What Is Sonic Similarity in Music?

Sonic similarity is a form of music search that finds connections between tracks using characteristics within their sound: rhythm, harmony, timbre, energy and structure, all of which define a track’s sonic identity and how it relates to others.

Tracks from different artists, genres and eras might share a similar tonal character or energy. To a traditional search engine, they might have very little in common, but when you actually listen to them, the connection could be obvious.

Sonic similarity is not just about finding two songs that sound alike; it’s about discovering a deeper and more subtle connection between recordings.

Until not too long ago, these connections were hard to put into words. The closest we ever got to it was through metadata, a critical component when it comes to music discovery. But as music catalogues continue to grow, metadata alone can’t always capture the sonic relationships between songs.

What the music industry needs are tools that can help explore catalogues in new ways, complementing trends and tags by focusing on sound characteristics. Today, technology can analyse some of the sonic characteristics behind those connections, highlighting threads between songs that, on the surface, seem very far apart.

For DJs and curators, this could translate into new music and unexpected connections to explore; for labels, it offers another way to navigate increasingly large catalogues and resurface recordings that might otherwise be missed.

Welcome to sonic similarity: a different way to discover music.

What Makes Two Tracks Sound Similar?

Tracks can sound alike for a variety of reasons, most of which can’t be described by tempo or energy levels alone, including:

• Rhythm and groove
• Harmony and tonal relationships
• Energy
• Timbre and texture
• Spectral characteristics
• Structure

If you want to find out more about how two tracks can connect, no matter how far their genres are, check out our recent article on What Makes Two Electronic Tracks Sound Similar.

Sonic Similarity vs Metadata

Genre, artist, release date, BPM, mood and other information can help us go through even the largest music catalogues. But metadata doesn’t always tell us how a track actually sounds. Instead, it describes and categorises a recording using information attached to it.

Imagine you have a reference track in mind and want to find something similar. You could search by genre, BPM and mood, and that would certainly narrow things down. But what if there was a strong match that belonged to another genre altogether? This is where sonic similarity offers a different path.

Metadata tells us about the music, but sonic similarity offers another way to explore what’s happening within the sound itself.

How Does Sonic Similarity Work?

The process starts with a reference track. The audio is analysed across different characteristics, which can then be compared with other analysed recordings to identify sonic relationships.

This makes music similarity search possible: instead of starting with keywords or categories, the recording itself becomes the starting point for the search, which means you can start with one song you already know and explore other recordings based on how closely their sonic characteristics relate to it.

In simple terms, you no longer have to translate what you’re hearing into genres, moods or keywords before you start looking for a match. The only thing you need is a reference and the knowledge that you want something sonically related to it.

Why Sonic Similarity Matters for Music Discovery

If a relationship exists between two songs, genre doesn’t necessarily have to stand in the way. A techno track might share certain sonic characteristics with experimental electronic music, ambient, or something from a completely different era.

By looking beyond metadata alone, sonic similarity can help you discover hidden connections within music, across styles and decades. This can help you explore music you might never have thought to search for in the first place.

For DJs: Start with a Track That Works

For DJs, finding tracks that mix well together is critical, and with sonic similarity, you can start with a song you already know and explore tracks with related sonic characteristics.

Some might be obvious choices, but others could come from artists, genres or eras you wouldn’t normally explore. The final decision remains yours: sonic similarity simply gives you another way to dig deeper.

BPM and key are important, but they offer a limited overview of a song. Groove, energy, texture and structure can make two tracks with identical BPMs behave very differently on the dancefloor.

How Resomix Finds Sonic Similarities

Resomix’s Sonic Discovery is a music similarity search tool designed to find sonically related tracks using a reference recording.

PYXIS-1 analyses 70+ measurable audio characteristics across areas including rhythm, harmony, energy, timbre, spectral content and structure.

Starting from a reference track, Resomix’s proprietary engine can compare its sonic characteristics with other analysed recordings, returning ranked sonic matches.

This means tracks don’t necessarily need to share a genre, era or descriptive tags to be connected; if their sonic characteristics are closely related, that relationship can bring them into the results.

A Different Way to Discover Music

In an era when genres are as fluid as they’ve ever been, genre tags can’t always capture everything we hear. Sometimes the most interesting connection is between two tracks that, on paper, have almost nothing in common.

Sonic similarity gives us a way to explore those relationships, starting with the sound itself rather than only the words we use to describe it.

Want to find similar tracks through sound? Try Sonic Discovery with a reference track.