Live jam with Stable Audio 3 inside my DAW | Session 03 | inc. hardware

Live improvisation recorded with Obsidian Neural V2.5.3, a real-time AI music generation VST3 plugin for live performance. This session is built primarily around Stable Audio 3. Audio generation on provider nodes across 9 specialized AI engines: stabilityai/stable-audio-3-medium — Latest generation from Stability AI, FlowMatching architecture stabilityai/stable-audio-open-1.0 — Full-mix textures and drum loops (80–160 BPM) RoyalCities/Foundation-1 — Melodic and harmonic phrasing (100–150 BPM) adlb/Audialab_EDM_Elements — EDM leads, supersaws and plucks (100–150 BPM) RoyalCities/RC_Infinite_Pianos — Grand and electric piano performances (100–150 BPM) RoyalCities/Vocal_Textures_Main — Choral and atmospheric vocal progressions (100–150 BPM) santifiorino/SAO-Instrumental-Finetune — Lofi jazz rap and indie stems (75–160 BPM) gab-gdp/StableBeaT — Trap beats and 808 grooves (75–160 BPM) atoof/gluten_v1 — Melodic trap and wavy motifs (90–160 BPM) All loops generated on the fly, triggered via MIDI, time-stretched to DAW tempo using Signalsmith Stretch. Plugin: https://github.com/innermost47/ai-dj Provider network: https://github.com/innermost47/obsidi... Website: https://obsidian-neural.com Developed by InnerMost47 (Recorded with additional hardware instruments: Behringer TD-3, Behringer Pro-1, Akai APC 20, Arturia MiniLab, CME UF6, Korg NanoKontrol 2 & Obsidian Neural Controller (custom Flutter app built for the VST))