audiergon 0.1.0

A custom educational DSP suite featuring a from-scratch iterative FFT implementation - built by Hamd Waseem

Audiergon Local

A suite of audio-related tools built with Python and AWS to demonstrate the use of the Fourier Transform!

Features

Cooley-Tukey Fast Fourier Transform

An implementation of the Cooley-Tukey FFT in Python, in fast_fourier_transform.py. It contains the following three methods:

• iterative_fft() performs a forward FFT
• iterative_ifft() performs a backward FFT
• iterative_fftfreq() calculates the frequencies outputted by an FFT

Local EQ Filter

A local Gradio client that uses the FFT to modify frequency bins in inputted audio files to analyse them.

• Requires the gradio library
• Uses process.py (Hann Window and Overlap-Add) for processing and local_gradio.py for the UI

Fourier Analysis

A local analysis tool to visually graph the Frequency Domain and Time Domain of a sound file.

• Requires inputted audio files to be formatted as .wav files using a Mono PCM codec.
  • Use ffmpeg -i in.mp3 -acodec pcm_s16le -ac 1 -ar 44100 out.wav in a command line to convert
• Uses the fourier_analysis.py file

Live Fourier Analysis

A local analysis tool that uses your microphone to graphically show the Frequency Domain of the ambient sound around you!

• Requires the sounddevice library
• Uses the live_fourier_analysis.py file

Upcoming Features

Audio Compression Tool

Utilises the concept that high frequency sound gets masked when lower frequencies are louder.

AWS Cloud Implementation and UI

A streamlined version of the FFT designed to process audio as fast as possible entirely within the cloud.

• Planning to use S3 Event Triggers combined with Lambda and a simple Vercel/Next.js frontend.

Devlog

Check out Audiergon Devlog Part 1 on YouTube for a detailed run-through of the theory behind the Fourier Transform!

License

Audiergon is licensed under the MIT License