Sound Intelligence Platform

Sound shapes trust.
Now you can build with it.

DoReMiFo maps acoustic properties to measured emotional response — so audio decisions have a reference point instead of an opinion. Built on a controlled study of 90 sounds rated by 490 listeners across 37 countries.

Academic originHAMU Praha · SGS Grant
Emotional modelRussell Circumplex (1980)
Acoustic frameworkEssentia descriptors · active-segment contract
01 — The problem

Audio branding runs on gut feeling.

Sound is the most emotionally direct medium a brand has — and the least rigorously managed. That gap is expensive.

🎯

Briefs that don't land

"Make it feel premium." "More trustworthy." Without a shared reference, every audio brief is a guessing game — and revisions are expensive.

📉

Sound that contradicts your brand

A single notification tone can undermine months of visual brand work. Most teams don't find out until users complain — or churn.

🔄

No way to validate

You can A/B test a headline. You can user-test a colour. Sound? You ship it and hope. There has been no tool for evidence-based audio decisions — until now.

02 — What you get

A common language for sound.

DoReMiFo translates acoustic properties into measured perceptual responses — so designers and clients can describe the same sound in the same terms.

Know what a sound actually communicates

Before you ship. Before the client hears it. DoReMiFo maps acoustic properties to the perceptual responses measured from listeners — so you can design with intent, not hope.

Brief smarter, revise less

Replace vague adjectives with measurable targets. A brief a sound designer can act on — and a client can approve without endless rounds.

Validate before you commit

Compare candidates against a perceptual reference library. Know whether your shortlisted sounds land where you intend — before they go into production or on air.

Evidence your clients can present upward

Audio decisions are hard to defend in a boardroom. DoReMiFo gives you data — perceptual scores, emotional positioning, benchmarks — that make the case without requiring the room to trust your ear.

03 — Products

Four ways in.

Whether you're a researcher, a sound designer, or just curious about how you hear — there's a DoReMiFo product for you.

04 — Research

Built on evidence, not opinion.

DoReMiFo starts as doctoral research at HAMU Prague. The first study is complete and pre-registered; what follows reports it as measured, including the parts that did not work.

Institution

HAMU Praha

Funding

SGS Academic Grant

Emotional model

Russell Circumplex (1980)

Feature extraction

Essentia · active-segment v2

Study participants

490 across 37 countries

Stimulus library

90 controlled 2.5-second sounds

Wave 1 result, stated plainly: perceived texture and activation are substantially predictable from acoustic descriptors; pleasantness is not. One perceptual dimension — fullness — depends on temporal structure that static descriptors do not capture, and is reported as outside the current model rather than modelled anyway. Two methodological failures are documented, one caused by a feature-extraction artifact that had produced apparent predictive signal until it was found. Predictions for the second study were registered before any data was collected.

Pre-registration and methods available on request — for researchers, the dataset is available under an academic-use agreement.

Early Access

Be first when the tools launch.

The DoReMiFo plugin suite and Sound Intelligence App are in development. Leave your email — we'll reach out when access opens.

No spam. Used only to notify you about DoReMiFo product launches.