code/ │ ├── on-this-site/ │ │ │ ├── NKOAPP — voice synthesis system │ │ ├── nkoapp · python, praat, vocaltractlab │ │ └── system overview v2 │ │ │ └── Automatergon — generative text tool │ ├── JpKrNe · 2d slider — current │ ├── original │ ├── JK · 1d slider │ ├── JpKrNe · mouth suites │ └── shadow image │ └── repositories/ ↗ github.com/hogobogobogo ├── VIS — Vocal Insistunit Synthesizer · python │ └── github.com/hogobogobogo/VIS └── Audio-2-Tract · python └── github.com/hogobogobogo/audio2tract
→ nkoapp — voice synthesis system
An advanced articulatory speech synthesis system for creating natural and unnatural vocalizations. Built with Python, Praat, and VocalTractLab, NKOAPP allows precise control over vocal tract shapes and glottis properties to explore the boundaries between human and inhuman voice.
→ automatergon japkone
A generative text tool based on Greta Monach's original work, branching out further on Emiel Van Miltenburg's implementation. Customizable phonotactic rules down to the onset, nucleus, and coda. Change grid sizes, color coding and each Automatergon can be downloaded in PDF, XML and TXT formats.
→ vis — vocal insistunit synthesizer
A direct synthesis insitunit generator for VocalTractLab. Vocal tract motor sequences are built from scratch — no audio input — out of source-target pair gestures shaped by oscillator and stochastic functions. The resulting base phrase acts as a template for controlled variations, exported as VTL tract-sequence files ready for synthesis. Python.
→ audio-2-tract
Vocal tract parameter manipulation tools: audio files are converted into articulatory parameters and then transformed. Patches cover plain audio-to-speech conversion, morphing between two audio sources, and segmentation with time-stretching via FluCoMa novelty detection. Python.