Originally posted by kypfer The K-5 is a fine sturdy camera, but I'm not so sure it'd be such a significant upgrade from a 14.6Mpixel camera to be worth the expense, especially for something like portrait work, where the viewfinder would normally be relatively well filled so little cropping should be necessary.
I tend to agree. It's a long, long time ago that someone told me to not bother with comparing MP counts unless there's at least a 50% increase. One of the reasons is of course that the megapixels are actually the square area of the sensor pixel counts in both dimensions. What looks like an awesome jump in megapixels, quickly fizzles if you realize you're only getting a few more pixels on either side. (4672 x 3104 for the 14mp K20D against 4928 x 3264 for the 16mp K-5/K-5 II(s)
So the benefit of an upgrade has to come from something else and that something else needs to be dramatic enough to justify the cost of the upgrade. If all else fails, a lens upgrade will probably deliver far more of an IQ boost than a body upgrade (unless of course you already have top-rated lenses, that is).
As the upgrade from a K20D to any of the 16mp models doesn't make sense from a pure pixel count perspective, one needs to look at high-ISO performance, AF speed and accuracy and overall usability. That's why an upgrade to a K-5 II(s) makes a lot of sense when high-ISO and indoor shooting is a large part of what you do. If it's landscape or cityscape, a wide-angle large-aperture lens makes more sense as an upgrade and if it is extra cropping room you need, a jump to any of the 24mp bodies should be a decent consideration.
Now, you might ask: was I this smart when I upgraded in the past? Nope! I made the mistake of doing exactly what was suggested here and going from the K100D-Super to the K20D (great upgrade) and then on the the K-5 (disappointing upgrade). It took me over a year to stop regretting the sale of my K20D which was one of the best Pentax bodies I ever owned. (still miss that awesome form factor)
I actually still don't see a lot of difference between my K20D files and the K-5's and it was only when I upgraded to a K-5 IIs that the usability improvement made me feel that I was actually pushing the envelope. The upgrade from the K-5 IIs was a careful consideration between going for the KP and the K-70 which the latter one won because of more bang/buck and the fully articulated LCD. Both of course had the 50% jump in pixels from 16 to 24 over the K-5-series.