CZ 16-35 reviewed on PopPhoto

[NealS]

Just in case you were not aware of it, the July issue of Popular Photography has a review of the new CZ 16-35 zoom. It tests well, but does not seem to be a truly outstanding lens, not in the the Nikon 14-24 class, even at the shorter end of its focal length range.

As the other CZ lenses in the Alpha mount are so good, I am surprised that CZ did not aim higher with this lens.

Anyone know if more CZ lenses are due for the Alpha mount?

[David Kilpatrick]

I'm not surprised. It's slightly annoying to have used a pre-production model, sussed out exactly what level the lens was at (having already used the Nikon 14-24mm a few times), then being told by Sony to remove all pictures I had taken with the lens and not to publish any comments based on the sample. From all the results and reviews I have seen so far, the sample was normal and the comments would have been relevant.

My money went on a new Sigma 12-24mm, extending the range usefully, and keeping my 17-35mm KM D lens which is not as good as the new CZ, but not so much worse that pictures are being lost or spoiled as a result of using it.

David

[NealS]

Re-reading the review it would appear that quality-wise the lens is aimed at the Canon 16-35 Mark II and similar optics.

David, now that a review of the lens has been printed, would you mind summarising for us your impressions based on the sample you tested?

One more question - yes, I know it is like reading tea leaves - what exactly are Sony's intentions with respect to CZ optics? In the Yashica/Contax days a full line of CZ primes and zooms was available within a few years of the system being offered for sale. Since this is not the case with the Sony system, it would be good to know if it is Sony or CZ who decide which lenses will be offered for sale?

[David Kilpatrick]

Yes, I thought the lens was similar to the 16-35mm Mk II Canon- that is, not a technically accurate lens. The waveform distortion resembles the 16-80mm CZ on APS C at 16mm, and can be corrected almost perfectly by using the LensFixCI preset for the 16-80mm on A700, applied to the 16-35mm on A900. The corner falloff is similar to the Canon, and is caused more by a sudden deviation from a flat field (mapped by the waveform distortion).

Whenever a lens has a clearly identifiable 'moustache' distortion pattern, you can be fairly sure that the plane of focus is also not flat, but resembles a large ripple if you dropped a stone in water - imagine a single wave spreading out and frozen (OK, without the really splashy bit in the middle!). Or imagine one of those old bullseye glass windows. With the Canon 16-35mm II this is very noticeable, and results in zones of the picture being sharper than others. A combination of BF or FF with the Canon, and any misalignment of the sensor parallelism, creates some unfocusing which users find frustrating. I have helped two Alamy photographers using Canon diagnose a serious assembly fault on 5D MkII bodies because of the results from the 16-35mm which has a strongly cup-shaped focus field.

I don't think the CZ is quite as bad as the Canon, and it's nowhere near as bad as the original 16-35mm MkI. Why does this never show as badly on film? Easy - film always tends to bulge in the centre of the frame unless a special method like a vacuum or ceramic pressure plate is used to keep it flat. The mirror action sucks it into a slight curve even if it appears to lay flat when observed normally. The curvature of the field of these lenses coincides with slightly convex film, and they were originally adjusted to allow for this (MkI 16-35mm Canon). Digital lenses must be optimised for a strictly plane sensor, and that applies to the CZ or the Canon MkII. But they do not have perfect focus planes, which means the AF must be tweaked to average out across the field - and this gets worse the closer you focus.

We have to consider that the 20mm f/2.8 is similar (slightly better than the 16-35mm I think), the 17-35mm Tamron is similar, not so sharp, but you expect that at the price. The Nikon 14-24mm and the Olympus 7-14mm are both very different - they have a simple distortion not a compound wave-form, and relatively flat fields with just a hint of deviation to the corners. The Sigma 12-24mm is no worse than the 16-35mm and actually rather better at 16mm. But it's also f/5 at 16mm, almost two stops slower than the CZ 16-35, and if you stop the CZ down much of the issue with slight lack of field flatness disappears - masked by huge depth of field.

There are very few wide-angles which have been made with a good geometry and flat field. They include the 18mm f/4 Distagon, 18mm vintage Nikkor, and the 18mm Olympus for the OM series. The early 21mm f/4 Rokkor (mirror lock-up) and other Super Angulon derivatives are nearly perfect, same goes for various Biogon versions (essentially similar, Zeiss would argue theirs was the original). Minolta never made a retrofocus ultrawide with technically perfect characteristics, but the 21mm f/2.8 Rokkor was as close as you could hope for. Leica made no attempt to keep the 19mm f/2.8 up to the level of the 21mm.

I've specialised in wide angles since 1974, in all formats from 35mm to 5 x 4 - with a lot of special work on 6 x 7-9 formats. I've used SWC, Veriwide, Super Angulon 58mm XL, the unique Docter Optik 45mm f/4.5 to cover 5 x 4, the Plaubel Makina 55W, Fuji's 6 x 9 and 6 x 4.5 widest angle rangefinders, the 16mm Hologon on Contax G, and many other lenses. So far no zoom apart from the 14-24mm Nikkor has come close to matching what you get from a fixed focal length of modest maximum aperture.

David