OVERRIDING MIDI pt.II

[Eli Brandt]

Steve wrote:
> >For instance, I think the biggest problem with MIDI is that it treats
> >everything as note-on note-off messages, and for most types of real
> >musical expression, say for example a sax player wailing, that's a
> >pathetic representation of a musical event.
> 
> Well, you do have all of those continuous controllers, but a resolution of
> 0-127  is just vile.

Yeah, those `continuous' controllers. :-)  The MIDI spec actually uses
the paired-controller technique you discuss -- it claims that controllers
32-63 are the low-order bits for 0-31.  I've never heard of a synth that
implements this by default...

> and remap the two bytes to a pair of cc's ("coarse" and "fine"), then set
> up your synth (if it allows for this) such that they both modulate the same
> parameter, the "coarse" having 128 times greater modulation strength,

I've never tried this, but I imagine you'll have a problem unless your
synth is smart enough to update the pair atomically.  Say you send 
        msb:0x23, lsb:0x7d, lsb:0x7e, lsb:0x7f, --
At this point you want to send (msb:0x24, lsb:0x00), but whichever you
send first will cause a glitch as big as the low-order bit in msb, which
is exactly what we wanted to avoid.

> What about a multiport MIDI interface *on the instrument*?

That's got a certain brute elegance, as well as a low development cost.
If I were pushing a new standard, though, I think I'd want more
performance increase for a given price increase.  And hanging more than
half a dozen MIDI cables off of a box is going to get inconvenient.

-- 
   Eli Brandt
   eli+ at cs.cmu.edu