Map the ecosystem
We study existing tools, papers, and workflows, compare approaches, and identify gaps. Technical reviews and plugin benchmarking help teams understand their options and choose where to invest development effort.
Research, engineering, and musical practice.
Our team brings together audio engineering, software development, interface design, and music research. We connect the study of a problem with the tools to explore it, test it, and put the results to use.
Build tools for developers
We build SDKs, reusable audio components, interfaces, and development workflows. These foundations help teams connect sound, controls, and delivery, alongside analysis tools for researchers working with music data.
Develop instruments and interfaces
We turn musical ideas and emerging technology into instruments, effects, and interactive software. Prototypes let musicians explore a method and help teams refine the sound and interaction together.
Investigate with partners
We contribute research and engineering to a shared problem, from model integration to datasets and benchmarks. The work can support a product decision, a reusable tool, or a joint publication.
Who we work with
Bring your domain expertise, a technical challenge, or a creative direction. We can work alongside your team or develop a specific part of the project.
Start a conversationInstrument and plugin developers
Explore a new instrument or effect, integrate a model, improve an interface, or modernize an authoring workflow. We help with DSP, development tools, host integration, and testing in the environments your product needs.
Get help with a product integrationAcademic labs and researchers
Work together on a specific question in music representation, generation, synthesis, or evaluation. We can build research software, interactive demos, datasets, and benchmarks, and help students make their work easier to run and extend.
Turn a research method into a usable toolCompanies working with sound
You may be building a game, a media platform, an educational tool, or an interactive device. We can help explore how audio fits your product, prototype an experience, and integrate music or sound technology with your existing software.
Artists and creative teams
Develop a custom instrument, sampler, or performance interface around your practice. Bring a playing technique, a sound collection, or an interaction you want to explore; we can build and refine it with you.
A collaboration can produce a working prototype, a benchmark, or a documented integration, with measurements and limitations your team can use to decide what comes next. We agree together on what stays private and what can be shared.
Some of our current projects
Text-to-Drumkit
An interactive research workbench for turning a text palette into a drum kit, inspecting individual sounds, and comparing saved benchmark evidence.
NewsAccepted at ISMIR 2026: see you in NovemberMaru Engine
Reusable audio building blocks and the tools to connect them into instruments, effects, and musical software.
NewsMeet us at ADC 2026 in BristolAgent Studio
A workspace in development for building musical tools with a coding agent.
Symbolic & Code Sequencer
A library and interactive visualizations for mapping musical relationships. Code Sequencer is our early experiment in expressing song structure as editable programs, using relative musical roles such as scale degrees and chord relationships.
Research interests and experiments
- More control over MIDI generation: shaping harmony, rhythm, and arrangement while keeping the result editable. We’re interested in small local models and the trade-offs between musical control, response time, and memory use.
- Musical agents that listen, anticipate, and play alongside musicians as a pianist or accompanist, with benchmarks that test their timing, responsiveness, and musical choices.
- Cohesive drum-kit generation and sampler instruments: exploring how sounds work together across a kit, with evaluation of individual sounds, the collection, and its usefulness to musicians.
- Work in progress: benchmarking classifiers for very short sounds and exploring sound embeddings for sample discovery and synthesis techniques such as granular synthesis.
- Visual and spatial interfaces, including 3D exploration, that help people navigate sound collections, shape synthesis, and understand musical relationships.
- Compatible audio systems and repeatable evaluation: studying how sound, controls, and saved state behave across browsers, plugins, and devices.
Exchange ideas. Share findings.
We’re always happy to connect with other labs and research communities, read papers for feedback, and discuss early ideas. Those conversations can grow into shared studies, models, datasets, or tooling around work in music information technology.
We welcome shared studies and publications, with methods, code, benchmarks, and interactive examples that make the findings easier to examine and reuse. In November we present Text-to-Kit, a benchmark and method for generating playable drum kits from text, at ISMIR 2026, and join the Audio Developer Conference in Bristol to discuss our cross-platform engine work.
What are you working on?
Tell us what you’re exploring, where you need help, and what a useful result would look like. A paper, a prototype, or a short description is enough to start.