Smiert Spionam wrote:
Looks like a very reasonable lens, and a bargain. I've got the 30/2.8 on the way, and looking forward to it.
From the crops I've seen (here and on blue skies' thread on DPR), I'm not sure the 19/2.8 offers much over the 16/2.8, though. The corners probably don't fall off as precipitously as the 16, though overall sharpness doesn't look any better, and might not be as good (stopped down, anyway -- the 16 is so weak at f/2.8 that I never shoot it there). I suspect the choice between the two would have more to do with focal length than absolute image quality....Show more →
Maybe that would also depend upon which NEX they are mounted. Seems like I recall many saying the 16 performed much better on the 5n than the previous NEX's. It may be that the 16 is poor on anything else, particularly the NEX-7 outside of the center- or have you used it on the NEX-7?
I sold my 16 a while back, even though I thought I had a decent copy of it, and I don't think it really compares to the Sigma 19, especially below f5.6, like you said. Granted, I usually cropped photos from the 16m to make it about 18mm-19mm, so maybe they'd be more comparable in that regard.
Tariq Gibran wrote:
Maybe that would also depend upon which NEX they are mounted. Seems like I recall many saying the 16 performed much better on the 5n than the previous NEX's. It may be that the 16 is poor on anything else, particularly the NEX-7 outside of the center- or have you used it on the NEX-7?
I use the 16mm on the NEX-7, and previously did on the 5n. For all practical purposes, it's fine on both (much worse on the original 3 & 5, apparently). F/2.8 is weak, but it's tack sharp in the center at f/4, very adequately sharp to the short edges by f/5.6, and best overall at f/8. It's not a perfect lens -- the corners only become really adequate by f/8, and overall sharpness never matches the center at any f/stop.
I've noticed color shift on the 16/2.8 on the NEX-7 wide open, though I never shoot it there. There's probably a little at f/4 or f/5.6, but it's slight enough that it is essentially invisible once you've done the automatic corrections for vignetting and distortion (in LR3/4, or the camera's own JPEG engine).
What I've seen from the 19/2.8 doesn't look dramatically better, though -- it's probably more consistent across the frame, but not as sharp as the 16 in the center, and with weaker edges.
That's not to say that the 19/2.8 isn't a fine lens -- it's almost certainly better wide open, and the focal length may be a better match for some people.
The Sony 16 is very sharp in the center, but its issues to me start around 5mm from the center. So, the middle third is great, but sharpness starts aiming downhill from there, and there isn't much evenness across the frame after than middle third. Then, of course, the extreme corners fall of a cliff.
I haven't had the 19 long enough to compare, but my initial impressions are that it is much more even across the frame.
The only thing that example shows me is that in the NEX7 shot with the Sigma, the shadows are blocked up but they are open with the Sony 16 in the bush detail. Thus, because of either exposure or light differences, it's meaningless as far as comparing like detail between both. That said, I have not shot with the 16 and have only seen terrible corner/edge results with it posted so I don't know personally how good it is.
Tariq Gibran wrote:
The only thing that example shows me is that in the NEX7 shot with the Sigma, the shadows are blocked up but they are open with the Sony 16 in the bush detail. Thus, because of either exposure or light differences, it's meaningless as far as comparing like detail between both. That said, I have not shot with the 16 and have only seen terrible corner/edge results with it posted so I don't know personally how good it is.
Maybe we're looking at different images -- or at least looking at them differently. What I see is that the Sigma doesn't match the 16 for either edge sharpness or CA control, on either the 7 or 5n. If it were simply an exposure difference, I'd expect the highlights in the 19mm shot to look better, not worse (as they do). With more exposure, the Sigma's obvious CA problems would be exacerbated, too.
Smiert Spionam wrote:
Maybe we're looking at different images -- or at least looking at them differently. What I see is that the Sigma doesn't match the 16 for either edge sharpness or CA control, on either the 7 or 5n. If it were simply an exposure difference, I'd expect the highlights in the 19mm shot to look better, not worse (as they do). With more exposure, the Sigma's obvious CA problems would be exacerbated, too.
If you are looking at the lower left crop from the trees and bushes in front of the house shot, then we are looking at the same thing. At least with the Sigma 19 I have here, there is not a lot of CA. Sigma's are known though for great sample variation.
The bushes do reveal the far corner softness of the 16, but look at the entire edge crop. Look at the tall tree just above the bushes -- it's hard for me to see the 16 as anything but superior there (a high contrast element right along the short edge of the sensor). It's worth noting that it's also better than the same crops from the CV21, which is also discussed as a superior alternative to the 16.
Sample variation (and more importantly, technique and focus accuracy) can always be an issue. But what is undeniable is that the 16mm is capable of quite decent performance out to the short edges -- and there's certainly nothing in the Sigma 19mm shot to suggest it's better (though it clearly outperforms the CV21 there). There may be other images out there that do show superior edge performance from the 19 -- but I have yet to see them.
A given shot that shows mediocre performance can always be dismissed, since technique errors and sample variation can make a lens seem to appear sub-par. The relatively weak performance of the 19, CV21, and CV35 in this sequence is not necessarily evidence of mediocrity, but it is most certainly a lack of evidence of excellence. What this particular sequence *does* show fairly convincingly, though, is that the 16mm is capable of excellent edge performance.
If the Sigma can beat it for stopped down edge performance, great -- I just haven't seen it.
Again, the exposure difference is too great to judge detail on the tree between the two lenses imo with the N7 shots. The sigma version is completely blocked up in the shadows to lower mid tones.
Not if you look at it full size -- the difference is fairly striking, if you ask me. If the problem is underexposure of the Sigma shot, why does it show not only poorer shadow detail, but poorer highlight detail and worse CA? You can't have it both ways.
Again, this is not evidence of the mediocrity of the Sigma lens (and even if this is typical performance for that lens, it's not bad). It's pretty clear evidence, though, that in this comparison, the 16/2.8 looks a lot better. It's not a subtle difference.
I'm not going to post images, since it's not my shot. A link to the full size version of the above image is available in the OP's thread on DPR, as well as the full images these crops came from.
I have to admit, I did not see where the full size sample was. His Sigma 19 looks really poor at the edge and the 16 he has looks better than anything I have seen thus far out of that lens.
Tariq Gibran wrote:
I have to admit, I did not see where the full size sample was. His Sigma 19 looks really poor at the edge and the 16 he has looks better than anything I have seen thus far out of that lens.
that looks about inline with what i see from my sony 16mm.
FlyPenFly wrote:
So this is surprising, the Sony 16mm is sharper across the frame at 5.6 and 8 than the new sigma?
I don't think it's conclusive. The Sigma 19 I have is pretty damn sharp across the majority of the frame. It's a little soft at the edges. I have decided to actually keep the Sig and NEX-7 based on it's performance. When I have more time, I will post more samples.
