I still think they should be counted as one lens design rather than two as it is pretty clear that the lens was designed to have the option of VR capability from the start and they either put in the additional mechanism or not. Equally, the Canon f/4 versions of their f/2.8 lenses do not represent a build from scratch.

From what I see of the Nikon range, pretty much every lens is a stand-alone design designed and optimized for a specific purpose. (OK the PC lenses and the ultra-teles maybe have quite a lot in common). Canon seem to be looking to make more lens variations than Nikon just to make up the numbers.

@soap While plastic lenses can have great quality, there is a point where the weight of the glass in a lens is so much that the body needs to be metal to keep everything stiff.

I think that the more likely replacement for the 18-135 will be another 18-135. It is a perfectly reasonable size – 24-200 equivalent. If they were to bump the f stop number up to make it a f/3.5-4.5 with a 77mm filter it would be a very handy lens to have about and make a perfect compliment to a DX ‘magic three’ of the 10-24, 18-135 and 80-400 AFS. That gives you coverage from 109-4 degrees (15-600mm equivalent) in three lenses with matched 77mm filters (if you care about such things). If forced to leave home with only one lens the 18-135 will have you covered 90% of the time.

One article I read after writing the above pointed out that like soft corners, falloff is typically a benign effect. Trying out the 10-24 I found it absolutely impossible to compose a non-contrive test shot with interesting stuff in all four corners at 10mm. Falloff is easily corrected in software, so maybe avoiding falloff is not worth a bump up in filter size. If Nikon thinks that is the case, some folk will be waiting a long time for that 70-200 f/2.8 replacement.

In general I see the point of getting a Nikon over other brands is that the Nikon engineers have a much better idea what lenses to make and what features to support than the average bozo on the net (including me). Nobody saw the 35mm DX coming or the 10-24 DX, but they make absolutely perfect sense when you think about the market and will sell in much bigger numbers than the typical suggestions here or on dpreview.

Lens design is inevitably a compromise. You can’t have perfect sharpness, weight, speed, cost, bokeh and so on at the same time. But in general the choices Nikon makes make a good deal of sense to me. As I see it they are trying to optimize the features that are hardest to correct for in software.

I am pretty sure we will see a replacement for the 18-135 soon enough. It is a lens that makes a lot of sense as a kit lens for the cheaper DX models. It doesn’t have the range of the 18-200 but it doesn’t have the weight either. It may not be a stellar performer but it definitely isn’t aimed at this particular audience and the replacement will be another mass market lens.

The replacement will certainly not be a 24mm f/1.4.

Well, I think we do. The current 70-200 2.8VR is just too expensive. $1800-$1900 is not something most prosumers can afford. That leaves us to what else? An used 80-200 AF-D (the AF-S is very hard to find these days) for $800 where Canon’s non IS version with USM for that price. Many pros would argue that they would like a lighter F4 lens that has great IQs for their backup body while carrying the big one on the main body.

The whole thing is about competition. It’s ok to have different strategies, but ultimately, it’s what users want.

@Soap. only naive fools believe that zooms are ever actually constant in aperture, and metering was never their advantage, it was constant access to a decent aperture throughout the whole range, and the ability to zoom without adjusting settings. most constant aperture zooms are really variable within 1/2 stop, which frankly is good enough.

Be this true or not is not for me to say – but that is the perception which seems to be widely out there.

I would expect the VR lenses to be dropped if the D4000 supports VR.

And the fact that Canon is claiming 63 lenses to Nikons 55 is probably the explanation for why they sell four versions of the 70-200 rather than two. Meaningless multiplication of variations is what companies do when they are playing catch-up.

Go try a ‘DX’ zoom on a full frame camera and you will find that it actually covers the frame across at least 2/3 of the range. The 10-24DX works as a 14-24 FX.

Coverage is not the same as falloff. The 70-200 is performance optimized, not cost optimized. Avoiding falloff on full frame was not a major design goal, using a 77mm filter, size and weight were.It is a pretty safe bet that the falloff will be corrected in the next iteration. But that will at most mean stepping up by one filter size, not doubling the size of the lens.

If by ‘DX’ you mean cheap, well, obviously Nikon could make cheaper lenses with smaller f-numbers. But that is completely different. Canon’s f/4 telephoto lenses are full frame. And they can use hybrid aspherical elements instead of all glass and plenty of other cost-cutting measures. But that is nothing to do with the DX issue.

I don’t think its likely we will see an f/4 constant aperture lens range from Nikon either. Nikon clearly prefers the Variable Aperture design. They just launched a $900 Variable Aperture lens.

If you want small size, you want a VA design. Nikon makes f/3.5-4.5 zooms. What exactly is the benefit to f/4 all along?

Canon made those f/4 zooms because it was playing catch-up. So they recycled the f/2.8 design as a cheaper ‘lightweight’ alternative. Then they pushed out a bunch of marketing blather about the CA design being ‘professional’. Its all huey.

In terms of total lenses, there isn’t that much difference anymore: Nikon’s got 55 (including the supposedly discontinued 18-135) and Canon’s got 63. Within another 2-3 years, I think Nikon and Canon will have the same number of lenses in their lines, but they won’t be the same lenses.

One can make the case that the Canon line is overkill. Do we really need four versions of the 70-200: 2.8 and 4.0 with and without IS? Do we really need two 75-300 consumer zooms, with and without IS, when street price is only $40 apart? Of course, you can also ask whether Nikon needs both the VR and non-VR versions of the 18-55 and the 55-200. I think that as a bigger company, Canon is able to support all these models and different price points, but Nikon’s manufacturing seems a lot more limited. Up until the recession, Canon was always better than Nikon in terms of keeping lenses in retail stock. With the recession, even Canon has many lenses out of stock or on backorder.

Canon does seem to be paring their consumer line a bit. They’ve recently discontinued the 24-85mm, 28-90 III and 55-200 II. Before that, they discontinued the 18-55, 20-35, 28-90 II, 28-105, 28-80 II and 80-200 II. In the pro line, all recently discontinued lenses have been replaced by newer versions.

Nikon seems to be attempting not to have overlaps in FX pro zooms, so you have 14-24, 24-70, 70-200, 200-400. Cannon doesn’t seem to care about the overlaps as they have 16-35 (or 17-40), 24-70, 70-200, 100-400.

The DX line is different as Nikon doesn’t expect consumers to buy more than two zooms, so you have 18-XX and 55-200. Canon seems to have dedicated their APS line to wide angle zooms, except for the 18-200, which is obviously for people who want to buy only one lens.