REAKTOR COMMUNITY

DESCRIPTION

This synth is part of an ongoing effort to understand the 909 kick circuit. The key pieces from the circuit analysis are:

* The use of the triangle oscillator through a diode limiter
* The phase reset of the oscillator
* A DC bias in the oscillator output
* The timing of the three envelopes
* The attack noise filtering
* The inclusion of the gate signal in the attack noise

I also added a smoothing decay to the output of the enveloped oscillator to avoid digital artifacts during the phase reset.

The three 909 presets were made by comparing the output of an unprocessed 909 to the kick synth. It was surprising to hear that the original 909 oscillator contained a lot of harmonics, even after the diode saturation.

In general, the efforts to emulate the 909 are only partly about recreating the original. The other motivation is to better understand the ideas behind the original synth, and to think about how these ideas can be extended in a digital synth like Reaktor. To this end, the synth provides a range of options for tweaking and creating new sounds.

Lastly, I left the scopes in place to show how the envelope shapes and outputs change.

I'm continually working on this, so I'll push updates as I improve things. Hope you enjoy :)

TODOs:

- The tanh saturator is a decent approximation of the diode limiters, but I think an improved saturation model would get us even closer to the timbre of the original 909. The Hyperbolic Saturator was a good choice as well.

- The phase reset of the oscillator actually silences the output for 2ms in the original 909. This would seem to imply that the attack sound happens first, followed by the oscillator sound. I have each of these parts of the circuit built in LTSpice, but I have to combine them to see how they interact at the output. Currently, the attack sound and the oscillator play at the same time.

- The original 909 seems to have a more pronounced "thwack" or "thud" at the start of the wave. I think this has to do with how the phase is being reset. I'm currently starting the wave at 90 degrees, which seems to be how the 909 behaves, but I need to look into this more.