← back to home
technical work

plugins & tools

Software I have designed and built — from VST plugins and standalone apps to open-source teaching tools and research datasets. Built at the intersection of signal processing, machine learning and sound design.

plugins & apps // built_with_juce/

Audio plugins and standalone applications built in JUCE and C++, developed for research and commercial use.

C++ · JUCE · DSP
HandySFX — Voice Character VST Plugin
A real-time voice character plugin with six DSP presets — Radio, Monster, Chipmunk, Cave, Megaphone and Robot. Each preset has its own signal chain and parameter controls, with a live oscilloscope visualiser (like Karen from SpongeBob). Built from scratch in C++/JUCE as a portfolio piece exploring DSP fundamentals.
VST Plugin C++/JUCE DSP Real-time Audio Ring Modulation Bitcrusher
→ View on GitHub
PYTHON · C++ · MACHINE LEARNING
QuAP — Quality Audio Prototyping
AI-powered tool for searching and generating sound effects. Combines ML-based classification with C++ signal processing to help sound designers find and create sounds faster... without losing creative agency or serendipity. Validated with 16 professional audio practitioners.
Audio Search ML Classification DSP Human-in-the-loop
→ Demo → DAFx 2026 Paper → GitHub
JUCE · C++ · DSP · COMMERCIAL
Nemisindo VST Plugins
Developed 6–8 procedural audio VST plugins at Nemisindo, implementing synthesis models for action-packed sound effects. Applied machine learning methods to optimise perceptual quality parameters across all models.
🚀 Rocket ✈️ Jet 🔥 Fire 💥 Explosion 🚁 Helicopter 🔫 Gun
↗ nemisindo.com
teaching tools // open_source/
JUCE · C++ · OPEN SOURCE · QMUL
JuceSteps — Audio Programming Labs
7 hands-on labs for learning audio programming in JUCE/C++. Each lab includes a starter template and a full solution. Originally developed for the Music and Audio Programming module at QMUL.
01 — Introduction 02 — Synth ADR 03 — Filters 04 — Delay 05 — Pan Laws 06 — FM & Compression 07 — Reverb & Distortion
↗ view on GitHub
PYTHON · LIBROSA · MATPLOTLIB
Audio Analysis Toolbox
A collection of Python scripts developed during my PhD for spectral analysis and visualisation of audio signals. Includes spectrogram generation, windowing function comparisons (Hann, Blackman-Harris), loudness calibration protocols, and perceptual evaluation pipelines used in listening test sessions with professional practitioners.
Python Librosa Spectral Analysis Visualisation Perceptual Evaluation
🎓 PhD Research Tool
datasets // research_output/

Publicly available datasets produced as part of my PhD research at the Centre for Digital Music, QMUL.

ZENODO · ARXIV:2501.17198 · 2025
6KSFX — Synthetic Sound Effects Dataset
6,000 synthetic audio samples across 30 sound categories, designed to advance research in procedural audio synthesis and evaluation. Includes 50 noise-augmented samples for ML applications.
6,000
audio samples
30
categories
2.56GB
total size
↗ download on Zenodo ↗ read the paper
ZENODO · DOI:10.5281/zenodo.17453307 · 2025
MR20VI — QVIM Challenge Dataset
Evaluation and development dataset for the Querying by Vocal Imitation Challenge (QVIM), hosted at AES AIMLA 2025. Contains reference sound effects paired with up to three vocal imitation recordings per sample, across multiple sound categories.
450MB
total size
2
splits (DEV + EVAL)
10
teams submitted
↗ download on Zenodo ↗ GitHub