Tokina 50-135mm f/2.8 AT-X Lens Review

Very good, but with a fatal flaw

Copyright 2012 Thom Hogan

Lens Formula
18 elements in 14 groups; 3 SD elements.
Other Features
Manual focus override, internal focus, 67mm filter size. Comes with BH-671 hood, 9-blade aperture. Focuses to 3' (1m).
Size & Weight
About 5 " (135mm) long from mount (8.5" with hood!), 29.8 ounces (845g).
no longer available new

The Basics
Back in mid the middle of last decade we had two third parties step up to the plate with DX compatible telephoto zooms that had the "right" focal length adjustment to mimic a 70-200mm on a 35mm film SLR. Sigma had the 50-150mm f/2.8, while Tokina released their version of the Pentax 50-135mm f/2.8.

This review is long overdue, to the point where the lens was actually taken out of production before I finished it. While I've made numerous comments about the lens elsewhere on this site, this is the first time I've formally committed to a full commentary on it.

As noted, this is a Pentax design originally for the K-mount that was licensed and produced by Tokina for other mounts, specifically the Nikon DX mount. Unfortunately, the lens appeared just before Nikon made their move to cameras that required in-lens focus motors, so within a short period of time Tokina found themselves in a position where half of Nikon's DX DSLRs couldn't use this lens (no autofocus). Unfortunately, adding an in-lens motor isn't that simple with some lens designs, and I believe this lens died because of the complexity of that. A shame, because it shows some of what can be done in DX.

What this lens tries to accomplish is solve the "regular telephoto" zoom problem for DX. In film SLRs and FX DSLRs, the 70-200mm f/2.8 is one of the more popular lenses, and you'll find one in virtually every pro's kit. On DX, the 1.5x crop really makes that into a 105-300mm equivalent lens, which is okay for some things, but is a little long at the short end for many indoor uses.

If a DX camera body is scaled down from FX, then the 70-200mm equivalent lens should be, too. And that's exactly what the Tokina 50-135mm tries to do. It is a 75-203mm equivalent lens, which is pretty close. And it is scaled down. The Nikkor 70-200mm f/2.8 is 8" long and 54 ounces in weight. If we could scale that by 1.5x we'd have 5.3" long and 36 ounces in weight. Guess what? Tokina did even better: 5" long and about 30 ounces in weight. In the following shot, the Nikkor 70-200mm f/2.8 is on the right, the Tokina 50-135mm on the left.

Copyright 2012 Thom Hogan

Of course, to be truly equivalent you'd need a 50-135mm f/2, not f/2.8, which would push things back up, but DX is mostly about compromises, and I frankly I find I can live with the f/2.8.

This lens is old-school, meaning it requires an in-camera focusing motor. In the recent DX lineup, that would include the D2h(s), D2x(s), D200, D300, D300s, D80, D90, and D7000, all higher end cameras. It doesn't include the D40, D40x, D60, D3000, D3100, D3200, D5000, or D5100.

This lens does not have optical stabilization, another likely factor in its demise and one of the reason's why I rate it only average in features. There's a focus distance window on the lens, but no DOF or IR markings. The lens itself is only marked at 50mm, 70mm, 85mm, 100mm, and 135mm for zoom, so if you're looking for another specific focal length to match you'll be guessing.

Like most traditional lenses, the focus ring is furthest from the camera, the zoom ring closest to the camera.

The 50-135mm is an internally focusing lens, so the front element doesn't rotate during zoom or focus. Close focus is about 3' (1m), giving a maximum reproduction ratio of about 1:6. The lens uses 67mm filters. There's a rotating tripod mount on the lens, and it is snout and very dramatically unprone to vibration (why can't Nikon make such a mount?).

The lens is made in Japan.

While this is a largish lens, it's not nearly as large and heavy as mounting a real 70-200mm f/2.8 on the front of your DX DSLR. Indeed, that's sort of the point: it's nicely scaled to the DX size, and you don't feel like you've got a whale on the hook at the front of the camera.

The zoom ring on my sample is silky smooth. The focus ring uses the push/pull method of taking the lens out of autofocus into manual focus, and it's a very sure mechanism: there's no imprecision in snapping between one and the other. For manual focus, the ring is a relatively smooth, but you can feel and hear the engagement with the lens elements (i.e., it has a bit of drag to it compared to the free-wheeling of the ring in autofocus mode). It takes a half rotation to move from near focus to far, which is nice if you're trying to do precise manual focus wide open.

The butterfly hood (supplied) is deep. Indeed, it adds a full 3.5" to the front of the lens. As you'll see, there's a reason for this, and it actually doesn't fully solve the problem. The drawback from a handling standpoint is twofold: the lens becomes quite long with the hood mounted, making you more prone to hitting things, and when reversed on the lens, it covers enough of the lens that you'll probably refrain from shooting that way, even in a rush.

Overall, the handling of this lens is quite good. It feels very old school in its build and precision, and I barely notice that I'm zooming and focusing manually.

Autofocus: Bzzt. Bzzzt. Bzzzzzzt. Yes, this lens is noisy. It'll wake you up if you're falling asleep. Like all screw-driven autofocus lenses, some of the autofocus performance is dependent upon the camera. The pro cameras (in DX that would be the D1 and D2 series), have stronger motors and snap the focus a bit faster. The consumer cameras (D80, D90, D7000) have weaker motors and you can often see the focus proceed slowly in very low light, especially with off center focus sensors. The prosumer cameras (D200 and D300) would be somewhere in between. Note the long focus ring throw I described in Handling: that's an indication that the focus elements have to move a long distance to go from near focus to infinity. So some of your impression of focus performance will be determined a bit by how far you've moving focus at any given time. In general, with cross-type sensors and not a lot of distance variation, you should find the focus performance quite good, even on the consumer cameras. With the line-type sensors, in low light, or if you're going from 3' to infinity for focus, you'll find the speed more leisurely. Interestingly, I found that this lens doesn't do a lot of hunting, a common trait of screw-driven lenses. That's probably partly due to the telephoto focal lengths it covers, and partly due to the high level of positioning the focus elements can be put to. Whoever designed this aspect of the lens did a good job of keeping the screw-drive characteristics to a minimum.

Sharpness: One thing I've come to expect from Tokina is sharp lenses, and in the central area, this lens is no exception. Center sharpness is excellent at f/4 and f/5.6 at any focal length. Indeed, near the maximum of what I can obtain out of a D7000.

So let's talk about wide open, which is where a lot of you will want to use this lens: still very good, and even excellent in the mid-range of the zoom. 135mm is slightly sharper in the central region wide open than 50mm, which is pretty much the way I'd want it.

You'll note that I've been talking about the central region. It's the corners that will disappoint you, if anything does about the sharpness. At 50mm, the corners are what I'd call good and never get above very good. At 135mm, the corners are pretty poor wide open, but get just barely into what I'd call good range at f/4 and beyond.

Overall, I'd say this is a very good lens optically. That might surprise you from what I just wrote about corner performance. Unless you're using this lens for landscapes (and even then if you stop down), the types of subjects you're likely using a telephoto zoom for don't typically suffer when a lens has soft corners. Indeed, some of us actually like that effect for sports, people, and other types of telephoto photography.

Let's go to the corners and see what I mean. First, let me give you the full view (this is off a 6mp DSLR, by the way, but I retested and see the same thing off my D7000):

Copyright 2007 Thom Hogan

As you can probably tell, pretty crisp. Let's go out towards the edge and see what we see in actual pixels at f/5.6 and 135mm:

Copyright 2007 Thom Hogan

Yeah, there's a bit of softness there, but not terribly bad. Remember, this is 6mp, so some of what you're seeing is lack of resolution in the first place.

So again, if you absolutely need sharp corners, this might not be the lens for you, especially at 135mm. But you might find that it suffices.

Light falloff: Surprisingly little. Never more than a half stop, and completely under control at 50mm f/4 and longer focal lengths at f/5.6. I was actually curious about this, so I mounted the lens on an FX camera to look more closely at the image circle. It is abrupt. Often with DX lenses you see the image circle go from full coverage to a fade to black over a fairly large area. Not so on this lens other than perhaps 135mm. I'm not quite sure how Tokina achieved that, but vignetting isn't something I'm going to worry about with this lens. What little there is can be easily corrected.

Chromatic aberration: Not zero, but not high. The two ends (50mm and 135mm) have higher chromatic aberration than the middle, but frankly, not nearly as much as I'd expect. Moreover, the characteristics don't change much at smaller apertures, which is a little unusual for a telephoto lens. On a D7000 there might be two pixels of CA fringing on edges, but that's relatively on the low side and easily corrected by most DSLRs (and certainly in post processing). At mid-range focal lengths, there's not enough CA to get excited about. By the way, I did not correct CA on the above examples.

Flare: So far I've been mostly praising this lens, but here comes the fatal flaw. And I mean fatal. Consider this image, taken with the lens hood on:

Copyright 2007 Thom Hogan

Yeah, 60's Hollywood bad. This, by the way is 90mm f/6.3. I can't recall another lens I've used in the last decade that would do this at f/6.3. You'll note that the light source causing this problem is not in the scene. This is the full frame, not a crop. This is a lens you must use the lens hood with.

This flaw was the real reason why I stopped using this lens. I'd be perfectly happy with the results shooting down sun, then I'd swing up sun and get results like this (go ahead, count the aperture blades ;~).

Flare is a common issue with Tokina lenses, in my experience. But none of the others are like this.

Distortion: Modest. From about 1% barrel at 50mm to 1% pin cushion at 135mm. Much of the mid-range doesn't really need correction, the ends need only minor correction. Frankly, better than my 70-200mm f/2.8 ;~).

Bokeh: This lens isn't going to win any bokeh contests. There's just enough longitudinal aberration wide open that you get the dreaded color artifacts on the blur circles. This is mostly controlled by f/4, and fully controlled by f/5.6. I wouldn't characterize the bokeh as bad, but it isn't excellent, either. If you're worried about it, stop down to f/4.

Overall, this is a very good lens marred by one big problem: flare. If you can accept that or shoot in ways to keep it from showing up, it's a very worthy addition to your lens kit. But if you don't like the 60's Hollywood look showing up in your contre jour shots, it's probably not a lens you want to own.


  • Corners go soft. Especially true at 135mm and wide open.
  • Flare potential. The nine blades of doom.
  • No VR, no AF-S. Lacking in the latest and greatest technologies.


  • About right. Right size, right weight, right optical ability.
  • Nice build. Built well (especially that tripod collar), and handles smoothly.
Quick Evaluation

I neither recommend nor not recommend this lens.


Almost, but not quite.


The 50-135mm lens has limitations you need to be aware of:

Converters: no converters are recommended (rear element very close to mount at all focal lengths).

G: Since this is a DX lens, it's not likely that you'd mount it on an older 35mm body that can't handle the lack of aperture ring or makes restrictions because it is missing.

Focus: Probably the reason why it's been discontinued: no internal focus motor means that it doesn't autofocus with the D40, D40x, D60, D3000, D3100, D3200, D5000, or D5100.

Coverage Area: produces a circular image vignette on FX bodies at all focal lengths except 135mm.

Version of Review:
10/8/2012: initial post. | Nikon | Gadgets | Writing | imho | Travel | Privacy statement | contact Thom at

All material on is Copyright 2012 Thom Hogan. All rights reserved.
Unauthorized use of writing or photos published on this site is illegal, not to mention a bit of an ethical lapse. Please respect my rights.