On Fri, 2010-12-31 at 12:49 +0100, Julien Claassen wrote:
The lame encoder adds a noticeable warbling at 128Kbit, there is less on
the original, but still:
The attack part of the sound is composed of two short samples of a
Yamaha grand of slightly different velocities, but here just spread out
for a wide and clear soundfield. Velocity nuances are instead handled by
filtering. The short piano samples by themselves have a very audible
loop, as in: "Ding, Zing, zing, z...", so after half a second (or so), I
blend in a mild pwm with a saw, filtered to match the overall sound. as
well as to give a slight vocal, singing character. Real pianos do not
have this kind of extended sustain.
The pwm/saw is finally routed through a three way chorus which makes
them sound a bit as if it were the neighbouring strings that were
resonating and reverberating. Or, that is at least what was intended.
The modulation wheel is wired to vibrato (on pwm/saw ) as well as send
to delay line for a longer, even more "singing" sustain. This is not in
use here though. The second modulation wheel is wired to turn filtering
far down, plate reverberation off as well as release rate to none for a
very dry, round "fusion-like" micromoog sound.
> Was this
It will "beat you in the face" if you collapse the two channels into
mono - so don't do that!
> Warmly yours
Linux-audio-dev mailing list