By design, music-licensing platforms (production music libraries & stock music marketplaces) are built to deal with hundreds of thousands of unknown tracks, thus making search the core element of the client’s experience. Navigation through such a massive amount of music should be made as easy and intuitive as possible.
Up until now, media file search queries have been performed textually, making the retrieval results highly dependent on the quality and quantity of existing metadata.
No matter how brilliant your music is, TV/film producers or app developers will never license it if they can’t find it. Whatever they are searching for, they have a very specific brief, and imagery in mind that won’t be covered by “dreamy track featuring cello, electronics and guitar”.
To be able to answer any possible query, most music-licensing platforms have developed over-complicated thesaurus and set of labels. The trouble is the only people who fully understand each of these labels — “progressive trance”, “hard trance”, “happy hardcore”, “speedcore”, “freestyle”, are committed enthusiasts of the music. This is not really helpful for the average filmmaker or videogame developer.
In a world of apps, end-users have higher expectations when it comes to search and user experience
The fear of missing out from music-licensing platforms results most of the time in fuzzy interfaces and clunky user experiences. When designing systems that are information-heavy, we might not always consider that people have their own limits.
In a world now dominated by apps, music-licensing platforms need to think of themselves as web apps and not only boutiques anymore. In the last few years, with the advent of mobile experience, user experience has become the core component when you design a service. Everything is developed with the idea of the user first. This will become more and more prominent as websites continue to compete for user’s attention.
Your app should empower your user rather than make them feel lost.
Music is worth a thousands words
Niland’s technology brings an entirely new approach for music search experiences. Using machine-learning technologies, it is getting now possible to use music to find music. Users can perform sound similarity searches and get results ranked by acoustic similarity. This is a more natural way of searching for music, especially for a subjective medium as music.
Some tracks are a mix of several genres or styles and cannot be properly labeled. And sometimes words just don’t capture what you’re looking for. Most of your users do not have a musicologist background, and are not familiar with music expert taxonomy.
But most of them have music references in mind when it comes to search for music. That “song sounds like” is a phrase we are used to hear when it comes to describe a newly released track.
Using a music similarity search engine (or call it audio search if you want) is a way to uncover hidden gems in massive content catalogs. Let’s empower your end users with tools that match the way they naturally think about music.
How do I get API access?
Getting API access is simple. Just send an email to our sales team at contact[at]niland.io or fill out our contact form found here and one of our sales reps will get back to you right away.