Visual Microphone

Have you ever wondered if sound can be seen? Well, during my final year at Middle East Technical University, I had the chance to dive into a project that explored just that-capturing sound with light! Intrigued? Let me tell you more.

In a nutshell, this project is called the Visual Microphone, where we investigated how sound interacts with light, allowing us to reconstruct sound without using a typical microphone. Instead, we relied on an optical system to detect tiny vibrations caused by sound waves. Cool, right?

So, how does it work?

Picture this: We set up a laser beam aimed at a surface, and as sound plays near it (think of a speaker blasting a tune), the surface starts to vibrate ever so slightly. These vibrations affect how the light bounces off, and using a high-speed infrared camera, we capture these tiny changes in the light pattern. From this, we can actually reconstruct the sound being played. Think of it as turning light into sound!

Our setup was quite simple but super effective. We used a multi-mode optical fiber and a 1550 nm laser (the same kind used in telecom systems), with a camera that took snapshots of these vibrations. By analyzing the light patterns-or speckle patterns-we could reverse-engineer the sound.

The fun part:

During our experiments, we played some tunes (even Inspector Gadget’s Theme!), and by using a combination of filters and smart signal processing, we managed to clean up the noisy parts and extract the pure sound. After processing the data, we ended up with a sound reconstruction that was almost as good as the original. Who would have thought light could “hear”?

What can this be used for?

Not only is this a cool experiment, but it could also have some real-world applications! For instance, monitoring vibrations in buildings for safety, checking the structural integrity of bridges, or even eavesdropping on sounds at a distance (we promise, only for science!).

A Peek into the Setup:

Laser Beam: To light up the vibrations.

Multi-mode Fiber: To guide the laser and capture the tiny changes in light caused by sound.

High-Speed Camera: Our “ears” for this project, capturing thousands of frames to piece together the sound.

This project gave me the chance to explore the intersection of light, sound, and signal processing. It’s amazing how much information we can gather from the most unexpected places.

For more details, you can find this preprint at Arxiv.

A big thanks to Berk N. Gün, a PhD student who contributed to this project, and Prof. Emre Yüce, who guided us throughout this incredible experience.