Quicktime pitch resolution

Hi Paul,

> > I've done a test by setting the output for Quicktime to Midi Yoke
> instead
>> of its own internal synthesizer, then recording that in FTS.
> >
> > This shows that in that situation it changes the pitch bend
> range to 24 semitones
> > before relaying the notes - i.e. +- 12 semitones, giving a
> resolution of 0.15 cents.

> well this is neither 128 nor 256 parts per semitone, but rather
> over 600! seems like kind of a mystery, huh?

Yes.

In any case, it's a mystery to me why they would change the pitch
bend range at all just to play ones midi clips on another midi device -
you'd think they would just pass everything on unchanged. Instead,
they change the pitch bend range, and also remap all the pitch bends
in the file in accordance with the new range.

One could perhaps imagine that maybe there are two parts
to it - a pre-processor and the player itself, so that the
output for the other player has been through the pre-processor first

But then, if it is like that, why does it get changed to a range
which doesn't match up with their documentation for Quicktime?

Anyway I did sone measurements using FFT - and it was within around half a cent
or so of the correct pitches for a 30 second major chord, so not too bad
for a wavetable soft synth.

Robert