hyperframes-media

Files (1)

• SKILL.md

SKILL.md (first lines)

# HyperFrames Media

## Text-to-Speech (TTS)

```html
<div data-hf-clip
     data-hf-tts="Welcome to HyperFrames."
     data-hf-tts-voice="en-US-neural-f1"
     data-hf-tts-speed="1.0"
     data-hf-tts-emotion="neutral"
     data-hf-start="1s">
</div>
```

| Attribute | Values |
|---|---|
| `data-hf-tts-voice` | voice ID |
| `data-hf-tts-speed` | 0.5–2.0 |
| `data-hf-tts-pitch` | -20 to +20 semitones |
| `data-hf-tts-emotion` | `neutral`, `cheerful`, `sad`, `excited`, `calm` |

### Voice IDs

| ID | Lang | Gender |
|---|---|---|
| `en-US-neural-f1` | en-US | F |
| `en-US-neural-m1` | en-US | M |
| `en-GB-neural-f1` | en-GB | F |
| `en-AU-neural-m1` | en-AU | M |
| `es-ES-neural-f1` | es-ES | F |
| `fr-FR-neural-f1` | fr-FR | F |
| `de-DE-neural-m1` | de-DE | M |
| `ja-JP-neural-f1` | ja-JP | F |
| `zh-CN-neural-f1` | zh-CN | F |

## Transcription

```bash
npx hyperframes transcribe assets/interview.mp4 --out assets/interview.vtt
# --language en  --format vtt|srt|json  --timestamps word
```

## Background Removal

```bash
npx hyperframes bg-remove assets/speaker.jpg --out assets/speaker-nobg.png
npx hyperframes bg-remove assets/speaker.mp4 --out assets/speaker-nobg.webm --model quality
```

## Audio Ducking

```html
<audio data-hf-clip data-hf-start="0s" src="assets/bg-music.mp3"
       data-hf-duck="0.15"
       data-hf-duck-attack="0.3s"
       data-hf-duck-release="0.6s" />
```

## Waveform Visualization

```html
<div data-hf-clip data-hf-start="0s" data-hf-duration="10s"
     data-hf-waveform="assets/music.mp3"
     data-hf-waveform-color="#58a6ff"
     data-hf-waveform-style="bars"
     style="width:800px;height:120px;position:absolute;bottom:40px;left:560px;">
</div>
```
Styles: `bars`, `line`, `circle`, `dots`

From christiansalazar-wk/ai-lore, MIT.

More from christiansalazar-wk/ai-lore