Saturday 18 December 2010

Further comment on Carl Zeiss 85mm f1.4 Planar

I have been using the Zeiss 85 Planar lens for well over a year now and have established its strong and weak points, from a combination of practice and reading good quality articles.

It has the typical Zeiss rendition, which is neutral colour with a well-defned micro contrast and ability to separate a subject from a background.  The lens seems to tame scenes of extreme contrast and delivers a punchy result straight out of camera.  There is a quality in the background bokeh I haven't seen from any other posted example shots taken with an 85mm lens.  Its not as totally smooth and buttery as some choices all the time but it has a rounded smoothness that makes it different.  Note the rounded and soft specular highlights in the below image.

Image taken at f2
In this image taken at f4 the bokeh has changed slightly.  It all depends on the aperture,  camera to subject distance,  subject to background distance, brightness of the background, detail of the background.

Image taken at f4
It is certainly a great character of a lens that one has to get to know to make the best of.

Its qualities suit photography where one has the time to set up and focus accurately and compensate for its major weakness in real-world use....focus shift.  This actually makes it very tricky to use for its intended purpose as a portrait lens.  You think you have nailed accurate focus, which is hard enough anyway at wider apertures, only to find that the plane of sharp focus actually shifted rearwards off the eye when you took the shot.  This is because of residual spherical aberrations (RSA) or non-coinciding planes of focus between light rays passing through the outer and central areas of the lens elements.  You focus at F1.4 where the light rays converge slightly forward and then as the lens stops down to say f2.8 to take the shot the focal plane moves rearward.  You do learn to compensate for it but its calculated guesswork and hardly infallible.  Basically you need to focus bracket or have a subject who is willing to hold pose while you check results.  Therefore its hopeless for shooting faster moving stuff like kids or most sections of weddings, but fine where time is less of an issue.  The RSA and focus shift isn't unique to the Zeiss and is common in many fast primes, but this lens does exhibit these characteristics clearly.

Its unsuitability for weddings is a bit frustrating seeing as I am developing this area of my photography and an 85 f1.4 is an essential lens to get those special shots.  It is making me look at an autofocus alternative for those occasions when the Zeiss is effectively impossible to use with a satisfactory degree of reliability.  There are a number of choices and a lot of research to be done before deciding if it is worth it.  There is the venerable Nikon f1.4D, the new 84 f1.4G and an interesting 3rd party alternative from Sigma.

Meanwhile I love the Zeiss for a walkaround lens, shooting landscapes at wide apertures, where its great spatial separation adds a surreal air to images.   As on all Zeiss SLR lenses the long throw, well-damped focus ring is sheer joy.



Another main asset or weakness, depending on point of view, is its wide open performance.   Apparently the lens has been designed with deliberate RSA wide open at f1.4, which leads to a kind of halation or lack of contrast.  This can look really amazing if focus is nailed correctly and the rest of the image dissolves in unique blur.  There is a central area of high resolution and some features have a 3D embossed look to them.  It can be awesome for portrait work, but for landscape work its better to stop down to f2 where the character completely changes in one stop.  The above image was taken at f2 and has high contrast and sharpness in the obviously narrow focal plane.

If you look at the specs of many of the 85mm lenses from the main manufacturers, they do not tend to focus close and the Zeiss is no exception at 1 metre minimum.  Its simple planar design means it is poorly corrected close up, where image quality takes a dive.  If I want tighter portraits I simply pop it on a crop frame camera for that extra magnification.

In summary...a great portrait and landscape lens with a few quirks, but huge character.  When you spend this kind of money on a niche lens you know it isn't going to work for everything, just as a Ferrari won't do  the weekly shopping run very well, but take it on a track day and yay!  You certainly can't pop it on the camera and expect it will do a day's varied shooting easily, but when it all comes together you know why you want it.  The Macro Planar 100 f2 is probably a "better" technical lens overall as its performance is predictable, sharp and without quirks from wide open, but the 85 has a unique character and is a classic portrait length.

This sort of result reminds me of why I love the Zeiss 85 f1.4 Planar.

Image taken at f2 

Metering in the snow

Today was one of those rare days in southern England when it snowed!  We can wait years, but so far its happened twice this winter!  Go out and you will likely meet a number of keen photographers at well-known local spots.

For those experienced in using a camera in all sorts of conditions the fact that metering will be somewhat confused by snow is no surprise, but it may well catch out those newer to the hobby.

Its a good idea to have some idea what your particular camera meter will behave like when shooting snowy landscapes. While exposure errors can be fixed relatively easily in RAW files, its always more satisfying to get it right in-camera and pushing a file too far can increase noise textures and damage image quality.

Camera metering systems are designed to give a 0EV (exposure value) for a typical image that contains a variety and balance of tones between white and black.  This is what we would consider to be a normal exposure, not too dark or too light.  It is commonly referred to as 18% grey.

However, if a scene contains considerably more dark or light tones than "normal" the metering will try to compensate to achieve a 0EV exposure.  If the tones are dark the metering will tend to overexpose the image to bring the tonal values up and if the tones are light it will tend to underexpose the image to bring the tones down.  In snow we have a similar issue to the bride's white dress...there is an imbalance in the tones owing to the large amount of white, which causes the metering to underexpose the image.

The exact amount of underexposure in snowy conditions will depend on the quality of light and how much of the composition contains the snow or other tones, like sky, buildings or people.

Today I found that the quality of the light had a huge effect on metered exposure.

Images taken in shade required 1.3 stops of positive exposure compensation to get a histogram pushing nicely up to the right and a correct exposure.  This was one of those.


Images taken in direct sunlight only required about 0.3 stops of compensation, like the following one.  I am imagining that much of the difference is because when there is direct sun, the blue sky is a more neutral tonal value than the white or light grey cloud.


In most exposure modes, such as shutter or aperture priority, the photographer would dial in the appropriate amount of exposure compensation using the dedicated button and command dial or the camera menu.

Even if the photographer shoots in manual exposure mode the camera is still using the same metering algorithms as in any other mode.  It will therefore continue to show a neutral 0EV exposure on the viewfinder meter when the final image will actually be underexposed just as much as in any other exposure mode.  Rather than using exposure compensation (usually unavailable in manual mode) the photographer should dial in the required amount of "overexposure" above the 0EV value....in the case of the above examples, the same 1.3 or 0.3 stops.

With a basic understanding of what the meter is likely to do we can set the camera up as soon as we start shooting and make any minor adjustments much faster.  The advantage of digital is that we can take a test exposure, check the histogram and by the second shot we should have nailed correct exposure.