Ricoh has been plugging away in the consumer dSLR space, even using cornucopia of colors to attract first-time and family buyers to its Pentax models, away from more staid Nikons and Canons. With the K-S2, the company combines Pentax's two-tone tendencies and colored lights with a more staid design sensibility and the weather-sealed body, large viewfinder and solid photo quality that have long been the hallmarks of Pentax's entry-level cameras. The result is a really nice option, with a single big weakness: video capture.and a
Overall, the photos look excellent. JPEGs are clean up to ISO 800 -- ISO 1600 in good light. Beyond that, they get soft fast. Color is very good; bright and saturated yet still accurate in its default Bright setting. Auto white balance is generally good, though in indoor lighting, photos develop a yellowish/pinkish cast, depending upon the light sources.
At press time, the only raw support for the K-S2 was the bundled SilkyPix software, and as far as I can tell there isn't a lot of detail you can recover from blown-out highlights without simply turning them brown, though you can bring up shadows without introducing too much noise. (Unfortunately, despite its popularity as a bundle choice, SilkyPix is -- at best -- impenetrable. Nor can you save as DNG for use in another program. I saved as a 16-bit TIFF file and used that in .)
The big disappointment here is the video quality. If you leave shake reduction on, which most people probably want to do, the movies have a bad case of the wobbles; not just typical horizontal jello (rolling shutter), but vertical as well. If you turn off the SR, the wobbles go away but then it's got the shakes.
Those issues aside, video also suffers from serious moire -- brick buildings in the background are covered in not-so-groovy rainbows. The camera has Pentax's AA-filter simulator, which defeats the moire that results from removing the anti-aliasing filter from the sensor (which blurs the image just a bit in exchange for removing visual errors from high-frequency patterns). The simulator shifts the sensor slightly for a second shot and combines the two to cancel out the patterns. Unfortunately, the AA-filter simulator is not available during movie capture.
It's also limited for general movie recording because it lacks continuous autofocus in movie mode. You have to focus manually or prefocus and hope your subject doesn't move from the focal plane. Given how noisy the inexpensive Pentax lenses are, I can understand the company forgoing continuous AF. It does have good focus peaking (outlines for in-focus edges), which makes manually focusing relatively easy.
Editors' Note: We haven't tested any comparable cameras since reworking our test methods last summer, so we're not including a comparison chart. It will be included when we test other comparable cameras in the future.
I'd rank the K-S2's performance as good -- certainly good enough for the basic "kids, pets and travel" photography for which it's intended -- but not better than average. It's a little slow to start up at 1.2 seconds, which is a mirrorless-camera level performance rather than dSLR. it takes about 0.4 second to focus and shoot in good light, which slows to about 0.5 second in dim. Unfortunately, it doesn't always refocus when you press the shutter, and in general the kit lens is a slow mover, so the time it takes to focus depends upon how far out of focus you're starting from. It fares better when it comes to two shots in a row: 0.2 second regardless of whether it's raw or JPEG, which is good. And it only jumps to 0.9 second with flash enabled.
Though it's rated at 5.5 frames per second for continuous shooting, I couldn't get it to work that fast with autofocus enabled. With AF -- our normal test conditions -- it maxed out at 4.6fps for about 40 JPEGs and 4.5fps for about 10 raw. With focus fixed on the first frame and a shutter speed of 1/250 second I got it up to 5.4fps. However, 4.6fps should be fine for most burst-shooting situations.
The autofocus performance, however, was a bit frustrating. Aside from the aforementioned occasional reluctance to even try to focus, I found that in situations where focus was critical -- most notably during our lab testing -- it was off just a hair. I ended up having to manually focus for best results.
It could be the new collapsible kit lens, which, with the exception of the ability to focus fairly close, is pretty disappointing. Or it could be vibration from the thwack of the loud shutter and mirror slap.
But in most cases the focus is generally fine and continuous shooting with the 18-135mm lens had a decent hit rate where the panning focus accuracy (manually following bicyclists as they ride through the frame) was good enough for most uses. It does offer expanded-area autofocus, which uses surrounding points as backup when the subject moves off of your selected focus point.
The battery life isn't great, either, but that seems to be a trend thanks to the effort to make the bodies smaller.