add minimal PhISEM (Shaker) physical model

[lfesp]

This PR adds a minimal implementation of Perry Cook's PhISEM model, a statistically-derived physical model for particle synthesis.

Issue: this implementation is not complete, as there is an issue with the BiQuad filtering. However, switching to a tSVF as the model's resonator works correctly.

[spiricom]

Thanks, Liam! I’ll check this over. I need to think about whether the resonant filter should be included in the model or be a separate thing

Composer, musician, instrument designer scattershot.org * @snyderphonics

On Fri, Mar 29, 2024 at 2:17 PM Liam Esparraguera @.***> wrote:

@.**** commented on this pull request.

In leaf/Src/leaf-physical.c https://github.com/spiricom/LEAF/pull/18#discussion_r1544747767:

• p->soundDecay = decay; +}

+void tShaker_setProbability (tShaker *const pm, Lfloat prob) +{

• _tShaker* p = *pm;
• // set gain correction for number of particles
• float nBeans = 1024.0 * prob;
• float gain = (log(nBeans) / log(4.)) * (40/nBeans);
• p->probability = prob;
• p->gain = gain; +}

+void tShaker_setResonance (tShaker *const pm, Lfloat freq, Lfloat radius) {

I think that getting rid of the resonator field entirely, and instead returning the direct output p->level * tNoise_tick(&p->noise) on each tick, would make for a better API. This is fine because there's no feedback.

If we let the user handle the filtering/resonance, they can set up and modify as many resonating stages as they like (which is used on almost all of Cook's PhISEM presets), and we don't have to expose a wrapper to tBiQuad_setResonance for tShaker.

— Reply to this email directly, view it on GitHub https://github.com/spiricom/LEAF/pull/18#pullrequestreview-1969261369, or unsubscribe https://github.com/notifications/unsubscribe-auth/AAGAY7GIYORFUPHAFNP2RJDY2WV2XAVCNFSM6AAAAABFOYBG2GVHI2DSMVQWIX3LMV43YUDVNRWFEZLROVSXG5CSMV3GSZLXHMYTSNRZGI3DCMZWHE . You are receiving this because you are subscribed to this thread.Message ID: @.***>