I just had the most wonderful session with my piano. I haven't been practicing much lately, if at all, so I was mostly playing through old faster pieces to get my fingers up to speed.
A funny thing happened. As I'm playing the first movement of a sonata by Mozart, I suddenly realize that I'm struggling with the phrasing, articulating the bold and the faint passages clearly. This piece has lots of f and p markings (forte and piano, strong and quiet). I'm not too happy with the sound either. So I start focusing on getting it more precise, only to discover that the extra precision makes the sound worse.
The attention to detail draws attention away from the big picture, which in this case is the melody itself. Some pieces are simple, the music is in the tones. I think this is what most people relate to when they think of a melody. But this doesn't hold for most of the classical music I've heard and played. The tones are still important, but usually the overall dynamics are what will carry people away. Think of a symphony, an enticing crescendo towards the thundering finale, like the theme from Space Odyssey 2001 (it's from "Also sprach Zarathustra"). If it can't erect the small hairs at the back of your head, you need a treatment with a five-pounder shovel.
So by focusing on the details, which at most can make the music more interesting, although if you haven't played the piano you probably won't notice 90% of it, one can happily leave the most important of the piece to the coincidence.
It didn't work for me. So I gave up the details and started listening for the bigger picture, and voila! Pure joy. For five minutes, I was completely carried away.
The interesting thing is that the exercise turned out to be surprisingly harder than you'd think because all those little fs and ps kept breaking my concentration. I know this happens to professional players too, because I've experienced it at least once with a brilliant player, playing technically well and with lots of dynamics - but with no overall structure to push my emotions making the whole thing plain boring.
There's a long and interesting discourse into the land of interpretation of music here. But let me instead draw a parallel to software engineering.
For both piano playing and software engineering, often all we got is a clearly incomplete specification of how the thing is going to work. But if you (just) follow the spec, the result is guaranteed to be mediocre. And the more detailed the spec is, the more it'll distract you from the overall picture which is usually where the fate, success or failure, is determined.
Now witness Wilhelm Kempff, a legend whose playing to me represents the opposite of the brilliant player mentioned above. He may make a couple of mistakes and he obviously bends the spec into what he sees fit. But I'm carried away.