Friends:  As others eloquently have said, any "performance" of a piano
roll is only "authentic" if it includes the human being behind the
pedals, since the maker of the roll assumed that as a given.  Same with
classical compositions (except avant-garde tape music and the like).

As with any performance, the composition should be respected, with
dynamics and pedaling conservatively applied.  A musician can also
do his work at a coding desk, but that is many times more difficult.
Why do something in an hour when you can do it better in 3 minutes?

"Use a real piano when recording rolls" -- I respectfully differ with
the idea that a piano font presents a real piano.  I would say rather,
it presents 88 x 32 separate real pianos (or however many samples are
in use), in which simultaneously depressed keys are acoustically
isolated from sympathetic vibrations across a cabinet interior that
would otherwise occur.  (Hold down various keys silently and play
single notes loudly, and you'll hear a few varieties of that.)

If the piano being imitated is a player, we also miss the air-pressure
droop that robs subsequent strikes of a bit of their volume.  In a
player, the leading edge of syncopations have a bit more bite to them
than in a system where full power is available at every instant.  The
thicker the voicing, the greater the effect.

I agree with John Tuttle -- I have always been able to spot sampled
pianos a mile away.  You hear them on TV and in movies all the time.
That cultural shift is one of the driving forces behind the MMD
"movement", and those who endure the expense and inconvenience of
dusty, hulking musical machines.

Yet, paradoxically, all the electronic innovations extend the reach
of the music beyond what would be possible otherwise, and the harder
Warren Trachtman and friends work on the fonts, the louder we should
cheer.  The scientists will not put piano factories out of business,
any more than medical research makes hospitals obsolete.

However the Charley Straight project is recorded, my check is ready!

Jim Neher

 [ I hope the computer scientists are addressing the deficiency of
 [ no sympathetic vibrations when the sustaining pedal is pushed.
 [ Modeling the resonators (all the idle strings) isn't difficult,
 [ but it's a challenge to make the synthesized resonating strings
 ] to react realistically with the samples of the struck strings.
 [ -- Robbie