Text → MIDI on a patch-bay.
Prompt → GenerateMidi → roll audition.
This take uses a Bohlen–Pierce prompt —
non-octave tuning, offline render, Play = cache.
https://t.co/eP6WrIBBag
#GroovyUI#AIMusic#modularsynth#midillm#BohlenPierce
Fresh hop: Rescore Video.
AI video looks sharp. The music often doesn’t.
So we rescore on the patch: split → ACE-Step SFT cover → mux → A/B original vs new score.
https://t.co/d0Ym9l3VCk
#groovyui#aivideo#aimusic#acestep#modularsynth
Basic video support in #GroovyUI — on the patchbay. Everything runs locally. Still → GenerateVideo → Preview. Same graph as denoise, stems, TTS — wire it, A/B it, log timing / config / HW.
https://t.co/a1z86U4GTH
#aimusic#aivideo#LTX
What ran on this patch?
Provenance visualizer in GroovyUI: the audio hop chain, labeled human / AI by node type — not a black-box “% synthetic” meter.
https://t.co/yPypNNxfbG
#GroovyUI#modularsynth#AImusic#provenance#authorship
New in GroovyUI: offline ASR eval on the patch bay. Same script → Kokoro → Whisper small vs turbo. Drag-select two transcripts → Inspector A/B highlights WER/CER word diffs. Patch models. Compare. Decide.
https://t.co/nDInAqMErs
#GroovyUI#AIAudio#ModularSynth#ASR#WER
#GroovyUI is now on #ISMIR2026 site — a patchbay for #AIAudio, not a song button.
First listen on your material.
Between model repos and the DAW.
For people who already think in nodes.
https://t.co/ZWKhaaFedV
https://t.co/cfaE55o0Gs
#AIMusic#AISpeech#MusicTech
Same Tab reflex as in the IDE- propose, accept, keep the patch yours.
GroovyUI Tab to Grow: ghosts the next hops; Tab commits one module. Human in the loop on a modular patch-bay.
https://t.co/2KAvzwJOnv
#AIMusic#AISpeech#AIAudio#HumanInTheLoop#ModularSynth#GroovyUI
AI models as modules.
Demucs → Whisper → M2M100 on one patch.
Isolate. Transcribe. Translate.
Rewire any hop. Render offline. Hear the cache.
Inspect provenance. Share the JSON. Not a DAW. A patch-bay for local AI audio.
https://t.co/6oCMJW86Es
#AIaudio#ModularSynth#GroovyUI
AI voice as oscillator material.
Wire Prompt -> TTS -> Granulate. Shrink grain size, spray pitch, bloom width- the take liquifies on purpose.
Render offline. Hear the buffer. Rewire the JSON.
https://t.co/cVu0lk72hA
#ModularSynth#AIAudio#GroovyUI
Follow-up: GroovyUI is open source. Same patchbay enabling you to patch audio AI models like you would in modular synth. Not a DAW. Not a vibe-song slot machine.
https://t.co/fEJ9hUSGVR
https://t.co/ZWKhaaFedV
#AIaudio#ModularSynth#CreativeAI#GroovyUI
Local AI audio shouldn’t mean notebooks and final_FINAL2.wav. #GroovyUI: patchbay for AI audio. graph → install → render → hear cache → licenses → JSON/WAV. Not a DAW. Two design-partner slots. Comment pilot / DM what breaks. #AIaudio#ModularSynth
First post on the ONE Systemics Substack- Studio Sample #1: From Knobs to Prompts- and What's Still Missing (#NeurIPS, #CreativeAI, #MCP2OSC) English on one side. UDP on the other. The bridge is MCP2OSC. https://t.co/Tst0oeUMvg