I don't think it's conclusive, either -- I don't think any single comparison like this ever is.
I think "sample variation" is often a catch-all explanation that covers a lot of variety in technique and application -- but some degree of it seems likely to be in play with the 16mm. What I do know is that the 16, if you have a decent copy, is a very competent lens. It's a long way from perfect, and you obviously have to watch the corners very carefully, but edge performance stopped down is on par with all but a few of the best ultra wide DX zooms and the very best modern full frame wides. The Nikon and Tokina 12-24 wide zooms that many of us used for years on DX aren't really appreciably better (nor is the Sigma 10-20, though the 8-16 and Tokina 11-16 may be). Neither are stalwart FX primes like the 20/3.5, 20/2.8, or 24/2.8 Nikkors. To decisively beat the 16/2.8, you need to be in the class of the 16-35VR, 14-28/2.8, 24/1.4, the Zeiss wide primes, etc. Those lenses will clobber the 16, and they should -- their price and weight are roughly an order of magnitude higher than the 16. At the same effective focal length and pixel density, put images from a D3x with a 24/2.8 next to those from an NEX-7 with the 16/2.8. I'll bet a nickel that you'll find new respect for the Sony.
I'm not generally a jpeg shooter, but I recently shot my NEX-7 set to medium-fine jpegs, which generates 12mp files. A lot of us here would likely have bought into the NEX system if it were only a 12mp sensor, which was state of the art not long ago. The quality of the 16mm images at that setting, which is probably more than adequate for 95% of what most of us use the NEX system for, is exceptional. If you expect the 16 to be a premium lens competing with the best from Leica and Zeiss, you'll be disappointed; if you consider it a tiny snapshot companion, you'll giggle at how good it is.
To be honest, I found The Sony 16 mostly dissapointing for anything other than black and white shots. The colors weren't very good on my 5n and it wasn't that sharp close up. It Also had an astounding amount of vignetting. I sold mine when I found it color shifted on my nex7.
I dunno. I find them pretty easy to ignore, given that they're not "results" , but tests -- tests which don't match what everyone is getting with this lens. What's more important: a test, or a photograph (like the one linked above with direct comparisons to other lenses)?
As I've said a zillion times, there just aren't that many DX lenses that outperform the 16 at anything remotely like its size and weight (let alone price). No lens of that focal length will ever look as good as a shot from a 50mm prime -- or a good 28mm, for that matter. It's just not going to happen -- or at least it never has. If that's too hard to take, you're better off shooting a full frame camera (Leica M or Nikon FX, most likely).
When someone offers a general purpose moderate wide that shoots better than the 16, I'll take notice. The Sigma 19 is one of the more interesting alternatives to emerge, and I'll be considering it carefully.
At first blush, the Sigma seems to share some characteristics with the CV15, which has a lot of field curvature even stopped down. I think that contributes to the softer edges in some shots -- seems like the best way to get moderate distance shots in uniform focus is to focus on foreground edge elements, and let the distance fall within the DOF. Since I don't own either lens, that might either be easily manageable, or a fussy pain in the ass. I assume it's somewhere in between.
What I'd be even more interested in personally would be something like a $500-ish 14/4 with solid performance at f/4, and truly sharp to the corners by f/8. Until then, the 16 is fine.
Thanks! Using these small JPGs it makes it difficult to see the difference even though it's so obvious on my monitors. I like this lens a lot and if I was getting paid for this shot would use the 19 over the 16 every time.
Regarding Capture One and Lightroom, even though I love the colors out of Capture One, I think that in the end it does not resolve as well as Lightroom (either v3 or v4). The default sharpening of C1 is quite strong, and even though it might give more 'pop' to small details, I have the impression that I'm not getting the full resolution out of the 5N or my new NEX-7. I don't do landscapes, though, so maybe I'm wrong. Also, the color noise reduction (for high ISO) is terrible, it abruptly smears out all detail of anything with red textures... but did I mention I love the colors?
My Sigmarit 19 is coming in a couple of days, and I also have the Sony 16mm, so it will be cool to see how they compare. I shoot family stuff, so I don't need extreme resolution across the frame, nice rendering is key (thumbs-up for the Zeiss 24!). My copy of the 16mm is quite good I think, although I wish I could have an in camera 5:4 crop mode to automatically crop the very extreme borders.
Thanks for the info about capture one. That's good to know because I really don't want to buy a copy of it. I am finding that light room requires more care than I have been giving it as far as fine detail retention goes.Take a close look at the Sigmarit 19 when it comes, I had to send my first copy back because of softness on top third. This copy is phenomenal some very happy about it. He mentioned cropping out the borders on the 16 mm I have been doing that with the wide-angle adapter on the 16, seems to work pretty nice.
i pretty much ignore all tests that involve imatest and shooting test charts. there's just so many ways to screw them up. i just look at the best (technically) actual photo that somebody has posted at full resolution now to get an idea of what a lens is capable of.
the sony 16/2.8 will perform better at f/8 than the vast majority of full frame 24/2.8 lenses on a FF camera of similar MP. when you look at f/2.8 performance or at higher grade FF lenses the story changes.
my big issues with the sony 16mm are actually color and "look".
the sigmarit 19mm looks very interesting and is a better focal length for me, but it is also a bigger lens. it looks like it probably performs much better at f/2.8 than the sony, which could make it worthwhile combined with the focal length.
I personally like the 16mm for the NEX 5, but the Sigma looks like a good alternative. I think you just have to know how to get the best results. Infinity at f/8 is usable but rather disappointing for me, but anything closer than infinity it seems to do well at. Close up at f/2.8 in the center of the frame this lens shines. There aren't that many applications for it there, but it has been suitable for what I planned to use it for.
FlyPenFly wrote:
Can someone compare the kit lens to the Sigma 19mm at F4 and F8 at infinity on the NEX-7?
I'm curious if gaining half a stop is worth losing 20-55mm in focal length and OSS.
+1 I would also be curious to see that comparison in good light. My initial suspicion is that the Sigma color will be noticeably better. Not sure about the sharpness though. I have only used the kit lens on the NEX-5.
Sony coatings are pretty good for color though. I was never dissapointed with the color on the 18-55 on the 5N. The newest Sigma HSM lenses though seem to be using a much nicer coating than their old lenses.
