About
I'm an independent research engineer for hardware teams, immersive media companies, and research labs. Acoustic modeling, measurement, and the software that has to run on the finished device — handled as one piece of work.
My doctorate at RWTH Aachen was in acoustic field modeling, electro-acoustics, and signal processing. As a Senior Research Engineer at Huawei I applied it to audio hardware — earbuds, headphones, wireless loudspeakers, and home entertainment systems — carrying research to working prototypes.
How I work
- Models validated against measurement, not simulation alone
- Reproducible, version-controlled pipelines, not one-off scripts
- Code and data you own, not just a PDF report
- Deliverables your engineers can run and extend
Services
Spatial Audio System Design
Spatial audio pipelines accumulate small errors — truncated order, HRTF mismatch, arrays not matched to the processing — that together degrade the experience.
End-to-end review and design of spatial audio chains — microphone and loudspeaker array algorithms (beamforming, spatial filtering, directivity control, source separation), Ambisonics encoding and decode, and binaural or loudspeaker reproduction. Higher-order Ambisonics, head-tracked and head-locked binaural, perceptual assessment. Includes HRTF simulation from 3D head geometry, spherical harmonic decomposition and interpolation, near-field HRTF modeling, and personalization workflows.
Acoustic Simulation
Simplified models miss both the diffraction and near-field effects and the electrical-mechanical coupling that together determine real transducer performance.
Full-field boundary element (BEM) simulation on your actual geometry — microphone arrays, loudspeaker baffles, headphone cups, enclosures. 3D directivity, radiation impedance, surface pressure maps, and HRTF simulation. Combined with coupled lumped-element electroacoustic models of loudspeakers and microphones — frequency response, sensitivity, and enclosure performance from electrical drive to far-field output. Once validated, the model lets you explore design variants at a fraction of the cost of building and measuring each one.
Audio Data Pipelines
Simulation, measurement, and processing tools live in separate workflows that are hard to reproduce, audit, or hand off.
Integrated Linux-native pipelines connecting CAD geometry through BEM simulation, electroacoustic modeling, and signal processing — scripted, version-controlled, and documented. Bridges digital signal processing and real-world analog measurements: impulse responses, directivity, near-field and far-field surveys — structured to close the loop between simulation and physical reality. Deliverables your engineers can run and extend. Results you understand, not just data files.
Systems & Deployment
An algorithm that works in a notebook still has to run reliably on real hardware. The gap between a working prototype and a deployable system is where audio projects stall.
From algorithm to running hardware. Linux/OS-level system design and deployment of acoustic and signal-processing software on real targets — from workstations to embedded PCs and single-board computers. Real-time audio configuration, scripting and automation. Design and build product-like prototypes, integrating hardware and software. Hardware bring-up and integration: wiring transducers, sensors, and measurement rigs to the compute that drives them — delivering the whole chain as a reproducible system you can deploy, not just a proof of concept.
Typical Engagements
Simulation Study
2–5 days
Predict directivity and response from geometry — before you commit to a build
Measurement Campaign
2–5 days
Characterize a prototype and ground your simulations in measured data
Pipeline Development
1–4 weeks
Turn ad-hoc scripts into a documented workflow your team can rerun
Design Review
1–2 days
An expert second opinion on an acoustic model or spatial audio chain before you build on it
Acoustic Prototype Development
2–12 weeks
From concept to a working, product-like prototype — hardware and software integrated, running on real targets
R&D Collaboration
Ongoing
Co-investigation with research labs and funded project consortia
Fixed-scope projects available · Remote or on-site (Europe)
Projects
CCTrack: An open-source, open-hardware MIDI headtracker you can build in an hour
Built from off-the-shelf parts, the device streams head orientation as MIDI control changes over USB; a companion browser app reads it live over Web MIDI, drives a 3D head preview, and auditions the result binaurally through a measured HRTF set.
It covers the same ground as the work above — head-tracked binaural rendering, HRTF processing, embedded firmware, and hardware bring-up — end to end, in public.
LAC 2026: Interactive Impulse Response Measurement
A talk at the Linux Audio Conference 2026, Maynooth University, on a Python tool for acoustic impulse response measurement — driven from an IPython prompt rather than a GUI, native to both JACK and PipeWire, and able to measure across several sound cards in one run. Built on Fons Adriaensen's zita-jacktools.
Slides (PDF) · Linux Audio Conference 2026 · Source on GitHub
Contact
Available for project-based engagements. If you have an acoustic or spatial audio problem and want to discuss whether I can help, get in touch.
CCTrack, above, is a public example of the work — hardware, firmware and signal processing end to end. Happy to walk through it, or any of the published research, in detail.