sonicLAB has released version 2.0 of SSNN, a live audio processor plugin which puts a live spiking neural network — 960 neurons across 32 layers — at the center of the processing path, driving eight synthesis engines in real time with each spike.
Version 2 brings the following improvements:
- Layers Control: Adjust the active network depth and the length of its recurrent feedback loop.
- Time Control: Shift captured spectral frames between layers to reshape the network’s temporal structure.
- New Neural Buffer FM: Transform captured audio through audio-rate time and pitch modulation, shaped by neural voltages, connection weights and spikes.
- Improved Synaptic FM: More responsive neural modulation, hi-res depth scaling and better voice-state handling.
- Improved headroom between sections: Avoiding premature hard clipping and offering more responsiveness to DSP sections to input content.
- Four individual voice outputs: It routes the four NN columns to separate DAW tracks for independent mixing and processing.
- Rebalanced neural feedback:The recurrent return now operates at 50% strength.
- Improved project recall: Cancellation of stale background updates and correction of false retriggers.
- Updated NNnotes companion: Corrects MIDI-effect declarations and improves AU compatibility.
- Revised manual: Updated synthesis descriptions, network controls and DAW-routing instructions.
SSNN v2 is a free update for existing users. It is available in VST3 and AU formats for macOS and Windows systems. SSSN v2 is priced at £89 GBP and can be purchased at sonicLAB store.
YouTube/aVwc_XtflWs
YouTube/akx37fk3jFs
Read More