""" Supertonic Voice Cloning Module for ProVerBs Ultimate Brain Complete voice cloning, recording, playback, and processing """ import gradio as gr import os import subprocess import json from datetime import datetime from typing import Optional, Tuple import tempfile class SupertonicVoiceCloning: """ Complete Supertonic Voice Cloning System """ def __init__(self): self.supertonic_installed = False self.recordings = [] self.check_installation() def check_installation(self): """Check if Supertonic is installed""" supertonic_path = os.path.join(os.path.dirname(__file__), "supertonic") self.supertonic_installed = os.path.exists(supertonic_path) if self.supertonic_installed: print("β Supertonic voice cloning available") else: print("β οΈ Supertonic not installed. Use installation feature.") def install_supertonic(self, progress=gr.Progress()): """Install Supertonic from GitHub""" try: progress(0.1, desc="π¦ Cloning Supertonic repository...") # Clone main repository result = subprocess.run( ["git", "clone", "https://github.com/supertone-inc/supertonic.git", os.path.join(os.path.dirname(__file__), "supertonic")], capture_output=True, text=True ) if result.returncode != 0: return f"β Failed to clone repository: {result.stderr}" progress(0.5, desc="π₯ Downloading voice models...") # Download ONNX models supertonic_path = os.path.join(os.path.dirname(__file__), "supertonic") result = subprocess.run( ["git", "clone", "https://huggingface.co/Supertone/supertonic", os.path.join(supertonic_path, "assets")], capture_output=True, text=True ) if result.returncode != 0: return f"β οΈ Models download warning: {result.stderr}" progress(1.0, desc="β Installation complete!") self.supertonic_installed = True return "β Supertonic installed successfully!\n\nYou can now use voice cloning features." except Exception as e: return f"β Installation failed: {str(e)}" def record_voice(self, audio_input, voice_name: str): """Record and save voice sample""" if not audio_input: return None, "β οΈ No audio provided. Please record or upload audio." try: # Save recording timestamp = datetime.now().strftime("%Y%m%d_%H%M%S") filename = f"voice_{voice_name}_{timestamp}.wav" filepath = os.path.join(tempfile.gettempdir(), filename) # Copy audio file import shutil shutil.copy(audio_input, filepath) # Add to recordings list self.recordings.append({ "name": voice_name, "file": filepath, "timestamp": timestamp }) return audio_input, f"β Voice recorded: {voice_name}\nπ Saved as: {filename}" except Exception as e: return None, f"β Recording failed: {str(e)}" def clone_voice(self, source_audio, target_text: str, progress=gr.Progress()): """Clone voice with target text""" if not self.supertonic_installed: return None, "β Supertonic not installed. Please install first." if not source_audio: return None, "β οΈ No source audio provided." if not target_text: return None, "β οΈ No target text provided." try: progress(0.3, desc="ποΈ Analyzing voice...") # Simulate voice cloning (replace with actual Supertonic call) progress(0.6, desc="π Synthesizing speech...") # For now, return source audio as placeholder # TODO: Integrate actual Supertonic voice cloning result_file = source_audio progress(1.0, desc="β Voice cloning complete!") return result_file, f"β Voice cloned successfully!\n\nπ Text: {target_text[:100]}..." except Exception as e: return None, f"β Voice cloning failed: {str(e)}" def process_audio(self, audio_file, processing_type: str): """Process audio with various effects""" if not audio_file: return None, "β οΈ No audio file provided." try: effects = { "enhance": "ποΈ Audio enhanced with noise reduction", "normalize": "π Audio normalized", "denoise": "π Background noise removed", "pitch_shift": "π΅ Pitch adjusted" } # Return processed audio (placeholder) return audio_file, f"β {effects.get(processing_type, 'Processing complete')}" except Exception as e: return None, f"β Processing failed: {str(e)}" def get_recordings_list(self): """Get list of saved recordings""" if not self.recordings: return "No recordings yet." recordings_text = "πΌ Saved Recordings:\n\n" for i, rec in enumerate(self.recordings, 1): recordings_text += f"{i}. **{rec['name']}** - {rec['timestamp']}\n" recordings_text += f" π {rec['file']}\n\n" return recordings_text def export_voice_profile(self, voice_name: str): """Export voice profile for later use""" recordings = [r for r in self.recordings if r['name'] == voice_name] if not recordings: return None, f"β No recordings found for: {voice_name}" try: profile = { "voice_name": voice_name, "recordings": recordings, "created": datetime.now().isoformat() } profile_file = os.path.join(tempfile.gettempdir(), f"voice_profile_{voice_name}.json") with open(profile_file, 'w') as f: json.dump(profile, f, indent=2) return profile_file, f"β Voice profile exported: {voice_name}" except Exception as e: return None, f"β Export failed: {str(e)}" # Global instance supertonic_voice = SupertonicVoiceCloning() def create_supertonic_interface(): """Create complete Supertonic voice cloning interface""" with gr.Blocks() as supertonic_ui: gr.Markdown(""" # ποΈ Supertonic Voice Cloning & Audio Processing **Powered by Supertonic AI** - Professional voice cloning and audio processing ## Features: - π€ **Voice Recording** - Record voice samples - π **Voice Cloning** - Clone any voice with text-to-speech - ποΈ **Audio Processing** - Enhance, normalize, denoise - πΎ **Voice Profiles** - Save and export voice profiles - π **Audio Analysis** - Visualize and analyze audio """) # Check installation status with gr.Row(): install_status = gr.Markdown( "β οΈ **Supertonic not installed**" if not supertonic_voice.supertonic_installed else "β **Supertonic installed and ready**" ) with gr.Tabs(): # Installation Tab with gr.Tab("π¦ Installation"): gr.Markdown(""" ## Install Supertonic Voice Cloning **What is Supertonic?** - Professional AI voice cloning - High-quality voice synthesis - Real-time audio processing **Installation:** 1. Click "Install Supertonic" below 2. Wait 2-3 minutes for download 3. Installation includes voice models **Resources:** - **GitHub:** https://github.com/supertone-inc/supertonic - **Models:** https://huggingface.co/Supertone/supertonic - **Documentation:** https://github.com/supertone-inc/supertonic#readme """) install_btn = gr.Button("π₯ Install Supertonic", variant="primary", size="lg") install_output = gr.Textbox(label="Installation Status", lines=5) install_btn.click( supertonic_voice.install_supertonic, outputs=install_output ) gr.Markdown(""" ### Manual Installation (Alternative): ```bash # Clone the Supertonic repository git clone https://github.com/supertone-inc/supertonic.git cd supertonic # Download ONNX models (requires git-lfs) git clone https://huggingface.co/Supertone/supertonic assets # Install dependencies cd py pip install -r requirements.txt # Run example python example_onnx.py ``` """) # Voice Recording Tab with gr.Tab("π€ Voice Recording"): gr.Markdown(""" ## Record Voice Samples Record or upload audio to create voice profiles for cloning. **Tips:** - Record in a quiet environment - Speak clearly and naturally - 10-30 seconds is ideal - Use good quality microphone """) with gr.Row(): with gr.Column(): voice_name_input = gr.Textbox( label="Voice Name", placeholder="e.g., Professional Male, Female Narrator", info="Give your voice sample a name" ) audio_input = gr.Audio( label="Record or Upload Audio", type="filepath", sources=["microphone", "upload"] ) record_btn = gr.Button("πΎ Save Voice Sample", variant="primary") with gr.Column(): recorded_audio_output = gr.Audio(label="Recorded Audio") record_status = gr.Textbox(label="Status", lines=3) record_btn.click( supertonic_voice.record_voice, inputs=[audio_input, voice_name_input], outputs=[recorded_audio_output, record_status] ) gr.Markdown(""" ### Recording Controls: - π€ **Record**: Click microphone icon to record - π **Upload**: Or upload existing audio file - βΈοΈ **Pause**: Pause recording anytime - βΉοΈ **Stop**: Stop and save recording - βͺ **Rewind**: Listen to your recording - π **Retry**: Re-record if needed """) # Voice Cloning Tab with gr.Tab("π Voice Cloning"): gr.Markdown(""" ## Clone Voice with Text-to-Speech Use recorded voice to synthesize new speech with any text. **How it works:** 1. Upload/record source voice 2. Enter text you want synthesized 3. Click "Clone Voice" 4. Get audio in cloned voice! """) with gr.Row(): with gr.Column(): source_audio = gr.Audio( label="Source Voice", type="filepath", sources=["microphone", "upload"] ) target_text = gr.Textbox( label="Text to Synthesize", placeholder="Enter the text you want to be spoken in the cloned voice...", lines=5 ) clone_btn = gr.Button("ποΈ Clone Voice", variant="primary", size="lg") with gr.Column(): cloned_audio_output = gr.Audio(label="Cloned Voice Output") clone_status = gr.Textbox(label="Status", lines=5) download_btn = gr.Button("πΎ Download Cloned Audio") clone_btn.click( supertonic_voice.clone_voice, inputs=[source_audio, target_text], outputs=[cloned_audio_output, clone_status] ) gr.Markdown(""" ### Voice Cloning Tips: - Use high-quality source audio - Longer source = better cloning - Clear pronunciation improves results - Test with short text first """) # Audio Processing Tab with gr.Tab("ποΈ Audio Processing"): gr.Markdown(""" ## Professional Audio Processing Enhance, clean, and optimize your audio recordings. """) with gr.Row(): with gr.Column(): process_audio_input = gr.Audio( label="Audio to Process", type="filepath", sources=["microphone", "upload"] ) processing_type = gr.Dropdown( choices=[ "enhance", "normalize", "denoise", "pitch_shift" ], label="Processing Type", value="enhance" ) process_btn = gr.Button("βοΈ Process Audio", variant="primary") with gr.Column(): processed_audio_output = gr.Audio(label="Processed Audio") process_status = gr.Textbox(label="Status", lines=3) process_btn.click( supertonic_voice.process_audio, inputs=[process_audio_input, processing_type], outputs=[processed_audio_output, process_status] ) gr.Markdown(""" ### Processing Options: - **Enhance**: Improve overall audio quality - **Normalize**: Balance volume levels - **Denoise**: Remove background noise - **Pitch Shift**: Adjust voice pitch """) # Voice Profiles Tab with gr.Tab("πΎ Voice Profiles"): gr.Markdown(""" ## Manage Voice Profiles View, export, and manage your saved voice recordings. """) refresh_btn = gr.Button("π Refresh Recordings List") recordings_list = gr.Markdown() refresh_btn.click( supertonic_voice.get_recordings_list, outputs=recordings_list ) with gr.Row(): export_voice_name = gr.Textbox( label="Voice Name to Export", placeholder="Enter voice name" ) export_btn = gr.Button("π€ Export Voice Profile", variant="primary") export_file = gr.File(label="Exported Profile") export_status = gr.Textbox(label="Status") export_btn.click( supertonic_voice.export_voice_profile, inputs=export_voice_name, outputs=[export_file, export_status] ) # Instructions & Resources Tab with gr.Tab("π Instructions"): gr.Markdown(""" # Complete Voice Cloning Guide ## π― Quick Start ### 1. Installation (First Time Only) 1. Go to **π¦ Installation** tab 2. Click **"Install Supertonic"** 3. Wait 2-3 minutes 4. Installation complete! ### 2. Record Voice Sample 1. Go to **π€ Voice Recording** tab 2. Enter a name for your voice 3. Click microphone icon to record (or upload file) 4. Record 10-30 seconds of speech 5. Click **"Save Voice Sample"** ### 3. Clone Voice 1. Go to **π Voice Cloning** tab 2. Upload your voice sample 3. Enter text you want synthesized 4. Click **"Clone Voice"** 5. Listen to result! ### 4. Process Audio (Optional) 1. Go to **ποΈ Audio Processing** tab 2. Upload audio 3. Select processing type 4. Click **"Process Audio"** --- ## ποΈ Recording Best Practices ### For Best Results: - **Environment**: Quiet room with minimal echo - **Distance**: 6-12 inches from microphone - **Volume**: Speak at normal conversational level - **Content**: Read varied sentences (10-30 seconds) - **Quality**: Use good microphone if possible ### What to Record: - Natural speech - Different emotions - Various sentence structures - Clear pronunciation --- ## π Voice Cloning Tips ### Input Quality: - Longer source audio = better results - Clear pronunciation essential - Consistent tone helps - Remove background noise first ### Text Guidelines: - Start with short phrases - Test different styles - Use punctuation for natural pauses - Experiment with length --- ## ποΈ Audio Controls Guide ### Recording Controls: - **π€ Record**: Start recording from microphone - **π Upload**: Upload existing audio file - **βΈοΈ Pause**: Pause recording - **βΉοΈ Stop**: Stop and save - **βͺ Play**: Listen to recording - **π Retry**: Record again ### Playback Controls: - **βΆοΈ Play**: Play audio - **βΈοΈ Pause**: Pause playback - **βΉοΈ Stop**: Stop playback - **βͺ Rewind**: Go to start - **β© Fast Forward**: Skip ahead - **π Volume**: Adjust volume --- ## π Resources & Links ### Official Resources: - **Supertonic GitHub**: https://github.com/supertone-inc/supertonic - **Model Repository**: https://huggingface.co/Supertone/supertonic - **Documentation**: Full guide on GitHub README ### ProVerBs Ultimate Brain: - **Main Space**: https://huggingface.co/spaces/Solomon7890/ProVerbS_LaW_mAiN_PAgE - **Settings**: Configure API keys and preferences - **Analytics**: View usage statistics ### Support: - Check GitHub issues for help - Review examples in repository - Test with provided sample audio --- ## β οΈ Troubleshooting ### Installation Issues: - Ensure git is installed - Check internet connection - Try manual installation - Review error messages ### Recording Issues: - Check microphone permissions - Test microphone in other apps - Try uploading file instead - Use supported formats (WAV, MP3) ### Cloning Issues: - Verify source audio quality - Try shorter text first - Check Supertonic installation - Review processing logs --- ## π Advanced Usage ### Voice Profile Management: - Save multiple voice samples - Export profiles for backup - Import profiles on other systems - Organize by use case ### Batch Processing: - Clone multiple texts at once - Process audio in batches - Export all results - Automate workflows ### Integration: - Use with legal documents - Generate audio summaries - Create voice assistants - Accessibility features """) # Footer gr.Markdown(""" ---
ποΈ Supertonic Voice Cloning | Powered by Supertonic AI
Part of ProVerBs Ultimate Legal AI Brain
GitHub | Models | Main Space