Skip to content

audio_recording.md

Latest commit

 

History

History
126 lines (87 loc) · 3.78 KB

audio_recording.md

File metadata and controls

126 lines (87 loc) · 3.78 KB
Raw

Audio Recording Utilities

← Previous: Voice Control System | Next: ODAS Audio Processing System →

← Back to Documentation

This document describes the audio recording utilities available in the scripts/audio/record directory. These utilities are designed to work with the ReSpeaker 6-Mic Array and provide different ways to record audio data.

Prerequisites

Before using these utilities, ensure you have the required dependencies installed:

# System dependencies (macOS)
brew install portaudio

# Python dependencies
pip install pyaudio numpy

Available Scripts

1. Basic Recording (record.py)

The simplest recording utility that captures all 8 channels into a single WAV file.

Features:

  • Records all 8 channels simultaneously
  • Saves raw audio data without processing
  • Output: Single 8-channel WAV file (output.wav)

Usage:

from scripts.audio.record.record import record_audio

# Record 5 seconds of audio
record_audio(seconds=5, output_file="output.wav")

2. Multi-Channel Recording (record_channels.py)

Records from all 8 channels and saves each channel to a separate WAV file.

Features:

  • Records all 8 channels simultaneously
  • Saves each channel to a separate file
  • Uses numpy for efficient data processing
  • Output: Eight 1-channel WAV files (output_channel_0.wav through output_channel_7.wav)

Usage:

from scripts.audio.record.record_channels import record_channels

# Record 5 seconds of audio from all channels
record_channels(seconds=5, output_template="output_channel_{}.wav")

3. Single Channel Recording (record_single_channel.py)

Records from all 8 channels but extracts and saves only one selected channel.

Features:

  • Records from all 8 channels
  • Extracts and saves only one channel (default: channel 0)
  • Uses numpy for efficient data processing
  • Output: Single 1-channel WAV file (output_one_channel.wav)

Usage:

from scripts.audio.record.record_single_channel import record_single_channel

# Record 5 seconds of audio from channel 0
record_single_channel(seconds=5, channel=0, output_file="output_one_channel.wav")

Common Parameters

All recording utilities share these common parameters:

  • RESPEAKER_RATE: 16000 Hz (sample rate)
  • RESPEAKER_CHANNELS: 8 (number of channels)
  • RESPEAKER_WIDTH: 2 (16-bit audio)
  • CHUNK: 1024 (samples per buffer)
  • RECORD_SECONDS: 5 (default recording duration)

Device Selection

To identify the correct audio device index for your ReSpeaker:

from scripts.audio.record.identify_audio_devices import list_audio_devices

# List all available audio devices
list_audio_devices()

Use Cases

  1. Basic Recording (record.py)

    • When you need raw multi-channel audio
    • For simple recording without channel separation
    • When working with audio processing tools that expect multi-channel input

  2. Multi-Channel Recording (record_channels.py)

    • When you need to analyze each microphone separately
    • For channel-specific processing or analysis
    • When working with single-channel audio tools

  3. Single Channel Recording (record_single_channel.py)

    • When you only need one microphone's data
    • For focused audio analysis on a specific channel
    • When working with single-channel audio processing

Notes

  • All recordings are saved in WAV format
  • Sample rate is fixed at 16kHz
  • Audio is recorded in 16-bit format
  • The ReSpeaker 6-Mic Array must be properly connected and configured
  • Make sure to select the correct device index for your setup

← Previous: Voice Control System | Next: ODAS Audio Processing System →

← Back to Documentation