Theres a difference between "a bit more resolution" and "getting the best out of the sensor" though.
Looking at the NEX 7 lens reviews on photozone the problem seems to be less that you'd need to spend alot on lenses to fully exploit the resolution but that you might not find any that do so at certain focal lenghts, the new Sony 16-50mm for example has a significant dropoff in sharpness at the boarders.
For landscape use it just seems like the D3200 has pushed further into the realms of diminished returns than the D800 has with pixel density.
In order to get the best out of any sensor, one would need the perfect lens, which doesn't exist. In order to get the best out of any lens, one would need the perfect sensor, which doesn't exist.
Any resulting image, in terms of amount of detail recorded, can never be more than the product of the sensor performance and the lens performance. The law of diminishing returns applies in both cases, whether one is increasing the performance of the lens, but keeping the same sensor, or increasing the performance of the sensor, but keeping the same lens.
Nowadays, I tend to use zooms more than I use primes; not because I think that zooms are as sharp as primes (although the Nikkor 14-24 at certain focal lengths may be) but because the extra detail one gets from an expensive prime seems very marginal to me, and consequently the greater flexibility of the zoom is much preferred, especially now that ACR includes lens modules that automatically remove obvious distortions, color fringing and chromatic aberration,
A point that needs to be stressed here is the difference between acutance and detail. An increase in the pixel density or pixel-count of a sensor will not produce more accutance, that is, knife-edge sharpness. It will just allow finer detail to be recorded. If the finer detail doesn't exist in the scene, it can't be recorded.
The sharpness or contrasty 'pop' that may be very eye-catching in a photo can usually be created with skilled post-processing, but such processing cannot be applied to detail that was never captured in the first instance. A sensor cannot capture detail beyond its Nyquist limit. If it appears to do so, it's called aliasing or false detail.
Below is an example, taken from the D3200 studio shots from Dpreview. I posted this in another thread, but in view of Les Palenik's comment that the D3200 shot is merely
slightly sharper
, I've further cropped the relevant section where the D7000 reaches its limits as regards recording of fine detail, but the D3200 handles the detail with aplomb. The D3200 image is on the left, of course. The difference is not subtle. It's like chalk and cheese.
Years ago, when the only detailed information on lens performance readily available was from the Photodo site, I used to test a lens after buying it, using my own made-up target consisting of line charts, fabric, sandpaper, clumps of fine twine, fine text and wood grain. If the lens didn't perform according to my expectations, I would return it, and either get a refund or another copy of the same lens, and test it again. (I admit I was a bit obsessed in those days
).
If I were to test two lenses for comparison purposes, and see the differences as shown in the cropped image below, at 200% on my monitor, I would be flabbergasted.
In all my testing and comparisons of lenses, I've never witnessed such significant differences, using the same sensor.
