Nordsound
Precision Engineering from the Northern Frontier.
Based in Hokkaido, Japan, Nordsound is a specialized software engineering studio. We operate at the unique convergence of Creative Technology (Sound/Visuals) and High-Performance Infrastructure (Search/Backend).
Heavy Data. Light Code.
The Digital North
In a digital landscape cluttered with inefficiency, we seek the signal. Our "Nordic" philosophy drives us to build software that is robust, minimalist, and transparent.
Whether architecting a search engine for millions of records or a real-time visual system for a global event, we apply the same rigorous standard: absolute stability and speed. We believe that true creativity requires a backbone of flawless technology.
Core Competencies
A rare blend of "Soft" creative coding and "Hard" backend engineering.
High-Scale Backend & Search
We build the invisible infrastructure that powers modern platforms.
- - Search Engine Architecture (Elasticsearch, vector databases)
- - High-Traffic Infrastructure (AWS/GCP auto-scaling)
- - Massive Data Handling (millisecond-response optimization)
Video Processing & Media
We treat media as data, building complex pipelines that process, analyze, and generate content.
- - Real-Time Processing (server-side rendering, transcoding)
- - Low-latency video distribution pipelines
- - Computer Vision (image recognition and analysis)
Sound Programming
Audio is physics, and in the digital realm, it is mathematics.
- - DSP (Digital Signal Processing)
- - Sonic input to system output bridging
- - Audio-visual synchronization systems
Products
Tools we build and share with the community.
DCTL Workbench
DCTL development environment with color pipeline for VS Code
A comprehensive development environment for DaVinci Color Transform Language (DCTL). Write, test, and preview DCTL shaders with real-time EXR rendering — without leaving your editor.
- - Syntax highlighting, IntelliSense & 40+ snippets
- - Real-time EXR preview with GPU acceleration
- - Full ACES 2.0 color pipeline support
- - One-click copy to DaVinci Resolve
AGTL
Audio GPU Transform Language — write once, compile to any GPU
A C-like language for GPU compute shaders, specialized for audio DSP processing. Write portable compute kernels once and compile to WGSL (WebGPU), MSL (Metal), HLSL (DirectX 12), SPIR-V (Vulkan), and GLSL (OpenGL) — powered by a TypeScript frontend and a Rust/Naga WASM backend.
- - C-like syntax with GPU-specific primitives (__KERNEL__, __SHARED__, sync())
- - 5 backend targets: WGSL, MSL, HLSL, SPIR-V, GLSL
- - Two-tier compilation: TypeScript parser + Rust/Naga codegen via WASM
- - Built for audio DSP: FFT, MFCC, spectral analysis kernels included
au-renderer
Offline Audio Unit renderer for Node.js
Load any AU synth plugin on macOS, set parameters, send MIDI, and render audio — all programmatically, no DAW required. A native C++/Objective-C++ addon built on Apple AudioToolbox, designed for headless offline rendering at scale.
- - Discover and load any AU synth plugin by name
- - Full parameter control with get/set and batch rendering
- - Sample-accurate MIDI scheduling with mono and stereo output
- - State persistence via ClassInfo (save/restore full plugin state)
Engineering-First Creativity
Our Approach
Most creative studios struggle with scale; most system integrators lack artistic sensibility. Nordsound bridges this gap.
We are the technical partner for clients who need "Creative Tech" that actually works in production—stable, scalable, and secure.
From R&D in Hokkaido to global deployment, we deliver code that stands the test of reality.