Originally posted by Adam I haven't really investigated this, but the low-light performance didn't seem to be better using the 85mm F1.4 compared to the 18-135mm.
The f/2.8 AF sensor is not meant to be better in low-light. The "f/2.8" does not refer to light levels, but to the length of the baseline of the AF area. The longer the baseline, the more accurate AF can be.
Originally posted by Adam I have a hunch that the center point won't help much in low light, as you benefit a lot more from having all 11 points enabled then.
You seem to link "AF accuracy" with the capability of a camera to achieve good focus in action situations. I believe most people link "AF accuracy" with the level of precision at which a camera can focus on a subject (typically stationary, as to not measure AF speed).
Originally posted by Adam Daytime tests are planned for tomorrow, but TBH, I have a hunch that we won't observe much of a difference compared to the K-5.
Unless Pentax have screwed up the implementation, an f/2.8 AF area
must be more precise than an f/5.6 area (the standard areas).
Actually demonstrating the superiority of the centre f/2.8 AF area in terms of AF accuracy is not easy since every AF attempt inevitably reflects on small tolerances and inaccuracies. However, in practice, on average a higher precision f/2.8 AF area will yield more shots in better focus than an f/5.6 area.
That's why high-end bodies all have f/2.8 cross sensors and Pentax went through the trouble of adding at least an f/2.8 line sensor to the K-5 II.