Dockyard Acoustics Sound Technology

🌆 Here is something Michael Kozák and I are cooking up as another addition to our growing portfolio. 🛁It all started with measuring the acoustics of a shower room, and slowly grew into a collection of softwares that were suboptimal. 🛰️This one, however is more of a tried and tested concept that we will be sharing with musicians, not acousticians. The first of what we hope to be many GPU accelerated plugins. 🌌More on this and the others soon ... Audio:- TRINIX - Quedate Luna (feat. Natalia Doco) https://lnkd.in/eZCTdkzj #audiosoftware #vstplugin #spatialaudio #musicproduction #roomacoustics #juce

💽 Music to Visuals 🌈 🐈⬛ For all the summer parties where people may have a spare projector ready: AURA, an audioreactive desktop visualiser for Windows. 🎶 ⬇️ Download: https://lnkd.in/dFpiEFBG 🔗 Source: https://lnkd.in/dE5GtdJ4 It listens to whatever's playing on your system audio, Spotify, YouTube, a DAW session, doesn't matter, through WASAPI loopback and renders it through 9,795 Milkdrop presets organised into a category-sidebar browser. ⚙️ Built on top of the excellent projectM engine + frontend-sdl2 (LGPL). My contribution is a custom overlay: FFT spectrogram pipeline, lock-free SPSC PCM ring buffer, ImGui floating-chip UI, category browser with global search. #AudioVisualisation #DSP #FFT #SignalProcessing #ProjectM #Milkdrop #CPlusPlus #ImGui #OpenGL #WindowsDev #OpenSource #IndieDev #Acoustics

Wayverb – Post 2 of 2 Room Acoustics software. Ported to Windows. It's Live. 🟣🔧 📥 Download (Windows, 16 MB): https://lnkd.in/ePnvTega 📂 Source code: https://lnkd.in/ePnvTega The original macOS-only C++ codebase and rebuilt it on Windows 11. It compiles. It runs. It simulates. 🟣 8 patches to the engine. A new results window. A full dark-theme GUI. Pre-built binary release: download, extract, double-click. No MSYS2. No terminal. No build steps. What you get: → Load any OBJ room geometry → Hybrid simulation: FDTD waveguide + ray tracing + image sources → Full results: impulse response, spectrogram, energy decay, frequency response → Acoustic metrics: RT60, EDT, C80, C50, D50 → Auralization: convolve your own audio with the computed IR → FDTD and ray animation replay What it took: → Fixed OpenCL kernel bugs for NVIDIA GPUs → Fixed a silent numerical crash in the stochastic raytracer → Fixed shader and OpenGL issues that produced black viewports → Capped GPU memory usage for consumer cards (6 GB VRAM) → Built the entire analysis and playback layer from scratch Original project by Reuben Thomas. This port is open source. 💜✨ #Wayverb #ComputationalAcoustics #RoomAcoustics #AudioEngineering #OpenSource #SpatialAudio #GPUComputing #Binaural #EngineeringInnovation #Windows #FDTD #RayTracing

🚗💨 Wind noise is one of the dominant comfort challenges in electric vehicles. This prototype visualizes near-wall flow unsteadiness around a vehicle using a GPU-accelerated solver which highlights regions of elevated aerodynamic activity often correlated with broadband noise generation (qualitative proxy). The long-term vision: rapid aeroacoustic diagnostics during early-stage vehicle design, before physical prototyping. #Aeroacoustics #EV #AutomotiveTech #CFD #DeepTech #Simulation #Startup #EngineeringInnovation #DockyardAcoustics

Wayverb – Post 1 of 2 Whatever Happened to Wayverb? 🟣🎛️ If you’ve worked in computational acoustics or physically based audio simulation, you may remember Wayverb, built by Reuben Thomas. For a short but important period, it was one of the most ambitious open efforts in hybrid room acoustics simulation. Then it quietly faded. Wayverb was not another reverb plugin. It aimed to compute acoustics rather than design them. By combining ray tracing and image source modeling with wave-based techniques, it attempted to simulate sound propagation from first principles. The architecture was hybrid, balancing physical realism with computational practicality. Its objective was straightforward but bold: generate room impulse responses directly from 3D geometry, informed by material absorption and diffusion, and render them spatially using HRTFs. Instead of tuning parameters until something sounded convincing, compute the acoustic field and let physics define the result. The engine was built in C++ with OpenCL acceleration. It ingested geometry, ran hybrid solvers, and produced impulse responses for convolution rendering. Low frequencies were approximated with more computationally intensive methods, while higher frequencies were treated geometrically. It was research-grade by design. And that exposed real limitations. The hybrid crossover between wave and geometric domains was difficult to make perfectly consistent. Diffraction modeling was limited compared to full-wave solvers. Low-frequency accuracy came at high computational cost. Large scenes scaled poorly. OpenCL, while portable in theory, often meant driver instability and hardware inconsistencies in practice. Beyond solver physics, workflow was a barrier. There was no seamless DAW integration, no real-time interaction loop, and no UX layer built for creators. It required technical fluency and patience. Wayverb did not fail because the idea was wrong. It struggled because sustaining a GPU-heavy scientific codebase is demanding. Toolchains evolved. Dependencies drifted. Hardware targets shifted. Meanwhile, the industry leaned toward convolution libraries, ML-driven reverbs, and real-time game engine acoustics. Physically rigorous simulation became niche. But the landscape has changed. Modern GPUs are dramatically stronger. Real-time ray tracing is mainstream. Compute APIs are more mature. The hardware constraints that once limited physically based acoustics are no longer what they were. So the real question is this: Was the limitation the concept or the timing? 🟣 Perhaps geometry-driven acoustics deserves a second generation: open, modernized, GPU-native, and built for today’s workflows. This is not nostalgia. It is a reconstruction. I am resurrecting it for everyone to access. Wait for Wayverb – Post 2 of 2. 🟣✨ #Wayverb #ComputationalAcoustics #RoomAcoustics #AudioEngineering #OpenSource #SpatialAudio #GPUComputing #Binaural #EngineeringInnovation