Splice: Similar Sounds
New music discovery tool for producers, singer-songwriters and DJs.
From tens of thousands of hobbyists, to the elite producers behind Grammy-winning singles from Billie Eilish, Lil Nas X and Justin Bieber, Splice has become the go-to marketplace for music samples, and the software used to organize and work with music files for production.
But “searching” for music samples is not the same as searching Google. Keywords are poor queries for sound. So how do we take the millions of music samples on the Splice platform, and create a search experience that is intuitive for musicians who “search” by ear?
As Design Director (contract) on the Search & Marketplace team, that’s the problem I set out to solve, alongside a team of music industry-savvy designers, machine learning engineers, and product managers.
Looking at search less as a tool to pinpoint something specific (in this case), and more as a gateway to discovery, Similar Sounds mirrors the behavior one might have on Google — but with sound.
Let’s say you search “2020 Election” on Google. Sandwiched between articles about the presidential election, one about the Senate pops out and grabs your interest. So you type “2020 Election Senate” into the search bar, to narrow your search to similar articles.
That’s really hard to do with sound… if you use words. But what if machine learning allowed you to take a music sample you like, and provide an easy way within the UI to query similar sounding samples? Some that are a super close match, and others that are related yet a bit different, injecting new, relevant inspiration into your music production process?
That’s Similar Sounds. You can learn more about it, and it’s impact, via TechCrunch.