[sdiy] VCO - sine output - why bother?

[Mattias Rickardsson]

On 29 August 2016 at 20:39, Donald Tillman <don at till.com> wrote:
>> On Aug 27, 2016, at 9:21 AM, Mattias Rickardsson <mr at analogue.org> wrote:
>>>
>>> Subtracting the sine waveform only works as intended when the phase of the sine wave is aligned with the fundamental component of the other chosen waveform.
>>>
>>> But the standard way of deriving a triangle and sine wave from a sawtooth core (ie., an absolute value circuit) leaves the phase of the triangle and sine 90 degrees from the sawtooth.  So one can't actually subtract out the fundamental in that situation.
>>
>> That would be nifty, but I see the subtraction of the fundamental as a
>> minor feature after all.
>
> I disagree; I think it's a very musically useful feature.

Hopefully I'm wrong! Sorry for my quick dismissal of the idea, I'm not
heavily into additive-ish Buchla-esque synthesis (yet) so I didn't
trig immediately.

> On an acoustic guitar the fundamental often fades out faster than the harmonics, on some notes much more significantly than others.  Same with chimes.  This happens a lot with acoustic instruments.

Is it so? I wouldn't have guessed on a string instrument, but it's
true that more metallic sounds tend to have energy transferred from
lower to higher frequency modes over time.

>> In fact, I'd say it's often useful to have
>> the sine wave unaligned with the fundamental by 90 degrees or so.
>> Because in a Yamaha CS-like filtering scenario, when filtering out the
>> fundamental (plus some more low harmonics) and adding a sine wave - if
>> the filter has an envelope or LFO (or is tweaked by the user) that
>> sometimes brings back the fundamental, then it will not collide and
>> add as strongly with the sine wave. They will work more together,
>> ensuring the fundamental is there, but not doubling the amplitude of
>> it when both sources are present.
>
> It depends on the phase shifts introduced by the filter, which depends on the number of poles, and will be changing a lot as we're dynamically tuning it, and the phase of the complex wave relative to the sine wave, which will probably be different with a square or sawtooth wave.  There are a lot of dependencies there.

Very true! This would have to be ensured if the effect is supposed to
be used. I'm not sure how Yamaha's synths perform in this regard.

/mr