Great Bustard wrote:
Abu Mahendra wrote:
Lee Jay wrote:
Since I went all-crop, I traded my 35/1.4L for the Sigma 18-35/1.8.
I have two complaints about it:
- The zoom ring turns the "wrong" way (not the Canon way).
- I had some difficulty calibrating it, and had to buy the Sigma USB dock to make it happen. Even then, it took a while to narrow in on the right numbers. Seems okay now.
I'll get used to the first one, and the second one seems to be history now.
I have to say, it's really nice to have a "prime" that has a zoom range! My one and only issue with my 35/1.4L was the lack of focal length range. Spectacular lens in every other way.
I bougt the Sigma 18-35 Art in August 2013. Great optics, rubbish AF. Divested myself of that liability recently. Sigma, no more. In the articulate words of George W Bush, "Fool me once, shame on you. Fool me twice...ahem...uh...i won't get fooled again."
There are people who have had problems with Sigma's AF. However:
http://www.lenstip.com/374.10-Lens_review-Sigma_A_18-35_mm_f_1.8_DC_HSM__Autofocus.html
When it comes to the accuracy of the autofocus, our assessment is very positive. The number of misses in studio conditions reached less than 4% which is a very good result. What’s interesting, most of these misses were registered at 18 mm.
We didn’t notice any back or front focus tendencies and that result didn’t change with the zooming. You can see it in photos below where, both at 24 mm focal length and at 35 mm the depth of fields encompasses nicely the middle of the scale. In order to be meticulous it is worth adding that the 18 mm focal length diverges from this trend. In that case, in order to get the sharpest images possible we had to microcalibrate the lens by -2 units. The difference was visible in measurements but in real life photos it would be difficult to notice.
http://www.lenstip.com/359.10-Lens_review-Sigma_A_35_mm_f_1.4_DG_HSM_Autofocus.html
When it comes to the speed it depends on the working conditions - if there is not much light running through the whole distance scale can take even 1 second. When it is lighter that time shortens to 0.6-0.7 of a second. What’s interesting, the expensive Nikkor 1.4/35 didn’t fare better here. Actually the elderly Canon 35L, which can be even two times faster that the Sigma or the Nikkor, remains an example to be followed.
We don’t have any reservations about the accuracy of the autofocus. The number of misses in studio conditions reached 2.5% which is a very good result. On both bodies, used in the test (the 50D and the 1Ds MkIII) we haven’t noticed any back or front focus tendency which can be also observed in sample photos below.
http://www.lenstip.com/430.10-Lens_review-Sigma_A_24_mm_f_1.4_DG_HSM_Autofocus.html
Still the speed of the mechanism is something you can complain about. Even in good lighting conditions running through the whole distance scale and setting the focus can take over one second. It also seems that running from the infinity to the minimum focusing distance takes a bit less than the other way round. Both the Canon and the Nikkor were almost two times faster and in the case of the Canon its ring turns through as much as 160 degrees, having definitely a bigger range than 90 degrees of the Sigma ring.
When it comes to the accuracy of the mechanism we don’t have any serious reservations. The average number of misses in studio conditions didn’t exceed 5%. During the outdoor photo session the Sigma coupled with the 5D MkIII missed more often so we decided to perform another test. We put the lens on a tripod in our studio, in front of our biggest resolution testing chart and took 50 photos by f/2.8 changing the focus each time. The highest vertical MTF50 value amounted to 2112 LWPH and the lowest was 1968 LWPH; the median amounted to 2048 LWPH. The worst result we got was just 7% or so worse than the best so you can assume the repeatability of the autofocus settings is very good.
http://www.lenstip.com/400.10-Lens_review-Sigma_A_50_mm_f_1.4_DG_HSM_Autofocus.html
The speed is good in most of cases – running through the whole distance range with the focus confirmation take less than 0.5 of a second. What’s interesting, from time to time that result can be prolonged by almost another 0.5 of a second without any change of lighting conditions.
The autofocus performance depends strongly on the camera body. When the lens was attached to the Canon 50D and the 5D MkIII the whole set was pretty much predictable, with the number of misses reaching 6-7% which is good but also only a good result. The cooperation with the 1Ds MkIII was strange for a turn. The lens, even if positioned in a stable way in the same place before the testing chart, could oscillate for a short spells, jumping between two close focus positions several times; then it chose one of them but in a rather random way. More often than not the chosen position was correct but there were a lot of misses as well. The fact that the lens had also a pronounced front focus tendency didn’t help – once again you could observe it only when it was attached to the 1 Ds MkIII. While cooperating with the 50D and 5D MkIII the Sigma didn’t have such problems.
The main complaints are about AF consistency after long periods of use. I have friends which can't get through a wedding with consistent AF with the lens calibrated before the wedding.
Regarding your poll, I'd take the 35 IS because it's sharp and fast enough and IS is a huge benefit for video work.
