I took some care checking my lens' AF issues using a tripod, 2 sec MLU, a long ruler, a good target and all that. The focusing error was so severe that I couldn't draw any other conclusion than that the lens has a front focusing problem. I measured at 1,2,4 and 10m and there was no difference. Here is the 2m result: http://photos.imageevent.com/jonas_b/fmforumsmonthly/fmforums2008/118_Siggy_front_focusing.jpg
Ouch. I have now agreed with the Sigma rep to bring the lens and the camera to their shop to have it examined. The camera is a 450D and it focuses just fine with my other Canon lenses (which also worked fine with my 5D). I live not very far from the shop and I'll be at place when they adjust the lens.
Glassbottle wrote:
The longer the lens, the less the effect, it seems to me. But it must also depend on which of a whole heap of optical compromises the designers have made for a particular wide-aperture lens.
I compared a 180 f2 to a Contax 200 f2 and a Canon 200 f1.8. Both the Contax and the Canon had visible colour separation in the o.o.f. highlights, whereas the Leica did not. Overall I preferred the Canon and kept it, but the Leica 180 seems to have practically no colour separation in bokeh.
I sent mine into Sigma on Friday without the 5D. I asked them if they had another 5D to test it with and they said yes, but that the AF points vary slightly from camera to camera. I didn't feel good about sending my camera for the reasons FretNoMore mentioned, so if they can't get it right using their 5D I think I'll just return the lens. I should get it back in about a week or ten days, and I'll report to this thread when I do.
I drove down to Kungsbacka, Sweden, where Tura Scandinavica has their repair shop. I have met Robert, who works there earlier, at one occasion last autumn. Then he fixed a focusing issue with my Sigma 30/1.4. Now he promised to try to fix the 50/1.4.
Robert is very friendly and didn't having me hanging around during the process. I learned a few things:
* Sigma still uses 40xFL as their standard adjustment distance
* The chips in the lenses aren't very smart; there is no possibility for Robert et al to adjust for focus shift due to RSA
* The 450D and the 50/1.4 are playing reasonably well, but not very well, together. Sometimes the camera and lens "micro-hunts" for the, hopefully, correct focus adjustment
Robert confirmed my findings about the bad front focusing problem. Then the chip was reprogrammed, twice, before I got the camera back for a short walk and some real life images.
It wasn't possible to adjust the lens (at least not to this camera) to work perfectly. Sometimes the focal plane ended up 1-2mm in front of the target, sometimes 1-2mm behind the target and sometimes spot on. We agreed not to continue the adjustment process but stop there.
All in all a 40 minute process. Normally the process takes less time but the fast primes are tricky, and Robert suggested the extra work with taking and examining some test images.
I'm very happy with the friendly Sigma service here. Frankly, if I didn't live close to their service center I would be more reluctant buying Sigma lenses.
well I was having my backfocus issue that was bad with anything over 6feet. I sent it to sigma NY and said it was working fine on their xxD bodies and had no 1D's to test with.
Came back same as it was. So I was playing around and put my flash on so it would fire the focus assist and it seems to be pretty accurate now even a target about 18feet away for me. so is it my camera or lens causing the miss focus then. Since it will be accurate with the focus assist light
I think we will also soon see some test results from my lens. Joseph James (joe mama) has
my lens right now. His 5D died right after he received my lens, so he wasn't able to complete
his tests, but Canon is fixing his camera right now (new shutter and mirror box -- $650),
so I think he will have it in his hands in a couple of days. I asked him to verify autofocus
performance on his body. I didn't see much problem on my 30D, but I didn't really look for it, either.
kinda sucks. I cant return it cause its past the return window. No one will buy it. and sigma claims it works perfect with their 20-30-40D's and they want me to send my camera in with the lens to do its adjustment. I dont wanna not have a camera for a week or so
just like the pic shows..center focus point and one shot. the entire focus point is taken up by the basketballs hoops poles. there should be no issue. This lens either has a BAD backfocus issue like I have been seeing..or its just really really inaccurate
Jonas B wrote:
It wasn't possible to adjust the lens (at least not to this camera) to work perfectly. Sometimes the focal plane ended up 1-2mm in front of the target, sometimes 1-2mm behind the target and sometimes spot on. We agreed not to continue the adjustment process but stop there.
So are you happy? 1mm-2mm at 40x FL (2 metres if I'm not mistaken) sounds perfectly ok to me.
I assume you're not seeing the behaviour that others have reported, severe misfocusing at distances greater than 3m/10ft while focus is correct on closer targets. Am I right?
Glassbottle wrote:
So are you happy? 1mm-2mm at 40x FL (2 metres if I'm not mistaken) sounds perfectly ok to me.
I assume you're not seeing the behaviour that others have reported, severe misfocusing at distances greater than 3m/10ft while focus is correct on closer targets. Am I right?
Hi,
I'm happy as the lens now can focus... Earlier I had severe front focusing at all distances and couldn't make any conclusions at all.
Now after having the lens adjusted I have noticed the same problem as other s have reported. I'll have to do a more controlled test next week (I'm again leaving home for a couple of days). I would say that around 3m is where things start to happen. Aiming at something very far away, "infinity" things aren't reasonably sharp until I have stopped down to around f/5.6.
But the bokeh is the best I have seen any modern lens with high contrast and resolution wide open make. Maybe we can't have it all?
Is it right to draw the conclusion that AF is maybe not as fast as one would have hoped given the HSM motor?
It's also really disappointing to read about these AF problems. It's a good thing I wasn't planning to buy until later in the year anyway, hopefully Sigma has tightened up their QA by then.
Glassbottle wrote:
So are you happy? 1mm-2mm at 40x FL (2 metres if I'm not mistaken) sounds perfectly ok to me.
I assume you're not seeing the behaviour that others have reported, severe misfocusing at distances greater than 3m/10ft while focus is correct on closer targets. Am I right?
I got a call today to say that my lens is ready for pickup tomorrow. I've got work the exact same hours the repair centre is open so I'm trying to get that changed and will add my findings either tomorrow or early Thursday.
I think that is good enough, yes. 1-2mm out of 2000... In the fortunate cases that the camera really locks on what I have in mind this is good. I agree that 2mm focus error can look ugly sometimes but in most cases it doesn't matter much. Then we can add the photographer (at least me) swaying and the target is probably moving a little as well.
With a static target and the camera on tripod there is no reason not to check the focus manually.
I also see 1-2mm as a great improvement to the 150mm error I had when arriving to Sigma.
