Audio Toolset

Python library and command line interface for processing audio signals.

audio_toolset provides an intuitive and flexible way to process audio files in Python. Each Channel object handles a single mono audio file, allowing you to apply filters, EQ, dynamic processing, gain adjustments, noise reduction, and more.

[!TIP] Stereo files can be handled by first splitting them into mono channels using split_to_mono and then recombining them with join_to_stereo. see example below

Installation

pip install audio-toolset

Examples

Quickstart

from audio_toolset.channel import Channel

track = Channel("my_audio.wav")

track.lowpass(cutoff_frequency=12000)  # Remove harsh highs
track.highpass(cutoff_frequency=80)  # Remove low-end rumble
track.normalize(target_db=-1)  # Normalize

track.write("my_audio_processed.wav")

Command line interface

All methods in Channel are exposed as subcommands in the audio-toolset CLI and can be chained in any order. this command will produce an identical result to the python script above

audio-toolset --source my_audio.wav \
  lowpass --cutoff-frequency 12000 \
  highpass --cutoff-frequency 80 \
  normalize --target-db -1 \
  write --output-path my_audio_processed.wav

Process multiple files

This is very handy when you need to apply the same processing to multiple files, for example a directory with sound effects.

import pathlib

from audio_toolset.channel import Channel

input_dir = pathlib.Path("input_audio")
output_dir = pathlib.Path("cleaned_audio")
output_dir.mkdir(exist_ok=True)

for wav_file in input_dir.glob("*.wav"):
    track = Channel(wav_file)

    track.lowpass(cutoff_frequency=12000)  # Remove harsh highs
    track.highpass(cutoff_frequency=80)  # Remove low-end rumble
    track.normalize(target_db=-1)  # Normalize

    track.write(output_dir / wav_file.name)

Working with stereo files

from audio_toolset.audio_data import AudioData, join_to_stereo
from audio_toolset.channel import Channel

# Load stereo audio file
stereo_data = AudioData.read_from_file("my_audio.wav")

# Split stereo into left and right channels
left_data, right_data = stereo_data.split_to_mono()
left_channel = Channel(left_data)
right_channel = Channel(right_data)

# Process each channel
for track in [left_channel, right_channel]:
    track.lowpass(cutoff_frequency=12000)  # Remove harsh highs
    track.highpass(cutoff_frequency=80)  # Remove low-end rumble
    track.normalize(target_db=-1)  # Normalize

# Combine channels back to stereo and save
join_to_stereo(left_channel.audio_data, right_channel.audio_data).write_to_file(
    "my_audio_processed.wav"
)

Features

• High level Channel object for simple user interface
• Load audio from a file or AudioData
• Gain processing
  • gain, normalize, fade
• Filters
  • lowpass, highpass, eq_band
  • noise_reduction (spectral gating)
• Dynamics
  • compressor, limiter, soft_clipping
• Plotting tools (see examples)
  • plot_signal - generate signal time and frequency plot
  • Dynamic plots - compressor, limiter can generate attenuation plots
  • Bode plots - lowpass, highpass, eq_band can generate filter bode plots

Plotting

audio_toolset provides multiple plotting tools to visualize signals, filters, and dynamics. All plots are generated interactively using Plotly and can be saved as images from the interactive viewer.

Signal Plots

Use plot_signal() to visualize the time-domain and frequency-domain representation of an audio channel.

2. Filter (Bode) Plots

Filter plots show the frequency response (Bode plot) of filters applied using eq_band(plot_filter_bode=True), highpass(plot_filter_bode=True), or lowpass(plot_filter_bode=True).

3. Dynamics Plots

Dynamics plots visualize the response of compressors and limiters applied to a channel. Generated by compressor(plot_compressor_response=True) or limiter(plot_limiter_response=True)