CD Baby just inked a partnership with Audible Magic to identify potential rights infringement claims on tracks before distribution.
Infringement claims are becoming increasingly common and contentious. Now, CD Baby is hoping to address potential issues before they crop up — Minority Report style.
Audible Magic is a digital copyright, licensing, and monetization management company. The company’s software uses algorithms to identify copyright-protected songs quickly.
CD Baby, just recently swallowed by Downtown Music, will begin using Audible Magic on new uploads effective immediately.
Here’s how the pair-up will work. When an artist uploads files to CD Baby, the Audible Magic ‘RightsRx’ program will scan the file against its library of 30 million samples. If the track contains potentially copyright-infringing material, CD Baby can flag or flat-out decline the file.
The scanning process reduces the chances of CD Baby receiving a DMCA takedown notice after a track has been published.
CD Baby says the vast majority of tracks will be uploaded seamlessly and ready for distribution.
Downtown Music Holdings acquired CD Baby and its parent corporation, AVL Digital Group, for roughly $230 million last month. The new partnership will help CD Baby police its growing catalog of independent artists more efficiently.
CD Baby’s director of digital operations, M.J. Woodis, says the partnership is a necessary step. CD Baby has a 19-person team dedicated to the issue.
“We work closely with DSPs, and it’s often more of a conversation, not just rules laid down. The heart of the matter is not to restrict as we get them music and data, and to enable search and other discovery methods to help our artists. It’s a growing process.”
Last month, CD Baby began offering audio and banner ad opportunities to self-managed and unsigned artists. At present, the company is one of the largest distributors of independent music with more than 650,000 artists under its umbrella. CD Baby says it manages more than 9 million tracks spread across 100 digital service platforms across the globe.
This is really interesting. Music is very identifiable by a sonic fingerprint in the waveform of a recording, but how is the similarity of separate songs measured by AI? If every I-IV-V with a similar beat was ruled out, we would have never had many of the great artists we know today. I wonder what the metrics are and how sophisticated the decision rules are. If anything, I wonder how many tracks it will flag that would have been out otherwise.