Creart:
This may not affect the actual timbre (the harmonic content) of the treble notes, but it greatly affects the sound of the treble in general: play around with the Body Resonance slider. Reducing the body resonance gives more body to the upper treble notes. Increasing it, to my ears, pushes the decay of the note away, as though the frequencies are sliding away into a large space. When it is reduced, the decay seems closer at hand. (Is this a problem? Does the body resonance affect treble notes more than it should, while it sounds good for the midrange and bass?) At times, in the model, I'm not sure if the piano body resonates more than a real piano would for high notes or too little--I'm not sure if it's vibrating too much (and the sound I hear is the resulting vibrations), or if it is too resistant, and is thus reflecting the sound out more than it should. I think I want it to reflect the sound, but push it around in the piano more, and towards the player and all around the player, instead of away from the player.
In other words, my impression is that the code is the same for the low and high notes--low notes vibrate the wood more and thus sound closer, but the higher pitches bounce off the wood and are dispersed. But for the player, for a realistic emulation, these higher frequencies could be more present, since the high freqs are both bouncing towards her\him and bouncing back and forth inside the piano.
I have no idea if this has been studied--the rate at which various freqs decline as they reflect back and forth off the inside of the piano. It's not convolution. Or is it? Or is part of the problem, if there is one, the point in time at which the reflection occurs--do the treble notes get reflected a little too fast off the soundboard? Or is it both the amount of reflection and the speed at which it leaves the soundboard? Am I making sense at all?