16-80mm initial thoughts and sample images

Started Apr 1, 2007 | Discussions thread
ForumParentFirstPreviousNextNext unread
Flat view
David Kilpatrick
Veteran MemberPosts: 5,436
16-80mm initial thoughts and sample images
Apr 1, 2007

I took the 16-80mm for a walk yesterday on the A100 and got a couple of dozen fileable images from a location which was a gap in my library, Coldstream right on the Scottish/English border. Conditions were not ideal, with weak sun (slightly hazy sky) and 80 per cent cloud cover occasionally breaking up to allow a bit of sunshine. It was cold and windy as well. I worked entirely hand-held with SSS, at ISO, using P for the most part with occasional program shift or exposure over-ride.

General points: the 16-80mm as delivered does not quite have the solid, smooth feel of the first prototypes to the zoom mechanism. When focusing, the image can make a lateral jump in the finder, depending on focal length and the direction of focus travel. It appears to stabilise very accurate and some kind of self-adjusting centering mechanism may be built in to the internal focus. Either that, or it's a very poor design! When firing, both the zoom and focus barrel can transmit a small jolt to the hand, which is surprising. It doesn't happen all the time, but indicates that the aperture mechanism is in contact or runs through other parts. Focusing is quite noisy, and not especially fast; the high contrast of the lens causes problems with wide area focusing, and the A100 would jump from sensor to sensor in many scenes - a case of 'too much information'.

The biggest negative issue is exposure. Using matrix metering or centre-weighted (I ended up using centre weighted as it seemed less affected) the 16-80mm can deliver between 1 and 2 stops underexposure in many everyday situations, something we have already learned to live with in the 11-18mm wide zoom. I was not attempting to correct this on my test shoot, so my sample pictures have received between 1.45 and 1.9 stops PLUS correction when converted using Adobe Camera Raw. The JPEGs are very dark. The A100 tends to underexpose by about 1/3rd stop compared to Canon 400D, and one full stop compared to Nikon D80 or Pentax K10D (though the K10D images are darker - it has a lower effective ISO). So the lens factor is probably about 1 stop. It's worst at 16mm, but affects the entire range. It is possible that calibration may vary from sample to sample, or with different camera bodies. The good side is that close-ups and views without sky, macro, still life etc were generally perfectly exposed (subjects which would normally call for -1 compensation).

Because of the correction for under exposure, the three tests posted here as full size JPEGs Level 12 show a level of noise equal to shooting at 320-400, not 100 (this is what happens when you brighten images from RAW). They also use ACR default sharpening 25, and a Strong tone curve because the conditions needed extra contrast (both these controls boost noise). At 16mm, the lens required -30 chromatic aberration correction; at 50mm, it required none; at 80mm, it required +20. These corrections completely eliminated the very sharp chromatic fringes. It has no visible mechanical vignetting, just the level of fall off expected at 16mm for well-known physical reasons with rectilinear wide angles. From these and other tests the optimum sharpness of the 16-80mm seems to be in the 35-80mm range.

I did not apply Luminance or Chroma NR, but I played with the sliders, and found that Luminance NR works unusually well with the 16-80mm as the software detects sharp edges and leaves them untouched. Because the edges are so well defined, Luminance NR actually works, hardly softening the overall image at all. For the same reason Chroma NR works badly, removing all colour from many small objects, because the Zeiss lens picks up pixel level variations in colour caused by textures and surfaces. ACR mistakes this for Chroma noise, and converts these to monochrome. It did this with the subtle coloured rendered houses in this view. On several subjects I found that the 16-80mm produced colour moiré which I have not seen often from the A100 - distant car grilles, mesh fences etc. What this means is that the lens resolution is exceeding the sensor resolution by a high margin.

16mm view - orginal full size http://www.pbase.com/davidkilpatrick/image/76472495

50mm view - original full size http://www.pbase.com/davidkilpatrick/image/76472585

80mm view - original full size http://www.pbase.com/davidkilpatrick/image/76472684

Focusing may be very slightly front-focus, which could be the camera or the lens. The exceptional sharpness of the 16-80mm makes it possible to see slight focus differences which would never normally be apparent.

Because of the noise level of these tests, and the detail of parts of the image, they make very large JPEGs at level 12 (7 megabytes in one case). Even with a low-noise, perfectly exposed ISO 100 shot I would expect a detailed image from the 16-80mm to be 50 per cent larger in compressed JPEG size, maximum quality, than you are used to. This also means - since the Sony raw file format is compressed too - you will get less shots on your card with the 16-80mm Zeiss! More information per image, less images per gig. That's a trade off I am happy to make.


ForumParentFirstPreviousNextNext unread
Flat view
ForumParentFirstPreviousNextNext unread
Keyboard shortcuts:
FForum PPrevious NNext WNext unread UUpvote SSubscribe RReply QQuote BBookmark post MMy threads
Color scheme? Blue / Yellow