MOTM

Roger writes:
>>Fantastic!  This demo illustrates wonderfully the usefulness of the
sort of filtering I had suggested.  The first sound is immediately
identifiable as an analog synth-type sound.  The others sound great.  I
think they sound a lot like real strings, but they could just as easily
have sounded like something else, I imagine.  I'd think that a tool
that makes this easier would be useful.<<

The other sounds in that first demo sound somewhat string-like but are more
or less tests to see how different filter configurations sound.

>>Obviously, you can tell the difference between the imitation and the
real sound, but of course that's not the point.  The point is that a
carefully filtered can sound good in and of itself (to some people) and
is less recognizable as just another analog synth sound. <<

Yes, it gives another level of obscuring the synthetic quality of the
original sound.  Not bad.  Before I read your notes I thought all of the
sounds were yours and thought, boy, I'm going to have to get back to my
incompleted Nord brass project using high band count filter banks.

Mark writes:
>>I don't think that's true.  I doubt anyone would build a modular
unless they were into using it.  Otherwise, there are plenty of other
ways to get standard synth sounds -- such as all of the software
plugins that are now available -- that have presets, are easily
controlled by a DAW, and cost much less space, time, and money.<<

People build modulars and buy synths all the time and use them in relatively
basic ways.  Most who are buying modulars are only hobbiests, many can't
even play an instrument.  There is no better evidence for this than what
I've been through with the latest synths I've bought.  Within days, I'm
finding all kinds of bugs and limits that thousands of other owners never
noticed.  That right there points to the limited use these synths are
getting.  In addition to that, often some of my demos will appear on a board
somewhere, where people finally convince themselves that my demos must have
been faked or I'm manipulating samples, etc.  This shows that people aren't
even aware of the capability of synths in the first place.  And as pointed
out by people on this list and others, as soon as it comes to doing
something difficult, then pull out a sampler, a physical modelling synth or
maybe a digital workstation synth and do it there.

>>While I think that the ability to share files, and having a massive
number of bands, and the ability to analyze sounds, and all these
other features, are great for a filter plugin, I do not see how they
would translate well into a hardware synth module.   If you have to
page through all these menus, or use editor/librarian software in
order to use it, then what's the point in making it hardware??<<

It depends how it's done.  In hardware it would look something like a
graphic EQ with other buttons to engage various features.  No menus or piles
of parameters to sift through.  It's just anytime I think about a product
like that I think, it's something that I would use but I don't know how many
other people would.  Sure a few other people would buy one just because
there are people who collect rooms full of gear, but admittedly it would be
a small market.

elle_webb writes:
>>This attitude seems pretty dismissive of today's synth pioneers. There's
as much creative work being done in electronic music as ever, imitative or
not.  If you don't think there's pioneering synth work being done today, you
should consider asking list members for suggestions. You might find some
great synth music that you're missing.<<

I've been on many lists and listened to people's recommendations and I read
magazines.  Believe me, if the next Tomita or Carlos popped up somewhere,
I'd have heard about it.  Sure, every now an then I'll hear a nice sound
done here or there, but these are not necessarily difficult sounds that
require dozens of modules and hours/days to program nor are covering new
ground and new programming concepts.  In fact, it usually seems to be
digital synths where people are getting more impressive sounds.  In the
music field things don't get better, they get worse as standards continue to
slip and as people won't even listen to the past virtuosos of synths.

>>In terms of imitative synthesis, some of the most innovative work is
being done by Eric Lindemann at Synful. His work shows how the
synthesis of phrase articulation is at least as important as the
sounds themselves in creating realistic and expressive imitative
synthesis.<<

He doesn't synthesize sounds.  It's resynthesis.  It's also additive
synthesis.  In fact he was going back to re-record (real instruments) some
of his sounds for resynthesis again because of some criticisms by people.
Yes, expression is also important which is why I have plenty of that in many
of my sounds such as those violin demos I posted.  Some people on another
list thought I had to be running all kinds of midi data from a sequencer to
acheive that result, but I had to correct them that it was just done with
velocity and a pedal.  It's just that many parameters are controled by those
two performance controls.  Synful on the other hand has to add some
expression after the fact so it can look ahead at future notes and calculate
what to do.  That's not really a good real-time instrument.

>>Imitating traditional instruments, though, seems to be a bit of a
Sisyphean task - the more realistic you make your imitation, the more
acutely you realize that it's an imitation.<<

There is something of the law of diminishing returns in how much time spent
per amount of improvement.  But it's possible to get to the point of
creating imitations that can fit right in with real instruments and you
can't tell the difference in many cases depending on the sound.  I've
already demonstrated some of that in the past, and those weren't even super
complex patches either.

>>Even if you could create a perfect imitation of an instrument, it
would be of limited musical interest, because it doesn't offer
composers or performers anything new.<<

I like to create some imitations to give me the real-time expression and
potential realism of physical modeling, but doing instruments that physical
modeling isn't good at.  As for limited interest, based on the market for
sample libraries and imitations of real instruments (such as synful) and the
lack of interest in new sounds by much of the market, people seem to like to
work with sounds they are familiar with and know how to use.

>>That's why the most creative synthesists aren't trying to imitate
traditional orchestral instruments. They're creating original sounds
and using these sounds in expressive ways.<<

Unfortunately they're not getting anywhere close the almost unlimited
rainbow of colors or Carlos, nor the complexity and attention to detail.
It's hard to get people off the butts to do so much as buy a filter bank to
get some different sounds, or to convince them to layer different sounds for
more complex new sounds.  And those who claim they want to use a synth for
only new sounds then end up mixing in samples, real sounds, mellotrons, drum
machines, digital pianos, etc., so we're right back to square one, not
utilizing the synth to its fullest.  That's the problem, when people want to
be expressive and musical they seem to end up not using analog synths.

-Elhardt