No, they didn't say one way or the other. They only had it for a day or two. My guess is they'll give me the same kind of B.S. about DOF, camera bodies, etc. they gave you. But we'll see.
Quick question for those who have the lens and have the AF problem at longer distances.
When you manually focus at longer distances using live view the lens can be focused correctly? It's just the AF that is fooled and sets the focus too far behind the subject?
I wonder if this problem is in any way connected to the way this lens' bokeh has been adjusted to be softer in the background but less pleasing in the forground.
This isn't strictly relevant, I suppose, but while we're beefing about Sigma ... I've been trying to buy a 12-24 for architectural work. I know this lens can be pretty damn good. An acquaintance has an exemplary copy.
So I've been shopping. I've tested two second-hand lenses from different sellers: both were defective, showing dramatic smearing of the image towards the edge of the left-hand side (on full frame). One was worse than the other but in neither case was it anything like ordinary edge softness.
So I went to my trusty local camera dealer and ordered one brand-new. It came today. A quick test revealed that it was even worse than the second-hand ones -- this copy was sharp only in the very centre of the frame, with motion-blur-like smearing everywhere else. As in the case of the others, stopping down didn't help much, if at all. The salesman was shocked enough to do his own test shots, which confirmed mine.
I asked him to order me another. He declined. Said his company had to bear some of the shipping and return costs, and he wasn't confident that I wouldn't be sending the next one back too.
I'm not a habitual Sigma-knocker, but this experience has not inspired confidence in the manufacturing competence of the brand.
In the case of the 12-24 I'm particularly pissed off because I know that good copies exist and I know the lens is a good design. I'd happily pay more if only they could sell me one that worked. It would still be a bargain compared with the Canon 14L -- which I may now be forced to stump up for.
Funny, I've been really lucky with Sigma lenses until now. I know QC is a HUGE issue for them, but I've escaped it somehow before the 50/1.4. I'm still waiting for UPS... should be here anytime. Very curious to know if they were able to fix it - or if they even acknowledged the problem.
A good Sigma 12-24mm is a rarity: I got a fair copy on my sixth try, if memory serves. My 50/1.4 is now being calibrated by Sigma UK – I'll report how it works when it returns!
hubsand wrote:
A good Sigma 12-24mm is a rarity: I got a fair copy on my sixth try, if memory serves. My 50/1.4 is now being calibrated by Sigma UK – I'll report how it works when it returns!
The Sig 15-30mm seems to be more consistent quality. My copy (1 of 1) is good on full frame. 15mm is wide enough for me on FF. Used these lenses are cheap, as many people seem to want as wide as possible.
LightonKids wrote:
Quick question for those who have the lens and have the AF problem at longer distances.
When you manually focus at longer distances using live view the lens can be focused correctly? It's just the AF that is fooled and sets the focus too far behind the subject?
/quote]
It is a pity that Sigma's QC seems to let them down so frequently. Makes one wonder how costly good quality control is during manufacture - that a general lack of it - seems to be a component in reasonable pricing - or is Sigma just addicted to shooting itself in the foot.
The only zoom I own is a Sigma 24-60/2.8 and it was the only copy of it that I purchased at the time. Build quality is more than adequate, AF is consistently accurate and the image quality is excellent. This should be par for anyone purchasing a Sigma EX category lens.
this was the first sigma lens I ever had an issue with.
I have 150macro x2..sold first one, 70-200 DG, a new 17-35 2.8-4, 10-20, 50-500DG every single one of these have been very sharp and perfect copies. my 50 1.4 was the only one I ever had any issue with
Edited by 05xrunner on Jul 31, 2008 at 07:42 AM GMT
I just got another reply from the Sigma technician Vuslat Tatar, where he continued to argue that the circle of confusion for different pixel pitches on the same sensor format was different. He equates the CoC to equal the pixel pitch? Others care to shed some light here? I thought it was a calculation that determined the point at which a circle is distinguishable as a point for a given magnification that was standardised. Also told me all EOS cameras had different flange to focal plane registration distances... not a standard 44mm... which is odd.
Anyway I digress, he skirted the entire issue even though I clearly spelt it out in all caps his response was many fast fixed focal length lenses from other manufacturers have focus calibration issues... *SIGH*
Okay, got mine back from Sigma today. A little pink form in the box had a the word "repaired" stamped in bold red ink. Forgive me if I was still pessimistic.
Here are the crops. All taken with my 5D on a tripod, cable release and center-point focus on the white book (on the left),
I think there was a slight improvement from last time. The first shot looks relatively close. The second shot is way off - it back focused on the book to the right, which is about 15 inches behind the white book. The third shot is spot on, but this is no surprise as it was only 6 feet away and my copy had no previous problems focusing at that distance.
I wish the results were more conclusive. If it was still way, way off I'd send it right back to Amazon. If it was spot on each time, I'd keep it. As it stands, the reality seems somewhere in the middle.
I think I need to do a bit more testing. I'm open to suggestions for how to test it more rigorously. These kind of tests are not my forte.
I forgot to add: quite honestly I may be making a mountain out of a molehill here. I rarely use f/1.4 at distances greater than 6 feet. I am very happy with the performance of this lens in every other respect.
I suppose I should try some shots at f/2 and f/2.8 at distances greater than 6 feet, since this is more typical of the way I shoot in the real world.
But there's something that bothers me about spending $500 on something that doesn't work the way it should.
That was the same problem I struggled with... It's very good when you can get in-focus shots and I also don't use f/1.4 over 10' *often* but the attitude of Sigma in regards to this issue frustrated me to the point that I returned it.
Also imagine the resale value if they do eventually fix it with a revision.
I only need AF for moving objects further away (like bands etc in low light) and this lens can't do that... so I spent the money I'm getting refunded on a great 50/1.2 lens and adapter... and ended up with some spare change I'll use my 85/1.8 for stuff I need AF for.
badlydrawnboy wrote:
... The second shot is way off - it back focused on the book to the right, which is about 15 inches behind the white book. The third shot is spot on, but this is no surprise as it was only 6 feet away and my copy had no previous problems focusing at that distance.
I wish the results were more conclusive. .....
I's say they are VERY conclusive. Similar reports from others would help strengthen the conclusion, but are not really needed at this point.
We now know the what, we just don't know the why.
Does this problem exist with many other third-party "Canon compatible" lenses? I don't own any so have no idea.
Are they all just gussied-up MF lenses?
Cableaddict wrote:
Does this problem exist with many other third-party "Canon compatible" lenses? I don't own any so have no idea.
Are they all just gussied-up MF lenses?
My Tokina 17/3.5 gets the focus right each time... but it has way more DOF so I don't think I'd notice as much if it didn't.
I don't have any fast primes from other brands so I can't really answer that question. Sigma do have a lot of issues with their 30/1.4 as well, but good copies of that lens do exist, it's merely a calibration issue instead of what appears to be a design flaw in this lens.
I think the real litmus test of whether this design is a dud will be when the Nikon equivalent launches and gets tested on D3/D700. If no issues there it's probably an EF specific communications issue... or we might suddenly see working copies appear in the EF mount meaning they fixed the issue.
I've had a slew of Sigma lenses as well as two Tamrons that focused perfectly. This was my only lemon.
I think I'll send mine back. Thrice's point about the resale value is a good one; I think I'd have a very hard time selling this right now considering all of the reports. I wouldn't feel good about selling it anyways.
I'll use my 85/1.8 for stuff I need AF for.