Originally posted by Stagnant I see a massive difference in IQ. The K-01 scene seems to be a bit more challenging, but in general the codec in the K-3 seems to compress much better than the K-01, despite having similar bitrates. The K-3 seems only a bit inferior to the K-5. The K-01 though... also contrast and sharpening seems to be way higher, which could give the codec problems. Sharpening gives the encoder way more detail that it has to compress, so if you run into blocking artifacts I'd suggest to reduce sharpening to the absolute minimum that is possible. That way the encoder can concentrate on the actual detail that is there, and you can still add sharpening in post.
When I reduce the contrast with my K-5 I seem to either, depending on natural or muted, get raised dark areas where I can see massive banding (say... some areas are 0, and the next step the sensor can retrieve above it is 10 already, then 13, 15, 16, ... especially in dark areas. That's not so useful, sadly) or bright areas where the sensor hits the limit (similar, just, well, 228, 230, 255. Something like that). Not sure if it is a sensor problem (there's just not more dynamic range). Could try that at some point by shooting a RAW photo together with a JPEG, I guess... See if there is more in the RAW.