me is that the amount of front-focus evident on the 35/1.4 could not be corrected even by a +20 adjustment. I noticed the 35/1.4 was front-focusing badly even before I did any MA. I will try setting the 24-70 back to zero on the tele end, and see if that helps. But the fact that I've had focus issues with almost all of my lenses does suggest it might be a problem with the 5D3. I'm going to send it to Canon to double check. Also, although I know Sigma is prone to AF issues, so many people (including Roger from LensRentals) say the 35/1.4 is different in that regard.

Who knows... it's so complicated!

I don't remember if you've mentioned the notion of sending the 35 to Sigma for attention. Perhaps there's something out of kilter with it that they could detect and correct, whether Canon can find anything amiss with the camera or not.

thw2 wrote:
badlydrawnboy wrote:
The problem for me is that the amount of front-focus evident on the 35/1.4 could not be corrected even by a +20 adjustment.

What about trying the new Sigma USB adjustment dock? Wondering how reliable it is.

Is it available? I haven't heard any reports of anyone using it.

Ernie Aubert wrote:
badlydrawnboy wrote:
Thanks for sharing your experience. The problem for me is that the amount of front-focus evident on the 35/1.4 could not be corrected even by a +20 adjustment. I noticed the 35/1.4 was front-focusing badly even before I did any MA. I will try setting the 24-70 back to zero on the tele end, and see if that helps. But the fact that I've had focus issues with almost all of my lenses does suggest it might be a problem with the 5D3. I'm going to send it to Canon to double check. Also, although I know Sigma is prone to AF issues, so many people (including Roger from LensRentals) say the 35/1.4 is different in that regard.

Who knows... it's so complicated!

I don't remember if you've mentioned the notion of sending the 35 to Sigma for attention. Perhaps there's something out of kilter with it that they could detect and correct, whether Canon can find anything amiss with the camera or not.

If a malfunctioning lens is brand new, I'd rather return it than send it to Sigma. I've heard of people sending their lenses to Sigma (and Canon, for that matter) with obvious problems and getting them back with the same problem and a note saying something to the effect of "the lens is operating within normal parameters". No thanks. Between myself and my company I've spent several thousand bucks at B&H over the past 6 months; one reason is their excellent return policy and customer service.

My current plan is to send the 5D3 to Canon after Christmas. When I get it back, I'll test the 35/1.4 and 24-70 again. If they're both still off, back to B&H they go.

Yeah, it's "Ya pays yer money and ya takes yer choice" for sure. There have also been instances (multiple for me personally) of two different scenarios: 1) Sending a lens (and indeed, a camera) in to Canon reporting an issue, and getting it back with the issue resolved; and B) Sending a new malfunctioning lens back to the retailer to exchange, and getting back another one with the same or some other problem - this happened to me once with three successive copies of one lens.

You never know ahead of time what the message on your fortune cookie paper is going to be.