transcribe

# transcribe - Speech to Text

Transcribe audio file to text using Google Cloud Speech-to-Text API.

## Usage

```bash
echo "recording.wav" | transcribe
echo "/path/to/audio.wav" | transcribe --lang es-ES
```

## Input

Accepts audio file path via stdin. Supports WAV files with LINEAR16 encoding.

## Arguments

| Flag | Default | Description |
|------|---------|-------------|
| `--lang` | en-US | Language code for transcription |
| `--rate` | 16000 | Sample rate of audio in Hz |

## Output

Outputs transcribed text to stdout.

## Requirements

- Google Cloud account with Speech-to-Text API enabled
- `gcloud` CLI authenticated: `gcloud auth application-default login`
- Audio file: WAV format, mono, 16-bit

## Examples

```bash
# Transcribe a file
echo "meeting.wav" | transcribe

# Different language
echo "spanish.wav" | transcribe --lang es-ES

# Chain with mic-record
mic-record --duration 10 | transcribe

# Transcribe and translate
mic-record --duration 10 | transcribe | translate --lang French

# Transcribe and save
echo "lecture.wav" | transcribe > notes.txt
```

## Supported Languages

Common codes:
- `en-US` - English (US)
- `en-GB` - English (UK)
- `es-ES` - Spanish (Spain)
- `fr-FR` - French
- `de-DE` - German
- `ja-JP` - Japanese
- `zh-CN` - Chinese (Simplified)

Full list: https://cloud.google.com/speech-to-text/docs/languages

## Cost

Free tier: 60 minutes/month

## Design

This tool only handles transcription. Recording is handled by separate tools:
- `mic-record` - Fixed duration recording
- Future: `mic-record-ptt`, `mic-record-wake`, etc.

This separation allows transcribing audio from any source (microphone, file, network).