[sdiy] Waveform phases and hard sync, sawtooth vs. triangle?

[Mattias Rickardsson]

Hi Florian,

Den fre 22 juni 2018 13:10Florian Anwander <fanwander at mnet-online.de> skrev:

> Hi Mattias
>
> Just two thoughts:
>
> 1.)
> > When you make drum synthesis, FM synthesis or even some regular bass
> > sounds, it's often important that the oscillator(s) behaves
> > identically each time the sound is started.
> If you start at the peak of the triangle, then you will have something
> like the overtones of a saw for the first half wave.


Yes... It's an impulse with all frequencies, since it's not repeating, so
not really overtones.

This seems to be
> unwanted, but it corresponds to the sound behaviour of all physical
> drums. They all have a bright  attack sound when the stick hits the skin.
> If you don't want this, then simply add something that acts like a VCA
> with an inverted decay envelope (like the autobend in the early Roland
> SH-synths), where the attacktime corresponds to the VCO frequency.
>

For drum synthesis I prefer having a customizable tick component and noise
component added to the oscillators, which themselves should be nice and
mellow. One can also add a quick pitch sweep at the very start in order to
make more or less of a tick from the oscillators.

The VCA is often there as well, but can be an issue - if it doesn't open
quick enough (for instance if it comes from a filtered DAC mux) your
oscillator disconinuity gets harmed in a more or less controllable way.
YMMV.

2.)
> > You need a way to start the tri/sin at the zero crossing.
>
> How do triangle core VCOs start after the power on? Is there a real
> reset function for triangle core VCOs?


No, there isn't one needed for its normal function.

At power on, the integrator cap starts uncharged. So the VCO starts in the
middle or at its peak depending on whether it's a bipolar or unipolar
triangle core. But that's not important, since it's not related to sync...
Unless you have one VCO per sound in your song, and power each of them on
when it's time for its sound to play. ;-)

The sync-behaviour in the CEM3340
> simply switches the load/unload "direction" of the VCO-core. So it does
> not(!) do a reset, like the hard sync does in a saw core VCO.
>

Right.
Btw, was this "reverse sync" designed with through-zero FM in mind?

/mr

>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://synth-diy.org/pipermail/synth-diy/attachments/20180623/3cda169b/attachment.htm>