A

audio processing

The Delianiser is a historically constrained audio-systems design project exploring the upper limits of what could have been built in 1964 using a network-oriented signal/control architecture (rather than a track-oriented workflow), informed by the experimental lineage of Daphne Oram and Delia Derbyshire. It also forward-tracks the architecture into modern technology to test whether a DAW derived from this pedigree can form a useful strict functional superset of the 1964 design.

Although the project is nominally audio, the 1964 Delianiser is basically a massive MIMD-style analogue computer for processing any wave system of almost any arbitrary complexity. The 2026 variant is a digital/analogue hybrid of the same.

The project has been split into modules:

This is the top level and contains the abstract views plus the top-level metadata

Delianiser-Graphs: Describes how the different parts of the project link up Delianiser-Implementation: This is where the updates as indicated by later technologies get merged in, until we reach the modern version Delianiser-Reference: Here we have a detailed reference implementation based on the knowledge and technology available in 1964 Delianiser-Relationships: How components connect up - this is lighter-weight than the graphs system Delianiser-Scans: This will eventually have links to scans that can be found online and actual scans for information that isn't otherwise available

The Delianiser is a historically constrained audio-systems design project exploring the upper limits of what could have been built in 1964 using a network-oriented signal/control architecture (rather than a track-oriented workflow), informed by the experimental lineage of Daphne Oram and Delia Derbyshire. It also forward-tracks the architecture into modern technology to test whether a DAW derived from this pedigree can form a useful strict functional superset of the 1964 design.

This module contains the 1964 reference implementation. It is possible to use this in isolation to simulate how the technology of the time would have worked. The Implementation module shows the changes relative to the baseline.

An utility software to download videos from YouTube, either in mp3 (audio only) or mp4 (video+audio).

Source repository for the DAFx25 paper "Real-Time Virtual Analog Modelling of Diode-Based VCAs" by Coriander V. Pines

Exploring Discrete Fourier Transforms in C++23, see rendered output: https://dft.germs.dev/

A prototype lossless audio compressor that I made as part of my undergraduate thesis (2016-2017). This audio codec uses techniques like linear prediction of samples, inter-channel decorrelation, and Huffman coding to achieve compression.

This project aims to help with voice-training by displaying the current resonance of the voice. Additionally features of comparing and analyzing the voice are planned.

A set of command line tools for processing a collection of music or audio. Can separate music by instruments, process with sound effects and mixdown back into composition. Built with SoX, Spleeter and Python