VoiceStudio

Operates the user's running VoiceStudio / debpalash VoiceStudio suite powered by k2-fsa OmniVoice and Coqui XTTS v2.

Capabilities & Workspaces

• Speech Synthesis: 600+ languages with neural diffusion and broadcast DSP mastering (broadcast, podcast, warm, bright, raw).
• Voice Design: Interactive attribute prompt generation (gender, age bracket, pitch, style, accent).
• Voice Cloning: Zero-shot reference voice conditioning from clean 3-10s audio samples.
• Model Context Protocol (MCP): Native tools for AI agents mounted on stdio or HTTP/SSE port 3900.
• Desktop Application: Bidirectional GUI in E:\Projects_D\LocalTTS_App\main.py.

Quick Execution via CLI

# OmniVoice 600+ languages with broadcast mastering
python E:/Projects_D/LocalTTS_App/cli_synthesizer.py --text "Hello world" --engine omnivoice --output out.wav

# Arabic speech synthesis
python E:/Projects_D/LocalTTS_App/cli_synthesizer.py --text "مرحباً بكم في استوديو الصوت" --lang ar --output arabic.wav

MCP Tools Exposed

• synthesize_speech(text, engine, language, speaker_or_instruct, speed, dsp_preset)
• clone_voice(ref_audio_path, speaker_id, text, engine)
• list_voices(engine)
• list_dsp_presets()
• check_system_status()

Attribution

Built on research and open-source foundations from Palash Deb (debpalash/VoiceStudio), k2-fsa (k2-fsa/OmniVoice), Coqui AI (coqui-ai/TTS), and maintained by Osama Alzahrani.