top of page

Sound, Smell, Light


Most scents are invisible, and different people interpret scents differently. For example, some people are addicted to the smell of durian, while others are disgusted with it. In short, there is no uniform measure of smell. The goal of this project is to standardize and visualize smells. So I'm going to design a set of algorithms that convert smells into sound and light.

Molecular Vibration Theory

As early as the 1920s and 1930s, Malcolm Dyson made a bold conjecture that smell could distinguish different chemicals by detecting the vibrations of smell molecules, and later the work of Robert Wright et al. further promoted the development of the theory of smell vibration. (The Science of Scent, Luca Turin, TED 2005)

Design Concept

I built a set of algorithms that could transform olfaction into hearing and vision, which amplified the molecular vibration frequency to a certain number of times and mapped it to piano keys and visible light frequency. For a detailed introduction, please see my paper Design of olfactory information visualization - Olfactory AR Watch ‘oWatch’

Example of Algorithm Application

I chose three scents -- jasmine, lavender and sandalwood. The algorithm was used to calculate the sound and color corresponding to their main components. So in the video, each smell released by the atomizer is accompanied by a sound and color release that represents it. (I made it with Maxmsp and Arduino)

Functional Test Video

Final prototype


Devoted to Innovation Design

bottom of page