[sdiy] Yamaha DXy DCO's

[phillip m gallo]

Peter,

In my lexicon, DCO is, yes, a digitally controlled oscillator.   Now if
you want to be specific you can call an all digital oscillator a
numerically controlled oscillator, but if it's preceded by an ADC and
interfaced to control voltages NCO sounds less than perfect.  I always
end up with DCO as an appellation.

DCO's range from numerically controlled phase accumulator with table
lookup, to timer based integrators, phase accum integrators, sampler
types, and DAC moderated saw integrator's (and more) , so it is indeed
extremely general.

I find that # of operators in FM is important in one way, but
flexibility for control of however few operators you have is just as or
maybe more important.  

This is why analog FM is still very powerful even though stability
issues and number of available operators impede the recreation of sounds
easily rendered by DX technology.   Analog folks seems to readily
understand "how" to implement dynamism, where digital folks take a while
to learn that conceptual dynamism and appreciable dynamism are very
different.  Now i apologize for this generalization as i intend it to be
illustrative not stereotypical.

If the Sampson FM machine is that you ref at Stanford, it  had very
sophisticated parametric "control" that really was an important part of
it's very comprehensive capability.  

The Yamaha programming model was simple in concept but tedious in fact.
The simple implementation of manual controls to FM sound generation
allows quick understanding of how to make FM sound interesting.  Instead
folks ended up either using "dead reckoning" or "the manual says press
this key".   

The 5 phase ADSR's (or is it ADS-SL-R) envelopes allow some interesting
control of operator amplitude and if you make the rule to turn left when
it starts to "Clang" FM makes some wonderful sounds especially in bass
registers where the harmonic dynamism is very effective.  I have heard
"piano bass string" that would make a BLÜTHNER cry.

Another thing making analog FM very valid is that it usually is
performed mono-phonically and "tracked" into polyphony, so it has more
independence of voice than a piano style technique on a poly instrument
(tending toward the organtuan to use that word twice in one day).   It
is the organ, that a poly analog or digital instrument often becomes
that is the performance and compositional limitation in a lot of uses.
Independence of voice as a result of multi-track rendering is as
available to poly digital instruments as monophonic's and always sounds
better (IMHO) than pounded chords piano/organ style even when the intent
is very homophonic.

regards,
p


-----Original Message-----
From: Peter Grenader [mailto:peter at buzzclick-music.com] 
Sent: Tuesday, February 17, 2004 4:06 PM
To: Cynthia Webster; philgallo at attglobal.net; synth diy
Subject: Re: [sdiy] Yamaha DXy DCO's


I've realized any sound with harmonics can be filtered nicely- even
one's favorite sound - one may be surprised.  Whether its coming from
digital analog, your pet,  etc.

Onto semantics -  Is not a DCO a 'Digitally Controlled Analog
Oscillator'? Sounds like Phil did a real-live digital oscillator which
is a different animal.  The filtering , which knowing Phil's work
probably sounded WAY better than Yamaha's is attributed to the nature of
digital VCOs being controlled by digital EGs.  Precision being the key
and the very reason this type of thing is impossible with Analog
components, because this level of tracking can't be had.  Not to the
point where phase cancellation creates filtering effects.

The same kind of deal this possible with the Analogue Systems six voice
digital polyphonic oscillator.

Now, not trying to stick up for Yamaha, BUT it's the manner in which
they incorporated polyphony which gave these instruments their trademark
thin sound - which I believe is actually cycling through single pitches
at an incredibly fast rate.  Also bare in mind (and this goes down as
the saddest moment in the biggest deal in synth history) the machine at
Stanford they bought the technology rights for consisted of something
like 256 operators (digital VCO/EG pairs) on one side, and 256  sound
modifiers (filters among other things) on the other .  They just elected
to squish that down to 6 operators and no sound modifiers.  It wasn't
until later they they released
FM based machines that had these effect modifiers in them.    With
someth9ing like 500,000 DX7s sold, it surely wasn't a bad decision on
some levels, but for us tweek geeks....

- P


Cynthia Webster wrote:

Hi Phil!

The recent thread comparing DX7 and Buchla FM synthesis techniques
pondered the use of Analog Filters on Digital FM Oscillators as
something worthwhile to play with.

This reminded me of the experience that I had when
you brought your Yamaha DCO board to a SciSound
meeting.  I was rather amazed at the quality of the sounds
when no filters at all were involved.

Being a total filter junkie myself, I was humbled by the filtered types
of sounds that you were getting without any.

So naturally, I would love to hear that wonderful DCO combination
through a juicy analog filter like the (4-pole) Nyle Steiner design!

Maybe we'll finally get to try this at Jim Patchell's next
DIY meet?  (March 27th same location as last year)

It is so odd that folks with fairly large systems typically seem to
dedicate only a single filter to each "voice" (when it get far more
interesting when multiple filters are involved) 

Clever use of Digital Oscillators suggests that if some
of the "filtering" is done without filters...
then some total magic may be possible when they're
combined with um, (forgive me)... The real thing!

There are some keyboard synths that combine the two,
anyone care to share their experiences here?

Cyn




on 2/17/04 12:26 PM, phillip m gallo at philgallo at attglobal.net wrote:


Posts regarding polyphony so far have only dealt with multiple "like"
voice
channels.   There is more to polyphony than implementing this way.
Polyphonically driving multiple "different" voices  is a very potent
technique.  It does require a more sophisticated "Note control".

As an example, in my pre-MIDI days, i built a controller which drove 9
Yamaha DXy DCO's with my SDIY modular tracing the highest key (descant)
, and my MG-1 tracing the lowest key(basso).  The DXy's where given all
notes but even here i did not program the DCO's to have the same voice.
This was a digital implementation of what i had done in an analog way
with a Lancaster Poly and the two synths.

This technique provides a really interesting ensemble effect that
resulted in two "performance" characteristics:

1) 
You playing stop's being pianistic or organ-tuan but becomes very
precise as notes allocate to "voicings" and repeatability requires
precise technique,

2) 
Alternatively, pianistic playing resulting in very cool ad hoc voice
assignment which provided a "uncertainty" to the "orchestration" which
had the dangerous side effect of keeping you playing the synth way past
bed time and bending your orchestrational "ear".  Reminded me of the
insite gained from playing a "prepared piano".

Now MIDI make all this very easy.  Your M2CV needs to have rules in it
for not to voice assignment, an example being how to detect new voice
activations vs. "overlap" between key activations.  Since MIDI provide
velocity information as well as channel info it's pretty easy to make
decisions as to voice allocation.  You do need to provide a buffer for
key's pressed but not presently assigned to help govern these rules.

I have always disliked the rotary note assignment that poly synths often
do and am sure this is a major contributor to polysynths sounding like
really neat but non-the-less organ-like instruments.

regards, 
p 

-----Original Message-----
From: owner-synth-diy at dropmix.xs4all.nl
[mailto:owner-synth-diy at dropmix.xs4all.nl] On Behalf Of Grant Richter
Sent: Tuesday, February 17, 2004 10:37 AM
To: Rainer Buchty; Andalong Dudigual
Cc: synth-diy at dropmix.xs4all.nl
Subject: Re: [sdiy] poly?

> Now for the fake solutions:
> 
> The cheapest way to turn a monophonic synth into some sort of 
> polyphonic machine is the arpeggiator. Works nicely with chords, but 
> is rather unusable for true polyphonic play.

My electronic music teacher (Dr. August Wagner) pointed out that echo
devices like the Echoplex allow you to achieve polyphony with a
mono-synth by layering in real time. It was one reason they were so
popular to use with early synths. 

He also pointed out that electronic instruments are the only ones
without inherent acoustic ambience, which all acoustic instruments have.
Hence electronic reverb is practically an essential for imitative
synthesis.

This was in response to my comment that using effects units with
synthesizers was "cheating". His response was that, not only was it not
"cheating", but imperative for the above reasons.

I know that is not what the original poster intended. But I thought they
were rather keen insights on his part (and for 1976). He died very
young, and I try to keep his memory alive in some small way.