17 March 2010

A practical look at K-7 autofocus

When I bought the K-7 last summer, the first thing I did was to compare Autofocus performance to its older sibling K20D. Under relatively controlled and predictable conditions, I found the K-7 to be a significant improvement both in speed and precision (blog article here).

Then I went to the South Atlantic, and to situations demanding more of the AF systems than did my initial testing. My experiences from this trip is a bit more nuanced than the first impression, and so I felt it would be natural to share here on the blog. It has been very long in the coming, however. Too long, maybe. Hope someone will find it interesting nonetheless.

I will take a practical approach, and present you with a few scenarios which made me think about AF performance in particular.

First, there was the case of the ordinary Bird-in-Flight (BiF) shots. In one respect the K-7 performed admirably; it is adequately fast at achieving a focus lock. However, when shooting from a ship, the background will be water in a large number of your shots. Water with a varying degree of froth and wavecrests. I found those white patches of foam to be very deceptive to the AF. If a bright spot appeared in the background, the AF would lock onto that even if the closer subject was many times the area in the viewfinder. Especially if the foreground bird had bland or darkish colours. In this situation, I really wish the AF would have more affinity for proximity and less for brightness. Compared to the other brands present on one occasion (Canon 7D, 1Ds MkIII and 5D, Sony A850 and A900, Nikon D300), the K-7 was definately lagging behind in AF performance/accuracy.

Pentax K-7, DA* 300/4
1/500s, f/8, ISO 400
The Cape Petrel with its checkered pattern was a good target for autofocus.
My AF hit rate was maybe 70-80% with this species.

Pentax K-7, DA* 300/4
1/500s, f/8, ISO 400
The Giant Petrel was a real PITA to keep in focus.
My AF hit rate was an abysmal 15-20%.

Then I would like to mention some not-so-ordinary BiF shots; obtained with fill-flash in low light. Made hand-held with the DA* 300/4. The flash was a Metz AF-58 flash set to -1.5 stops compensation, and a BetterBeamer flash condenser. On one occasion, I was on deck with 3 Canon shooters who all gave up, complaining their cameras wouldn't lock focus on the birds in the twilight. I kept going for nearly an hour afterwards, the K-7 remaining sure-footed as long as the background was not brighter than the bird. In fact, I kept going much longer than I had expected was possible. And for the record, I did not use the focus-assist lamp. So kudos to Pentax for low-light AF performance from me.

Pentax K-7, DA* 300/4, Metz AF-58 flash
1/180s, f/9, ISO 400, -1.5 flash comp.
My hitrate here was pretty low,
but then again I was the only one getting anything at all.

The next challenge for AF was blizzards. Technically speaking, however, this is a variant of the same issue as above; brightness affinity. When snowcrystals cluster together in big flakes, the AF tries to lock on them as they pass the sensor area. The net result is an awful lot of focus-zipping back and forth, and never locking onto the subject beyond the falling snow. A couple of Canon shooters confirmed the same thing, and our Sony guy just sighed and shook his head. Both he and I switched to manual focus only, and worked that way throughout the blizzard. I don't know how the Nikon guys fared under these conditions, and some Canonites seemed to be doing okay. However Pentax was certainly not the only struggler.

Pentax K-7, DA* 60-250/4
1/200s, f/7.1, ISO 200, manual focus

On the whole, I would say that the K-7 stands its ground pretty well. It is certainly the best-performing AF system ever produced by Pentax, it runs circles around eg. the *istD and easily outpaces the K10D and K20D. In comparison with the other brands, however, it's not leading the pack in any way. I found better performance than some Canons in low-light conditions, but I can't recall exactly which models those guys had. A comparison with non-contemporary models would be unfair and misleading.

It's still the Big Two that have the most advanced AF systems, and Pentax will probably forever be at least one camera-generation behind with regards to performance. That's a fact of life.

For myself, I find that I can either enjoy the advances Pentax has made with K-7, or continue to look longingly at other brands and say "Pentax makes too little, too late". I also find the latter attitude to be very counterproductive in terms of what kind of images one attempts, and succeeds, making.

15 March 2010

The Lure of the Laptop

When I got home today and had a look at the previously posted picture from Goksøyr, I realised just how much I had pulled the levers in Lightroom to produce what was posted. On my desktop monitor it looked completely over the top, with unreal saturation and a warm tone that did not exist as far as I can remember the scene.

So what happened? I think I was deceived by the laptop screen. Though calibrated, it renders the colours very differently from the desktop screen. It is also very sensitive to viewing angle, and the shadows will go either black or a blocked-up kind of grey depending on how I hold my head.

To save you the trouble of looking at the previous post, here's the laptop rendering too:

13 March 2010

Goksøyr, Runde

Another pic from our week-end foray.

Pentax K-7, DA* 16-50/2.8
1/250s, f/8, ISO 400

Shielded from the westerly and North-westerly winds by the island of Runde, the houses at Goksøyr are sheltered from most winter storms.

12 March 2010

Old Fence

Pentax K-7, DA* 16-50/2.8, tripod
20s, f/9.5, ISO 200

Observed at Runde tonight. If the highlights look a bit odd, it's probably because I got it full of sleet and tried to wipe it off with the sleeve of my almost clean fleece jacket.

I think it doesn't look too bad, all things considered. :-)