Fretless MIDI guitar

[S V G <vsyevolod@yahoo.com>]

At 8:33 AM -0800 1/2/03, S V G wrote:
>From there it is up to the pitch to MIDI converter to unravel all 
>this information...if you play a note within 50 cents (half a 
>semitone)
>either side of middle C with a fretless guitar, the pitch to MIDI 
>device will output the note "C" plus the appropriate pitch bend 
>information (or not).  Another note played within 50 cents either 
>side of C# will trigger a new note as "C#".  The only time this 
>doesn't hold true is when you pluck a note and then slide your 
>finger up or down the string.  The note has already been
>activated and chosen and then the pitch bend information is sent to 
>modify that note up or down an octave or more.

  RZ:  This system has both advantages and disadvantages. It assumes that 
pitch bends will always start within a semitone of the nominal pitch, 
but it's quite common for some players to bend down from a whole tone 
or a minor third above.

   SVG:  This shouldn't be too big of a difference to matter.

   RZ:  The system I mention as being available on the Peavey Cyberbass 
(each 
string outputs a single MIDI note plus pitch bend within a two octave 
range)...

  SVG:   This doesn't jive with my understanding of how MIDI works, one 
note that gets bent up or
down an octave?  Wouldn't that sound weird at the extremes?

  RZ:  ...has some interesting possibilities for fretless and even 
fretted MIDI guitars. For instance, it allows playing with flexible 
intonation. Microtonal players (whose pitch discrimination can be 
uncanny) would be able to play guitar synth in Just  or other 
intonations without the synth having microtonal features. Even though 
many synths have tuning capabilities, it requires switching presets 
in order to change from one to another. With a MIDI pitch bend based 
system the intonation would be left entirely to the player.

  SVG:  Again, I'm not sure that I understand what you are referring to 
here.  I've been building
microtonal instruments for the past 25 years, though my knowledge of how 
MIDI works is on the
scant side.  If a patch had a pitch bend resolution of +/- 1 semitones, 
wouldn't this be enough to
obtain most any microtonal interval as long as you don't bend the note 
outside of that semitone
range?  And the pitch bend accuracy would be to the nearest 1.5625 cents 
(100 cents divided by 64)
which is close enough for rock 'n' roll (though some would moan in a most 
pitiful way).

    RZ:  There are certain technical issues related to pitch bend 
resolution. 
Unfortunately most MIDI equipment seems to be designed with 7-bit 
pitch bend and this permits only 128 discrete values. This means that 
bends are limited to 63 or 64 steps up OR down. With a semitone bend 
the individual steps are about 1.5 cents; a while tone is about 3.1 
cents. Both are OK, but if you stretch 128 values over a two octave 
range you get pitch increments of nearly 19 cents!

     SVG:  Okay, I tried this on my keyboard synth and I found out what 
you're talking about.  If
I set my pitchbend range to +/- 12 semitones, my pitchwheel gives me a 
"stepped" output as opposed
to a smooth slide.  I must have been getting around a 10 cent jump with 
each increment.

   RZ:  The technical 
solution to this is to use the full two-byte, 14-bit resolution 
allowed by MIDI, resulting in 16,384 steps and pitch resolution of 
about 0.15 cents. I don't think there are many sound modules that can 
match that. The Cyberbass Voice Module and the Oberheim Matrix synths 
can handle it.

     SVG:  My Emu synths all have a resolution of +/- 64 per semitone 
which is 1.5625 cents.  Not
great but better than nothing.  Kurzweil has a resolution of +/- 100 per 
semitone which is in 1
cent increments which is a whole lot easier for a musician to work with 
(one of the rare moments
when Kurzweil had the musician rather than the nerd in mind...) and other 
systems tend to follow
one of these two models.  However, these numbers are referring to the 
gradations available when
constructing microtonal scales as well as the pitchbend range.  With the 
Kurzweil, you set up one
octave and all the octaves are locked into that same tuning.  With Emu, 
you have to set the tuning
for every single note which is more labor intensive yet it offers much 
more flexibility (i.e.
octaves don't have to be in tune with each other, which is essential for 
most ethnomusicological
scales).  Regardless of whether you are using a fretless or a fretted 
guitar, you can set a user
definable scale on your synth, turn off the pitchbend, and anything you 
play on your guitar will
ipso facto sound within that scale that you defined.  It would be 
imperative to not blend the
acoustic sound of the guitar with the synth sound if that were the case.  
Still it holds some
potential.

  RZ:  BTW - Middle C is "C4" no matter what Yamaha says.

     SVG:  I know that MIDI was developed with several musicians on the 
consulting panel.  What
were they thinking?  That they could just ignore the modern acoustic 
usage?  Especially since
there seems to be no overwhelming reason that they couldn't just keep C4 
as C4.


>I can personally vouch for the Godin/RMC combination, the sound of 
>my nylon Multiac is extremely satisfying to my ears as a classical 
>guitarist.

    RZ:  Any comments on the difference between Multiac and ACS-SA?

     SVG:  I actually owned the ACS-SA for a few weeks by accident before 
obtaining my Multiac-SA.
 The ACS is a less expensive guitar, being mostly a solid body nylon 
stringed instrument made out
of western maple.  Ergonomically, it was too heavy for me and it didn't 
hang on my body well. 
Though I don't play electric guitar and it was probably not too unlike its 
close cousins.  My
sense is that it tracked slightly better than the Multiac as far as my 
GR-33 was concerned,
probably due to fewer acoustic resonances getting in the way.  It also had 
a little better
sustain.  As for the Multiac, it is much lighter, still hangs a bit 
awkwardly on my body (I think
I'd do better with the Godin Concert Classical as the body joins the neck 
at the 12th fret rather
than the 15th fret) and it has a far superior sound straight off the 
piezos.  Acoustically it is
much louder than the ACS which is a drawback if you are using a VG 
processor with alternate
tunings, yet a plus in most other areas.  It it made with a spruce 
soundboard and a nice sized
chamber in the center portion.  If I was going mostly for the synth access 
qualities, I'd get the
ACS.  With it I found that I could get more expressiveness both in 
tracking, sustain, and
amplitude characteristics using the onboard synth of the GR-33.  If I was 
interested in the
acoustic qualities of a nylon string guitar blended with synth access, I'd 
go for the Multiac. 
Since I find myself blending the two sounds a lot, I'm a happy camper with 
my Multiac.

     SVG

__________________________________________________
Do you Yahoo!?
Yahoo! Mail Plus - Powerful. Affordable. Sign up now.
http://mailplus.yahoo.com