Re: [TANGO-L] MIDI simulation of the bandoneon

[CHARLES KELLY <carlitokel@USA.NET>]

On most PC sound-cards (and some MIDI sound modules/keyboards) the now
standard synthesized sound scheme "General MIDI" will contiain a sound (#24)
called "Tango Accordion," (a term which some would say is an oxymoron.!!)  In
any case, the sound is there; but the quality of it is even less satisfying
than the so-called "bandoneon stop" (octaves tuned pure, no musette) found on
some accordions -- and it is EVEN FURTHER from sounding like a real bandoneon.
Moreover, it has none of the characterisitc articulations or "bite" of the
bandoneon, nor the characteristic delivery effects (golpes, arrastres,
appoggiaturas, etc), nor the wonderful nasal quality found in the tenor region
of the bandoneon left hand, nor the sound of the bellows (fueye), etc, etc.  

Overall, I would say the GM #24 is not usable for any serious musical
purposes. I would think that any knowledgable listening public, particularly
those used to hearing the sound of real (or recorded) bandoneons playing tango
music, would find GEN MIDI #24 quite unsatisfying. It can be helpful in one
quite limited and clinical way: in conjuction with music notation software, to
give a VERY rough, approximate rendering for a composer or arranger to check
his notes and notation. But even for this purpose, it requires a stretch of
the aural imagination (another oxymoron). The composer/arranger must already
know, a priori, what a real bandoneon sounds like, and what to write for it
(or not.)  The synthesized sound, used in this limited context, might be said
to bear the same relationship that an architecht's rendering has, to the
actual finished building itself.  Perhaps useful for studying the blueprints,
but not a place where one would dwell very long with any profound pleasure.  
 
I have seen, recently, notices for a scheme of "sampled" bandoneon sounds, (a
detailed explanation of sampling technology being beyond the scope of this
discussion).  I believe these sounds have been produced in Argentina, but I
have not heard them, so I can't comment.  However, as with any sample library,
much depends on the method used to actuate the sounds (such as the KONTACT
software, sequencers, etc), and of course, on the person whose hands are on
the keyboard or other controling device(s).  The importance of articulation
and phrasing can not be overstated.

I heard once of a music/psychology study, in which the recorded sounds of
different instruments playing the same note, were played in an accoustic 
laboratory for traind musicians to listen to.  HOWEVER, the researchers had
chopped off the first 1/100th of a second of each note.  In the majority of
cases the subjects, even though they were trained musicians,  could not
differentiate between a single note of, say, a trumpet, versus a violin, in
this admittedly artificial setting.  It does however, serve to point up quite
vividly, how much musical "information" is contained in the attack (attack
"envelope") at the outset of each note. Indeed, the very identify of the
instrument is contained therein.   

Best regards to all.
SiempreTango