Audiergon

A Python Digital Signal Processing (DSP) processor powered by a custom Cooley-Tukey Fast Fourier Transform implementation.

💻 GitHub Repo | 🐍 PyPI Page | 📝 Docs | ☁️ Audiergon Cloud

Key Features

• Cooley-Tukey FFT
  • Pure Python iterative implementations of 1D Forward and Inverse Fast Fourier Transforms.
• Optimised Bit Reversal Permutations
  • Fast bit-reversal indexing using native compiled C-speed NumPy indexing.
• Overlap-Add Audio EQ Pipeline
  • Complete pipeline for applying Hann windowing, processing frames through an equaliser, and rebuilding raw 16-bit PCM .wav streams without clicking or imaginary artifacts.

Installation

Install the library using pip:

pip install audiergon

Dependencies

• numpy

Quick Start & API Examples

1 - Basic FFT & Frequency Analysis

Compute the Fourier Transform of a simple signal array and generate its matching frequency bins.

from audiergon import iterative_fft, iterative_fftfreq

# Input array must have a length that is a power of 2
signal = [1.0, 0.5, -0.2, 0.1, 0.8, -0.4, 0.3, -0.1]
sampling_interval = 1.0 / 44100  # 44.1 kHz Sample Rate

# 1. Perform Forward FFT
frequency_coefficients = iterative_fft(signal)

# 2. Compute matching frequencies for the spectrum bins
frequencies = iterative_fftfreq(len(signal), d=sampling_interval)

print("Frequency spectrum coefficients:", frequency_coefficients)
print("Calculated bin frequencies (Hz):", frequencies)

2 - Equalising an Audio File

Process an audio track using the built-in Overlap-Add (OLA) processing pipeline.

from audiergon import process_audio

# Modify the frequency bands using multipliers
output_path = process_audio(
    audio_filepath="in.wav",
    bass_gain=1.5,
    low_mid_gain=1.2,
    mid_gain=1.0,
    high_mid_gain=0.8,
    treble_gain=0.5,
    output="out.wav"
)

The process_audio pipeline requires 16-bit Mono PCM WAV files. You can convert standard formats using ffmpeg:

ffmpeg -i in.mp3 -acodec pcm_s16le -ac 1 -ar 44100 out.wav

API Overview

audiergon.fast_fourier_transform

• iterative_fft(arr)

  • Computes the 1D Discrete Fourier Transform (DFT).
  • Input: list or numpy.ndarray (length must be a power of 2).
  • Returns: list of complex numbers tracking amplitude and phase.

• iterative_ifft(arr)

  • Computes the Inverse DFT converting frequencies back to the time domain.
  • Returns: numpy.ndarray scaled by the sequence length.

• iterative_fftfreq(l, d=1.0)

  • Generates frequency values for each bin location.

audiergon.process

• process_audio(audio_filepath, bass_gain, low_mid_gain, mid_gain, high_mid_gain, treble_gain, output=None)
  • Applies windowed equalisation filters over an absolute file path.
• generate_hann_window(frame_size)
  • Generates a list containing real Hann coefficients to smooth frame boundary edges.
• apply_equaliser(transformed, frame_size, framerate, ...)
  • Directly apply scale gains on an existing complex frequency domain segment.

audiergon.bit_reverse

• bit_reverse(arr, convert_to_complex=False)
  • Applies bit-reversal array permutations optimized with high-performance NumPy vector operations.

License

Audiergon is open-source software licensed under the MIT License.