Re: [microsound] [.microsound] "sound" of max/msp

[jhno <ear@xxxxxxxxxx>]

>One of the reasons may be that these tools are complex to understand
>(in comparison with guitar (or fuzzbox etc.) where the technical
>construction is simple to understand).

whoops - a bit of dyslexia, i think you got this backward...

most of the negative qualities that people associate with electronic music
are, i think, due to the simplicity of the sound generation.

a resonating zither with electromagnetic pickups, overdriving the vacuum
tubes of a 40 year old amplifier in feedback - now that is complex, and
quite challenging to understand (scads of monkeys still trying...)

the electronic instruments that have been most highly regarded over time
usually owe their characteristic sound to the idiosyncratic behavior of
certain circuits, algorithms, etc. sometimes even their designers are at a
loss to explain why certain differences are perceived.

software synths have their own individual character as well, but currently
the range of variation is not nearly as rich as that achieved by their
analog counterparts. it is a very young art. obviously digital synthesis
has many other things to offer.

perhaps i think your description backward re: sound generation - but from
most musicians' point of view, and speaking of the actual experience of
dealing with the systems, it makes much more sense, and certainly agree.
hence the question is more a matter of interface. a guitar might seem
simple and obvious, but we have the advantage of hindsight. it has been in
development a long time.

>I might guess that another reason could be that these tools (and
>things such as soft synths) probably are not implemented
>mathematically correct with respect to dithering & truncation.

most audio synth software on the macintosh uses native 32-bit
floating-point for calculations, and hence do not deal directly with these
issues.

>The result is an accumulation of
>errors (digital distortion) no matter you want it or not.

we want distortion... but only certain kinds. most accidental,
precision-related digital audio distortion sounds "bad" because of its
annoying non-randomness. nyquist described the enemy, and the ally.

i have been investigating waveshaping lately (a quite broad class of
distortion techniques). it is one way i have found to diversify the
"character" of msp audio.

on good distortion vs. bad distortion, here is an eye-opening paper from
the 70's investigating tube vs. solid state.

http://www.dwfearn.com/tvst1.htm

moderately technical; fascinating.

jhno