There have already been people saying they are happy with the lens.
Chez Wimpy, posting over at dpreview.com has shown good results
and posted favorably about it. There are others there too.
LightRules tested it against the Canon 50/1.4 and found that the
autofocus accuracy was about the same.
But I understand your concern, and I'll try to report soon.
Given my past experience with Canon autofocus performance in
general, I don't expect my report to be unreservedly favorable, but
I am already enjoying the Sigma, even if it turns out to have
some AF weakness.
Stay tuned...........
Edited by trumpet_guy on Aug 09, 2008 at 03:17 AM GMT
I live in Japan and I just purchased the sigma (sn 1006XXX).
I was able to test out two out-of-the-box lenses in-store. I tried both of them with a 40D, 450D, and 400D. The AF performed about the same across all bases, great up until about 3m. One had more BF than the other, the better of the two had about 10cm at 3m and worse further out (wide open). I bought the better one.
The CA, spherical aberration, and vignetting are not bad at all, much better than the Canon 50 1.4, and better CA than the L (all wide open). Mind you these were rough tests that I could do in-store.
I'm a fan of the bokeh from this lens. I was able to get some nice round "perfect" bokeh from two overlapping light sources. The information from small bright light sources is not diminished as much as I thought it would be with this lens, and at the same time it's not busy at all. I need to experiment some more with this though.
I plan to do some more tests soon. But before that I'm going to go to the sigma service station in Kanagawa and get it calibrated. If the AF is still crap at >3m I'm just going to return it.
Jinashi wrote:
I live in Japan and I just purchased the sigma (sn 1006XXX).
(...)
I plan to do some more tests soon. But before that I'm going to go to the sigma service station in Kanagawa and get it calibrated. If the AF is still crap at >3m I'm just going to return it.
The post above ("Part 6") is one part of a progressing number of posts where I slowly share what I slowly learn about the lens. I am now looking forward to try the lens on a 5D (provided I find a used 5D for a reasonable price).
2)
Keep it or sell it? In the case Sigma doesn't provide any solution to the AF problems many reports about I guess we all have the choice to either keep the lenses , sell them or return them. As I have found the lens to be the best fast 50 I ever tried (to my eyes and taste) I contemplate keeping it and just use it as an auto aperture manual focusing good fast 50. I don't see the point in returning in and go back to a completely manual inferior 50mm. Maybe some frustrations are parts of life.
I really do want to keep it if I can get it close. Jonas, I read your post about your calibration experience. I think my lens is already fairly close compared to some other reports here, so hopefully they can calibrate it well.
After playing with it some more I think it's going to be a hard decision to return it, for my purposes stopping it down at larger distances might be ok. I'll see how it goes at my next photoshoot.
The nice thing is that the service is free here! But it takes 2 weeks....
I will also remind folks that I had two copies of this lens, both exhibited the back-focus at distances greater than 2m problem. The second copy was calibrated and apparently "fixed" by Sigma USA, but it the focus problems weren't remedied and somehow the sharpness at f/1.4 got worse even at close distances!
I think I'll wait a few months and see what happens.
I have tested the auto focus on my copy extensively on a tripod over the
last two days. I've compared the results to shots from the same distance
with my Canon 50/1.4 and found that the auto focus on the Sigma works
fine with my 5D -- over the tested range (2 feet to infinity).
That's the summary of my results, but of course there are peculiarities to each
lens, which I plan to report on soon.
Suffice it to say that I would not hesitate to use AF at any range of focus.
My copy of the Sigma seems to have a little more variability in its
AF [the variance of the distribution is greater, for you statistical types] than the
Canon lens does, but there is very little consistent bias toward back focus or
front focus. I say "very little" because I did observe a tendency to some front
focus at shorter focus distances around 3 feet to 10 feet, but it is a minor amount,
and it doesn't seem to be consistently apparent in non-testing situations.
To tell you the truth, I'm a pixel peeper, and the bad experiences of others on the
boards have made me pixel peep on this lens even more. This lens delivers the
goods when it needs to.
I'll put together a complete report and put it on my pbase site soon.
I shot a bunch of shots around Palo Alto last night, well after dark.
With the 5D, I was getting very good focus accuracy at a range of distances.
And the optical quality -- Man! No CA, wonderful sharpness wide open,
great contrast. And mixed in with these great shots were some that just
missed focus altogether. I've come to accept this with Canon's non-1-series
bodies.
With the parallel testing I've been doing with the Canon 50/1.4 I've been
impressed with it as well. If you need a flat field of focus, go with the Canon,
as the Sigma has field curvature that is quite noticeable when shooting flat
surfaces at under 6 feet. With apertures narrower than f/2.8 there's little to
no reason (bokeh, perhaps) to go with the Sigma. The Canon gives excellent
performance. But if you need f/1.4 and you need it sharp with great contrast,
a good copy of the Sigma will be hard to go without.
While I know negative attention is always disproportionately reported on internet forums....as such I thought I'd take the time to voice my experience.
The AF on my Sigma 50/1.4 appears to be accurate as either my 85L, 35L or 24L for that matter. I tend to do a lot of close portrait shots, which on a 50mm at f1.4 is around a half inch DOF. Subject or photographer swaying a smidge can make a photo a toss away. Shooting wide with my 85L, I've gotten in the habit of doing pre-focus prior to a final shot. I find my keeper rate due to AF is significantly higher. Perhaps that carry over on the Sigma 50/1.4 may be part of my success.... or maybe I just got lucky with a good copy. I have a theory on why pre-focusing helps. One is that if subject of photographer is swaying, the lag time to focus and trigger is reduced if the focus is near, making the chance that swaying will take subject and photographer out of that 1/2 inch DOF, smaller. Secondly, AF sensor reading taken closer to actual in-focus may be more accurate than further off. This one is speculation, but sounds plausible to me.
Ok, back to the Sigma 50 observations... Lateral CA is greatly reduced over the 85L. At f1.4, I believe the Sigma may do better in that regard than the 35L as well. Color and contrast is excellent at f1.4. I think even slightly better than the 35L at that aperature, though I want to do some more comparison testing before firming up that opinion.
Not that color tone matters much with dSLR's and AWB, but given that some Sigma's in that past (such as my former Sig 28/1.8) had a warm tone, it might be worth mentioning the 50/1.4 to be as neutral as I can tell.
Regarding center and edge sharpness, bokeh etc.... I'd probably just be repeating what has been already said a dozen times here. It's beautiful.
One_DaveT wrote:
(...)
The AF on my Sigma 50/1.4 appears to be accurate as either my 85L, 35L or 24L for that matter. I tend to do a lot of close portrait shots, which on a 50mm at f1.4 is around a half inch DOF. Subject or photographer swaying a smidge can make a photo a toss away.
Hi Dave,
I saw your post at DPR and replied there as well.
AS this is the third or maybe forth Sigma claimed to work properly I get a little curious. Do you mind telling us what camera you use and in what serial number range your Siggy50 is?
One_DaveT wrote:
The AF on my Sigma 50/1.4 appears to be accurate as either my 85L, 35L or 24L for that matter. I tend to do a lot of close portrait shots, which on a 50mm at f1.4 is around a half inch DOF. Subject or photographer swaying a smidge can make a photo a toss away.
They are all accurate AF 6feet and closer. mine was just fine..but as soon as you get to around the 10foot to infinity point the AF gets all over the place.
I am really enjoying my new Sigma 50/1.4 but I also just realized that I have yet to use it on AF. Maybe I am biased as I own so many MF only lenses and could like a lens even if it did not have AF.
Jonas B wrote:
Hi Dave,
I saw your post at DPR and replied there as well.
AS this is the third or maybe forth Sigma claimed to work properly I get a little curious. Do you mind telling us what camera you use and in what serial number range your Siggy50 is?
I've been shooting with a 30D. The Sigma serial number is 107xxx.
Just got my calibrated 50/1.4 back from Sigma. Having run more short/mid/long range tests for AF accuracy, I can only pity the Sigma engineer given the job of calibration! If you micro-adjust the 1Ds III for short-range accuracy, mid- and long-range accuracy is compromised – and vice versa.
My Mark III favours a Sigma 50/1.4 setting of about +10 for most working distances; I just have to remember that sub 50cm it will front-focus. However, the 50L displays similar behaviour: it's not possible to calibrate that for perfect near- and far-distance accuracy, as is well documented. It is slightly more accurate than the Sigma, though.
In contrast, the Canon 50/1.4 only needs a simply global correction to be good for all distances. On the downside, the colour and bokeh aren't as attractive and the vignetting is much worse. On the plus side, it is smaller and cheaper.
In many ways, then, the DG EX 50/1.4 looks like Sigma's attempt to copy the 50L at a cut down price, and not as you might have thought, their take on Canon's plain vanilla fast 50.
Mine seems to be generally good at distances greater than 4m or so.
I have seen some backfocusing in that range, but it's not consistent, and
I've also seen my Canon 50/1.4 do the same thing.
Matching what hubsand is seeing, my Sigma seems good at long distance with
a small degree of front focus tendency at short distance (around 6 feet and under).
I've seen nothing of the consistent back focus at long distances that some have
reported. Based on what hubsand has reported, I'm leaving well enough alone.
trumpet_guy wrote:
(...)
I've seen nothing of the consistent back focus at long distances that some have
reported. Based on what hubsand has reported, I'm leaving well enough alone.
Keep the reports coming....
Not a report but a thought: Sigma uses a distance of 40x the FL when adjusting focusing errors. If your lens were adjusted to the best a Sigma technician can do it would focus more or less exactly right at 2m. The front focusing would be gone. Then what would happen at longer distances... ?
One_DaveT wrote:
I've been shooting with a 30D. The Sigma serial number is 107xxx.
Thank you.
I have surely missed something but I think we now jave three users reporting lenses working flawlessly; one in the 1003*** and two in the 107*** (or should it be 1007*** ?) range.
I don't know if this means anything, especially as some first reporting having good working lenses later have tested again and found the long distance deviation.
To me this start to sound like a mechanical problem of some sort. Maybe Sigma can sort it out after all.
Good point. According to what hubsand has seen, perfect calibration at 2m would
probably lead to back focus at longer distances. And if this is the case, it would
be much harder to live with. It is not too difficult to adjust for a minor front focus
at close distance. You just sway your body slightly (we're talking a few cm here)
after focus lock and then you shoot. For problems with distance shooting, you have
to constantly be tweaking the focus ring.
I've started putting up my test results on my pbase site, and should be able to get
the rest of it up there tonight or tomorrow.
I have surely missed something but I think we now jave three users reporting lenses working flawlessly; one in the 1003*** and two in the 107*** (or should it be 1007*** ?) range.
I don't know if this means anything, especially as some first reporting having good working lenses later have tested again and found the long distance deviation.
Well I may just fall into that category. When I initially reported, I had been mostly shooting in the < 10 foot range with just a small few beyond. As I reported, these seemed fine. Having read the responses, I did several rounds of distance test shooting (10,20,30,40 feet). I found many of these to have a back focus aproximately equal to the distance of the DOF at f1.4.
Regarding serial number, Yes, I meant 1007***.
fwiw, I bought a used Sigma 28/1.8 two years ago. It had consistent and clear AF miss (front focus if I remember correctly). Even though it was not under warranty, I sent it to Sigma. It came back with an impressive dead on AF. No charge to me, and turn around time was a week + shipping time I believe. Seems to me that a Sigma 50/1.4 with dead on AF would be sine pari.
Putting this into perspective, this particular use case which may have an issue is an odd one for my style of shooting. I don't usually shoot at a distance with a 50mm, and when I do for a landscape, I would likely have the lens stopped down a bit. If I'm trying to close down the DOF, blurring out the foreground and background, something with a longer focal length like a 135/2 or 200/2.8 might be better suited.... just my preference.
