Labyricorn/voicebox
1
0
Fork 0
mirror of https://github.com/jamiepine/voicebox.git synced 2026-09-19 14:50:38 -07:00
Code Issues Packages Projects Releases Wiki Activity

docs

Browse Source
This commit is contained in:
James Pine
2026-03-16 05:11:21 -07:00
parent e16cc42d53
commit 192979a762
38 changed files with 1645 additions and 2124 deletions
Download Patch File Download Diff File Expand all files Collapse all files
docs/content/docs/overview/creating-voice-profiles.mdx
+20 -20
View File
@@ -28,32 +28,32 @@ Voice profiles are the foundation of voice cloning in Voicebox. This guide cover
### Ideal Sample Characteristics
<CardGroup cols={2}>
<Card title="Duration" icon="clock">
<Cards>
<Card title="Duration">
**10-30 seconds**
Too short: Poor quality
Too long: Unnecessary
</Card>
<Card title="Clarity" icon="volume">
<Card title="Clarity">
**Clear speech**
No background noise
No music or overlapping voices
</Card>
<Card title="Quality" icon="sparkles">
<Card title="Quality">
**High fidelity**
44.1kHz or 48kHz sample rate
Minimal compression
</Card>
<Card title="Content" icon="microphone">
<Card title="Content">
**Natural speech**
Conversational tone
Complete sentences
</Card>
</CardGroup>
</Cards>
### File Formats
@@ -63,9 +63,9 @@ Supported formats:
- **M4A** - Acceptable
- **FLAC** - Lossless alternative
<Tip>
<Callout type="info">
Use WAV for best results. Avoid heavily compressed formats.
</Tip>
</Callout>
## Recording Tips
@@ -108,20 +108,20 @@ Adding multiple samples can significantly improve quality:
### Why Multiple Samples?
<CardGroup cols={2}>
<Card title="Robustness" icon="shield">
<Cards>
<Card title="Robustness">
Model learns a more complete representation
</Card>
<Card title="Versatility" icon="palette">
<Card title="Versatility">
Handles different speaking styles better
</Card>
<Card title="Quality" icon="star">
<Card title="Quality">
Reduces artifacts and improves naturalness
</Card>
<Card title="Consistency" icon="check">
<Card title="Consistency">
More reliable across different texts
</Card>
</CardGroup>
</Cards>
### Sample Variety
@@ -144,9 +144,9 @@ Consider adding samples with:
- Phone call quality (if needed)
- Room acoustics
<Warning>
<Callout type="warn">
All samples should be from the **same speaker**. Mixing voices will produce poor results.
</Warning>
</Callout>
## Processing Existing Audio
@@ -286,11 +286,11 @@ The emotional tone of samples affects generation:
## Next Steps
<CardGroup cols={2}>
<Card title="Generate Speech" icon="waveform" href="/overview/generating-speech">
<Cards>
<Card title="Generate Speech" href="/overview/generating-speech">
Use your profile to generate speech
</Card>
<Card title="Build Stories" icon="film" href="/overview/building-stories">
<Card title="Build Stories" href="/overview/building-stories">
Create multi-voice narratives
</Card>
</CardGroup>
</Cards>
docs/content/docs/overview/generating-speech.mdx
+2 -2
View File
@@ -43,9 +43,9 @@ Use formatting to suggest emphasis:
- Bold for strong emphasis: "This is **very** important"
```
<Note>
<Callout type="info">
The model interprets these hints but results may vary.
</Note>
</Callout>
## Advanced Features
docs/content/docs/overview/generation-history.mdx
+12 -12
View File
@@ -9,20 +9,20 @@ Voicebox keeps a complete history of all generated audio, making it easy to find
## Features
<CardGroup cols={2}>
<Card title="Full History" icon="clock">
<Cards>
<Card title="Full History" icon={<svg xmlns="http://www.w3.org/2000/svg" width="24" height="24" viewBox="0 0 24 24" fill="none" stroke="currentColor" strokeWidth="2" strokeLinecap="round" strokeLinejoin="round"><circle cx="12" cy="12" r="10"/><polyline points="12 6 12 12 16 14"/></svg>}>
Every generation is automatically saved
</Card>
<Card title="Search & Filter" icon="search">
<Card title="Search & Filter" icon={<svg xmlns="http://www.w3.org/2000/svg" width="24" height="24" viewBox="0 0 24 24" fill="none" stroke="currentColor" strokeWidth="2" strokeLinecap="round" strokeLinejoin="round"><circle cx="11" cy="11" r="8"/><path d="m21 21-4.3-4.3"/></svg>}>
Find by text, voice, or date
</Card>
<Card title="Re-generate" icon="rotate">
<Card title="Re-generate" icon={<svg xmlns="http://www.w3.org/2000/svg" width="24" height="24" viewBox="0 0 24 24" fill="none" stroke="currentColor" strokeWidth="2" strokeLinecap="round" strokeLinejoin="round"><path d="M3 12a9 9 0 0 1 9-9 9.75 9.75 0 0 1 6.74 2.74L21 8"/><path d="M21 3v5h-5"/><path d="M21 12a9 9 0 0 1-9 9 9.75 9.75 0 0 1-6.74-2.74L3 16"/><path d="M8 16H3v5"/></svg>}>
Regenerate any past generation with one click
</Card>
<Card title="Export" icon="download">
<Card title="Export" icon={<svg xmlns="http://www.w3.org/2000/svg" width="24" height="24" viewBox="0 0 24 24" fill="none" stroke="currentColor" strokeWidth="2" strokeLinecap="round" strokeLinejoin="round"><path d="M21 15v4a2 2 0 0 1-2 2H5a2 2 0 0 1-2-2v-4"/><polyline points="7 10 12 15 17 10"/><line x1="12" y1="15" x2="12" y2="3"/></svg>}>
Download individual or batch exports
</Card>
</CardGroup>
</Cards>
## Viewing History
@@ -54,20 +54,20 @@ Drag generations to the Stories Editor timeline.
## Search & Filter
<Tabs>
<Tab title="By Text">
<Tabs items={["By Text", "By Voice", "By Date"]}>
<Tab value="By Text">
Search for specific text content
```
"Hello world"
```
</Tab>
<Tab title="By Voice">
<Tab value="By Voice">
Filter by voice profile
```
Select from dropdown
```
</Tab>
<Tab title="By Date">
<Tab value="By Date">
Filter by date range
```
Last 7 days, Last 30 days, Custom range
@@ -83,6 +83,6 @@ History is stored locally:
- **Windows**: `%APPDATA%/com.voicebox.app/data/`
- **Linux**: `~/.config/com.voicebox.app/data/`
<Warning>
<Callout type="warn">
Deleting the data directory will remove all history. Export important files first.
</Warning>
</Callout>
docs/content/docs/overview/installation.mdx
+19 -19
View File
@@ -7,19 +7,19 @@ description: "Download and install Voicebox on macOS, Windows, or Linux"
Voicebox is available for macOS and Windows, with Linux builds coming soon.
<CardGroup cols={2}>
<Card title="macOS" icon="apple">
<Cards>
<Card title="macOS" icon={<svg xmlns="http://www.w3.org/2000/svg" width="24" height="24" viewBox="0 0 24 24" fill="none" stroke="currentColor" strokeWidth="2" strokeLinecap="round" strokeLinejoin="round"><path d="M12 2c-1.5 0-2.8.4-3.9 1.1A5.5 5.5 0 0 0 4 2.5C2.5 2.5 1 4 1 6c0 3.5 2.5 6 5 7.5C5 16 4 18 4 20c0 1.5.5 2.5 1.5 3C6.5 23.5 8 24 9.5 24c2 0 3.5-.5 5-2 1.5 1.5 3 2 5 2 1.5 0 3-.5 4-1 1-.5 1.5-1.5 1.5-3 0-2-1-4-2-6.5 2.5-1.5 5-4 5-7.5 0-2-1.5-3.5-3-3.5-.9 0-2.1.4-3.1 1.1A6.5 6.5 0 0 0 12 2Z"/></svg>}>
Download for Apple Silicon or Intel Macs
</Card>
<Card title="Windows" icon="windows">
<Card title="Windows" icon={<svg xmlns="http://www.w3.org/2000/svg" width="24" height="24" viewBox="0 0 24 24" fill="none" stroke="currentColor" strokeWidth="2" strokeLinecap="round" strokeLinejoin="round"><rect x="2" y="3" width="20" height="14" rx="2" ry="2"/><line x1="8" y1="21" x2="16" y2="21"/><line x1="12" y1="17" x2="12" y2="21"/></svg>}>
Download MSI installer or Setup executable
</Card>
</CardGroup>
</Cards>
### macOS
<Tabs>
<Tab title="Apple Silicon">
<Tabs items={["Apple Silicon", "Intel"]}>
<Tab value="Apple Silicon">
Download: [voicebox_aarch64.app.tar.gz](https://github.com/jamiepine/voicebox/releases/latest/download/voicebox_aarch64.app.tar.gz)
```bash
@@ -30,7 +30,7 @@ Voicebox is available for macOS and Windows, with Linux builds coming soon.
mv Voicebox.app /Applications/
```
</Tab>
<Tab title="Intel">
<Tab value="Intel">
Download: [voicebox_x64.app.tar.gz](https://github.com/jamiepine/voicebox/releases/latest/download/voicebox_x64.app.tar.gz)
```bash
@@ -45,13 +45,13 @@ Voicebox is available for macOS and Windows, with Linux builds coming soon.
### Windows
<Tabs>
<Tab title="MSI Installer">
<Tabs items={["MSI Installer", "Setup Executable"]}>
<Tab value="MSI Installer">
Download: [voicebox_x64_en-US.msi](https://github.com/jamiepine/voicebox/releases/latest/download/voicebox_x64_en-US.msi)
Double-click the MSI file and follow the installation wizard.
</Tab>
<Tab title="Setup Executable">
<Tab value="Setup Executable">
Download: [voicebox_x64-setup.exe](https://github.com/jamiepine/voicebox/releases/latest/download/voicebox_x64-setup.exe)
Run the executable and follow the installation wizard.
@@ -60,9 +60,9 @@ Voicebox is available for macOS and Windows, with Linux builds coming soon.
### Linux
<Note>
<Callout type="info">
Linux builds are coming soon. Currently blocked by GitHub runner disk space limitations.
</Note>
</Callout>
## First Launch
@@ -76,9 +76,9 @@ When you launch Voicebox for the first time:
3. **Backend Server** — The bundled Python server starts automatically
<Tip>
<Callout type="info">
First generation will be slower due to model downloads. Subsequent runs use cached models.
</Tip>
</Callout>
## System Requirements
@@ -95,9 +95,9 @@ When you launch Voicebox for the first time:
- **GPU:** CUDA-capable NVIDIA GPU (for faster generation)
- **Storage:** 10GB+ free space
<Note>
<Callout type="info">
CPU inference is supported but significantly slower than GPU. A CUDA-capable GPU is highly recommended for real-time workflows.
</Note>
</Callout>
## Verification
@@ -108,12 +108,12 @@ After installation, verify everything works:
3. Navigate to **Profiles** and create a test profile
4. Generate a short audio clip to verify the TTS engine works
<Check>
<Callout type="success">
If you see a green status indicator and can generate audio, you're all set!
</Check>
</Callout>
## Next Steps
<Card title="Quick Start Guide" icon="rocket" href="/overview/quick-start">
<Card title="Quick Start Guide" href="/overview/quick-start">
Create your first voice profile and generate speech
</Card>
docs/content/docs/overview/quick-start.mdx
+10 -10
View File
@@ -46,9 +46,9 @@ Voice profiles are the foundation of Voicebox. Each profile contains voice sampl
</Step>
</Steps>
<Tip>
<Callout type="info">
For best results, use clean audio with minimal background noise and consistent speaking tone.
</Tip>
</Callout>
## Step 2: Generate Speech
@@ -74,9 +74,9 @@ Now let's use your new voice profile to generate speech.
<Step title="Generate">
Click **Generate** and wait a few seconds
<Note>
<Callout type="info">
First generation may take longer due to model initialization. Subsequent generations will be faster.
</Note>
</Callout>
</Step>
<Step title="Play & Download">
@@ -114,20 +114,20 @@ The Stories Editor lets you create multi-voice narratives with a timeline-based
## What's Next?
<CardGroup cols={2}>
<Card title="Voice Cloning Guide" icon="microphone" href="/overview/creating-voice-profiles">
<Cards>
<Card title="Voice Cloning Guide" href="/overview/creating-voice-profiles">
Learn advanced techniques for high-quality voice cloning
</Card>
<Card title="API Integration" icon="code" href="/api-reference">
<Card title="API Integration" href="/api-reference">
Integrate Voicebox into your own applications
</Card>
<Card title="Stories Editor" icon="film" href="/overview/stories-editor">
<Card title="Stories Editor" href="/overview/stories-editor">
Master the multi-track timeline editor
</Card>
<Card title="Remote Mode" icon="server" href="/overview/remote-mode">
<Card title="Remote Mode" href="/overview/remote-mode">
Connect to a GPU server for faster generation
</Card>
</CardGroup>
</Cards>
## Tips for Success
docs/content/docs/overview/recording-transcription.mdx
+2 -2
View File
@@ -59,6 +59,6 @@ Automatic speech-to-text powered by OpenAI's Whisper model.
</Step>
</Steps>
<Tip>
<Callout type="info">
Transcription is useful for creating voice samples from existing audio or generating subtitles.
</Tip>
</Callout>
docs/content/docs/overview/remote-mode.mdx
+6 -6
View File
@@ -41,9 +41,9 @@ In Remote Mode, the Voicebox desktop app (running on your local machine) communi
uvicorn main:app --host 0.0.0.0 --port 17493
```
<Warning>
<Callout type="warn">
This exposes the server to your network. Use a firewall or VPN for security.
</Warning>
</Callout>
</Step>
<Step title="Open Firewall">
@@ -108,9 +108,9 @@ In Remote Mode, the Voicebox desktop app (running on your local machine) communi
## Security Considerations
<Warning>
<Callout type="warn">
The API currently has no authentication. Only use on trusted networks or with a VPN.
</Warning>
</Callout>
**Best Practices:**
- Use a VPN (WireGuard, Tailscale) instead of exposing to the internet
@@ -129,9 +129,9 @@ Expected performance on various GPUs:
| RTX 3060 | ~5-7s per 10 words |
| CPU (12-core) | ~20-30s per 10 words |
<Tip>
<Callout type="info">
A GPU with 8GB+ VRAM is recommended for best performance.
</Tip>
</Callout>
## Troubleshooting
docs/content/docs/overview/stories-editor.mdx
+6 -6
View File
@@ -9,20 +9,20 @@ The Stories Editor is a DAW-like timeline interface for creating multi-voice nar
## Features
<CardGroup cols={2}>
<Card title="Multi-Track Timeline" icon="timeline">
<Cards>
<Card title="Multi-Track Timeline">
Arrange multiple voice tracks in parallel
</Card>
<Card title="Inline Editing" icon="scissors">
<Card title="Inline Editing">
Trim and split clips directly in the timeline
</Card>
<Card title="Auto-Playback" icon="play">
<Card title="Auto-Playback">
Preview with synchronized playhead
</Card>
<Card title="Voice Mixing" icon="users">
<Card title="Voice Mixing">
Build conversations with multiple speakers
</Card>
</CardGroup>
</Cards>
## Creating a Story
docs/content/docs/overview/troubleshooting.mdx
+6 -6
View File
@@ -25,9 +25,9 @@ Windows SmartScreen may warn that the app is unrecognized.
- Click "More info"
- Click "Run anyway"
<Note>
<Callout type="info">
This is expected for unsigned applications. We're working on code signing for future releases.
</Note>
</Callout>
## Server Issues
@@ -163,9 +163,9 @@ This is expected behavior. The first generation downloads the Qwen3-TTS model (~
Settings → Generation → Use CPU instead of GPU
<Warning>
<Callout type="warn">
CPU generation is 5-10x slower but uses system RAM instead of VRAM.
</Warning>
</Callout>
</Accordion>
<Accordion title="Reduce Batch Size">
@@ -322,9 +322,9 @@ bun run tauri build
**Solutions:**
<Warning>
<Callout type="warn">
This will delete all your voice profiles and generation history. Export important profiles first if possible.
</Warning>
</Callout>
```bash
# macOS
docs/content/docs/overview/voice-cloning.mdx
+8 -8
View File
@@ -28,20 +28,20 @@ Voicebox uses **Qwen3-TTS** from Alibaba to achieve near-perfect voice cloning f
### Sample Quality
<CardGroup cols={2}>
<Card title="Do" icon="check">
<Cards>
<Card title="Do">
- Use 10-30 seconds of audio
- Clear, consistent speaking
- Minimal background noise
- Natural speaking pace
</Card>
<Card title="Don't" icon="xmark">
<Card title="Don't">
- Very short clips (< 5 seconds)
- Heavy background noise
- Music or overlapping voices
- Heavily processed audio
</Card>
</CardGroup>
</Cards>
### Multiple Samples
@@ -51,9 +51,9 @@ Adding multiple samples from the same speaker can improve quality:
- Different emotions (happy, serious)
- Different recording conditions
<Tip>
<Callout type="info">
The model will learn a more robust representation from diverse samples.
</Tip>
</Callout>
## Supported Languages
@@ -65,9 +65,9 @@ More languages coming soon.
## Limitations
<Warning>
<Callout type="warn">
Voice cloning should only be used with consent. Ensure you have permission to clone someone's voice.
</Warning>
</Callout>
- Quality depends on sample clarity
- Works best with consistent speaking tone