A new precious sunny day (even though I needed to drive for a few hundreds kilometres to our French cousins' Riviera), a new test in the field. I concentrated again on the Sony Zeiss Vario-Tessar T* E 16-70mm F4, in the typical travel configuration: no tripod, not even a monopod, protection filter (B+W MRC UV Haze Filter 010M) always mounted, with a few noted exceptions. No memorable photos because most of them were taken in the harsh light hours of the day and with a sky mostly without clouds; even the sunset wasn't much interesting, and in any case I missed part of it because I ended up with the NEX-6 out of battery power (my fault, I already knew that the battery doesn't last for a whole day). But in the end I got another interesting round of tests; some confirmations (a few bad, others good), a couple of surprises, some new things to test in the next round.

This is the first time I went out with this lens with the attitude of taking landscape photos rather than taking test shots: so all the photos have been post-processed with my usual workflow in Lightroom; I only took care not to crop the vertical borders (with one noted exception) to keep the border analysis meaningful. For some photos it's possible to explore the 1:1 detail with a loupe, others can be maximized at full screen.

Some random shooting

At 70mm the lens confirmed my previous findings: sharp at the centre, sharp at the right border, soft at the left border. The photo below makes it possible to understand up to which extent softness affects the border: up to the building just right of the ship mast, let's say 1/5th of the width of the picture.

Then the lens was tested on a rolling boat, with a moderate amount of pitch and roll (the sea was apparently flat, but with the thing that in my town is called “long wave”). A first small surprise was a shot with some softness even at the right border — perhaps an effect of the stabiliser?

As a matter of facts, the stabiliser was useful, but can't be always trusted. A rolling boat, with its wide but slow motion, is probably not the kind of stuff for which lens stabilisation is optimized. Some bursts of three photos didn't make a single sharp shot; to get one I had to go up to seven/eight. It could have been a matter of auto-focusing too.

Off the boat and walking on the island of Saint-Honorat, I tried some shots with a high dynamic range, taking advantage of the harsh light of the sun and the shadow under the pinewood. The camera did a good job of preserving the details in shadows. The lens confirmed a very good sharpness at 35mm (probably its sweetest focal) from the centre to borders and up to corners. No visible defects, such as chromatic aberration, in the challenging areas such as shadowed branches against the bright background.

Same performance at 32mm — yes, the small light buoy at the right border is slightly soft, but I focused on the breaking waves at the centre, so it's probably a matter of DoF. The branches at the corner are still sharp, as well as the XII century fortified monastery.

At 28mm I still got a good performance near the right border, even though it's near the 24mm focal at which I've previously experienced the worst things. Again, the camera was pretty good in preserving the dynamic range of the scene, with shadows and a bright sky.

While in the past sessions “exposure to the right” using the live histogram always worked pretty good, this time I got some slight problem: while I never got burned highlights that couldn't be recovered in Lightroom, in some cases their rendition was not perfect and sometimes it required some further tweaking such as a local brush; for instance in the wave breaking area below. Probably in a bright day with harsh light it's better to be a bit more conservative and keep the histogram 1/3 EV far from the right side...

Occasionally a photo turned out to be just lightly softer than expected, including the centre. I suppose it was a fault of mine.

In some cases the lens stabilisation was really useful. For instance, I was able to take a decently sharp photo of one of the cloister in the monastery hand held, at 16mm and 1/10 sec. I think I've never been able, so far, to take such a photo at ISO 200 without a tripod. The trick was as usual to shoot a burst of photos and then pick the sharpest one. Thanks to the low ISO the NEX-6 was pretty good in delivering fine details in the recovered shadows. Using the monopod or boosting just a bit the ISO would deliver a pretty good sharpness. The photo was slightly cropped at the left border.

Another shot, out of an arrowslit, demonstrated again the good capability of the NEX-6 to deal with shadows: the wall details in the darkness were successfully recovered during post-processing.

While the border behaviour is repeatable most of the time, sometimes a small surprise appears... for instance, in the photo below at 19mm the left border is strangely soft (the problem at this focal should be at the right border).

At 20mm the lens confirmed a soft right border, while the left one was quite sharp.

At 16mm and with a strong backlight the default settings of Lightroom weren't able to completely remove the chromatic aberration in branches at the borders and corners: purple fringing can be seen.

Not friendly to the sun

The worst performance of the lens was shown during the return on the boat, when I took a few deliberate shots at the full sun, to confirm a problem that I saw in a previous session. Basically, this lens flares a lot (in this series of shots I removed the protection filter). Sure, flare is common in zooms, but my Nikkor 18-70mm ƒ/3.5-4.5G ED DX AF-S is much more capable of keeping it under control (see below). The Sony Zeiss produces a high amount of green blobs, quite hard to control. Their position and size dramatically changes by adjusting the focal and the aperture; they can go from a thing that totally ruins your photo...

... to a few, very small spots that could be easily removed in post-processing (e.g. with the dust spot removal tool). I really needed to disable lens stabilisation to prevent the green blobs from wandering around the frame without control (it's not a problem, because in this scenario you have plenty of light and the short shutter time alone avoids motion blur).

Between this two extrema the behaviour is more on the bad side than on the good one. Furthermore even in best cases there's a strange single diagonal streak of light around the sun, from the upper-left corner to the lower-bottom corner, that can't be avoided. I've never seen anything like that in all the other lenses I have, and it's quite a pity. Look, in comparison, a shot from the Nikkor 18-70mm ƒ/3.5-4.5G ED DX AF-S: just a few flare, very controlled, no streaks but pleasing sun rays (I purportedly stopped down to ƒ/22 for them) and I remember it was not too hard to find the optimal placement of the sun in the frame which mitigated the flare. This area has to be further investigated, with a side-by-side comparison of the two lenses.

While I've seen examples around the web where a partial masking of the sun did a good job in mitigating or even remove flare, it seems it's not as easy: I wasn't able to do much better than the thing you see in the photo below — that's a real pity, as after a few post-processing the tonal qualities turned out to be pretty good (the loss of contrast could be compensated).

So, shooting in face of the bright sun with this lens has to be further investigated. In a previous session I've noted that when the sun brightness is attenuated by some clouds, flares are less dramatic and more controllable. Perhaps the use of a moderate neutral filter could help? To be tested.

A disappointing sunset

The Massif of Estérel is known for his red porphyry that makes a pretty interesting chromatic contrast with the blue sea and sky, enhanced by the warm tone of the sunset. Unfortunately the sunset was disappointing because there weren't pretty, puffy clouds and the sky turned out to be pale blue (testing with a polariser was not planned for the session). Apart from this, I was not satisfied with the red colour of the rocks in some photos, such as the second image below. Perhaps this is another scenario where ETTR should be avoided... Lightroom could have recovered the three colour channels in different ways. Also, both images aren't as sharp as I'd like them to be (perhaps the stabilisation is to be blamed again). I made the mistake of not shooting side-by-side with the Nikon D5100, which could have helped in understanding whether there was an operator error. Stuff for a next round of tests.

Conclusion: a feral cat

The Sony Zeiss lens is a sort of feral cat in your garden: you like him and he can even give you some love back most of the time, but in an unpredictable way and mostly when he feels is in the mood; sometimes he might instead scratch you. But perhaps with knowledge and a bit of experience you can regain some of the control with him. Let's see...