Noise reduced: firmware update Sony A700

Good to have a photoclub, for I had not been paying attention to the Sony website lately: a new version of the firmware for the A700 was published already in September! Version 4 is meant to tackle my main issue with the A700: noise in high sensitivity settings. Or more completely (from Sony Europe):

* The choice to select auto exposure bracketing (single & continuous) with 3 shots in 2 EV steps has been added. [That is nice for HDR pictures! I'd like to start that discipline once.]
* The choice to turn [Off] the High ISO NR feature has been added. [Great! I want to be in control.]
* Improvement of the image quality in high ISO setting. [This is the main improvement for me]
* Improvement of the auto white balance and D-Range Optimizer performance. [One of my reasons for always shooting RAW is that the auto-white was not always to be trusted. I'll keep shooting RAW, though. I don't use the D-range optimiser; also a matter of wanting to control things myself.]
* Improvement of reliability for communication between camera body and vertical grip. [Don't have that thing.]

Of course I immediately made a quick try of higher ISO settings. My impression is that ISO 1600 is as good now as ISO 800 was: quite usable. There is a visible reduction in quality from ISO 1600 to 3200. See first set of compared photos. This (out-of-focus) detail from the top-left corner of a picture shows this effect.

The settings of the noise reduction do make a difference. My impression is that setting the noise reduction to 'high' is a big improvement for larger areas, but may make edges and shadows a bit too 'rough'. Look at the left side of the picture frame in the comparison of two photos below.


Thanks to Dieter for telling me about the firmware update!

Test, relevance and reputation

The November 2008 issue of Black & White Photography has a column in which Mike Johnston defends his way of testing cameras and lenses. He argues against trying to be completely objective: it is not the figures that matter, but the trial of the gear in dfficult, actual photography circumstances. I completely agree with Mike: too often testers focus on what is measurable rather than on what is relevant--not just in photography but also in my professional field of higher education. There, rankings are made of things that can be measured but whose relation to educational quality are uncertain. Similarly for photo gear: their 'performance' in laboratory circumstances is only a remote indication of what you can do with the thing in practice. In practice, all kinds of circumstances and especially your own needs are much more important than the 'objective' figures. You use different focal lengths, differnt distances, and handheld instead of on a tripod as they did in the test, or you take different types of pictures. For instance, my new standard zooom lens got fairly good test results, but with my preference for architectural pictures, the slight distortion in the wide-angle setting weighed more heavily than what I could read in the mags. Or take the high-quality lens I looked at: it was indeed great from a technical point of view, but way too heavy to be practical on a day-long hike. So let me quote Mike's conclusion: 'The bottom line is that observation is at least as important a method of inquiry as measurement is. It is no less relevant and no less reputable.'

Quality or Compromise?

When I bought my new camera, a few months ago, I had the firm plan to go for quality. No compromises! So I ordered a standard zoom lens with the camera that would give maximum quality. All of the web said so (for one example: click here). The shop where I had bought the body did not have the lens immediately. 'Next week', they said, and later it was 'In two week's time'. In this way, more than two months have now passed and still they did not have it. 'In a month from now', they promised today after another phone call. I had enough of them and decided to try another shop; all over the net you could find this lens, and quite a few shops said they had it available--not quite for the price of my first shop, but they had it. Well, they said they had it. I called a shop or two, but discovered that the web sites had been too optimistic and in fact they had run out of stock. The third shop did not answer the phone. But that one was not too far from my home, so after dinner I just drove there, à la bonne foi. Lo and behold! Konijnenberg had four of them on their shelves: bigh & beautiful Carl Zeiss 2.8/24-70 lenses! And would I like to try? No, I did not want to try, I wanted to run home with this rare beauty at once. Still, I did give it a try in the shop.
Two quick examples, made inside the shop give some impression of its quality. The first is in the wide-angle setting (24 mm), the second in tele (70 mm). For your inspection they are uploaded full-size (though in jpeg). Of course these are not real test pictures, but still: there is not even a hint of barrel or cushion type distortions in the lines near the margins of the picture. There is no chromatic aberration around the lamps on the ceiling and above the counter, in the corners of the picture. In short: all the high expectations from the rave reviews were made true.
But what the reviews had not told, was that this uncompromising beauty also did not make compromises on size--and especially not on weight! Of course when you look up the specs you can see that weighs nearly 1 kg. But you only realise what that means once you hold it in your hands. This baby completely undid all the good of my camera body's built-in stabiliser! I could not hold it still for a long time. And I imagined how it would feel on a day-long hike with a backpack on my back and the camera hanging in front. Or what an outing with the family would be like: 'Don't say a word to daddy, he's busy carrying his lens!' It took a couple of minutes to say goodbye to a dream of quality, but I clearly felt--literally!--that I must make a compromise here. So I went home with a mid-class Sony lens instead of the Carl Zeiss one. Not a bad one, but visibly in a different class. See the third example--I still was not able to hold my camera horizontal, apparently... Yet I'll be in a much better mood to take pictures, so if the technical quality of the photos may be a little less, my (and my family's!) quality of life is optimised by this decision.