Welcome to the next installment of my review of the new Tamron SP AF 17-50mm 2.8 XR Di II VC LD Asp IF. Today I look into the “VC” part and talk about the consequences of having a stabilized lens in that focal range.
This lens is not meant for full-frame (FX) cameras. On my Nikon D300 it is equivalent to 25.5-75mm, that’s roughly the same as the familiar 24-70/2.8 on FX. Canon covers both ranges with stabilized lenses, Nikon covers them as well, but without stabilization. The Nikon AF-S DX 17-55/2.8 is generally regarded as a fine lens, though slightly overpriced, and the Nikon AF-S 24-70/2.8 G, the new version that was introduced along with the D3, is frequently touted the best 24-70 on the planet, even though it is not stabilized. This is the field where the new Tamron tries to compete.
We have already seen quite satisfiable optical qualities. I don’t have any of the Canon lenses to compare, I have neither of the two Nikon lenses and no Nikon FX body to use the Nikon 24-70 as it was meant to be used, thus I have a hard time putting this new Tamron into reference. From what I’ve seen so far, and from images taken with the Nikon 17-55 that I’ve seen online, I suppose that the Nikon is better as regards corner sharpness wide open, and certainly as regards distortions. The distortions on the Tamron are more complex, of the mustache type, and therefore not completely correctable in Photoshop, although specialized applications like PTLens can do it, once they support the lens. Again, from what I’ve heard, the Tamron is clearly no match for the Nikon 24-70 on an FX camera.
That is all well. The Nikon 17-55 costs twice as much, the Nikon 24-70 even almost three times. They are well supposed to be better, and as is so often the case, you don’t get twice the performance for twice the money. The question is, how does Vibration Control, Tamron’s term for what Canon calls Image Stabilization and Nikon calls Vibration Reduction, change the overall verdict?
We have to look into two questions: The first is, does it work? The answer is a resounding YES. It works, and I’d say it works at least as well as Nikon’s VR, maybe even a little better, but that may be my subjective impression, founded in nothing but my satisfaction with the present purchase.
Well, just take a look at the technical data. Btw, you can do that yourself. All the images on my new blog have intact EXIF data. Firefox does not show it, of course Internet Explorer doesn’t either, and while some browsers like Opera do so, you can always save the image and open it in Photoshop or any other image processing program.
Enough of that. This image was taken at ISO 3200, f2.8 and 1/2s. Handheld. Sure, it is not completely sharp, I could have very likely got a better result out of a burst (hehe, a “burst” at 1/2s :)) of three or five images, and maybe even trying one more time would have sufficed. This is just a single, first, half-hearted attempt – and it is pretty enough for web use. I wouldn’t print this big (especially not this), but for documentary purposes even this image is OK. Or maybe not, because in reality it was so much darker, that this is rather forensic research than a documentary image 🙂
So, this is the horizon. Obviously it works and it works well. So far I have taken images at 1/8s and 50mm. I can’t hold this all of the time, but maybe every other time. 1/15s is what I have set as limit before the camera automatically raises ISO. That’s what I feel confident to be able to hold almost always at 50mm and always at 17mm.
The other question is: Do we need stabilization in such a wide lens? In general the answer to that question depends on who you ask. Nikon users tend to deny it, everybody else tends to acknowledge it. Basically: if you can have it, you want it, if you can’t, you pretend you don’t. There are exceptions, but go to any camera forum, ask the question, I bet it’s mostly along these lines.
Now that Nikon users can have it as well, how is it really? Is it worth the premium? The stabilized version of this lens costs almost twice as much as the non-stabilized.
As so often, it depends. For instance on what you photograph and when and how you do it. Stabilization is useless for sports. You want to photograph players, not moving ghosts. It is a mixed situation in the wedding market, stabilization may work well during the ceremony within a dark church (that’s exactly where you need it most), it won’t work for dancing couples. But then, you can always turn it off.
Some landscape photographers work principally from the tripod. Obviously stabilization is wasted on them. For my own uses, photographing on the street, always without tripod, not caring about motion blur, to the contrary, using it as an artistic opportunity, for me image stabilization is tremendous progress. So far I had it only on the Nikon 18-200 VR and on the Nikon 70-300 VR, direly needing it in the long focal ranges, now I have it in this very useful range on a constant f2.8 zoom and I am exalted.
Would I change to a D3s and the non-stabilized Nikon 24-70? Sure! Now that the D3s has sensor cleaning like every other DSLR on the planet (with the exception of the D3X, but that’s only a cheap one, isn’t it?), I would gladly take the superior sensor and give up stabilization, at least if I had the money to burn.
Would I exchange lenses with someone offering me the twice as expensive and optically slightly better Nikon 17-55? Never!!!
That’s it for stabilization so far. I probably should look into the question of panning, because this lens does not have two different stabilization modes like the Nikon lenses do, one for panning, one for total stabilization. Maybe with this lens VC must be turned off in order for panning to work, but maybe Tamron applies some magic and this is not necessary. I’ll have to try to find out.
As regards the images, the Image of the Day was taken at f2.8, ISO 3200, -0.7EV and 1/10s. I have removed a car in the background and the white license plate 🙂
The image with the street corner is straight from the camera. f2.8, ISO 1400, -0.7EV and 1/15s.