Last week’s announcement that the U.S. Copyright Office had successfully accepted a bulk submission of notices of intent (NOIs) for compulsory mechanical licenses in electronic form marked a major milestone, both for the Office and for Music Reports Inc., which delivered the NOIs on behalf of music streaming service Guvera.
Music Reports has been working with the Copyright Office for more than a decade as part of the Office’s fitful, and at times halfhearted, effort to upgrade the creaky, pre-digital process for submitting and accessing music publishing information to at least 20th century standards if not quite 21st. Last week’s successful test run on the Office’s new, electronic submission system, involving about 100 tracks, is believed to be the first such hand-off.
“We’re now ready to start doing this at scale. It’s a big, big step,” Music Report’s VP and general counsel Bill Colitre told RightsTech.com.
But it was only one step toward solving what Colitre says is a much bigger problem: the vast and fast-growing amount of music being released on digital platforms today for which publishing information is not available, if it was ever collected in the first place.
“There are over 6 million new recordings being released each year, and the pace is accelerating,” Colitre said. “The reason it’s accelerating is obvious. Recording technology has gotten so cheap that every kid in his garage is making a record. And the aggregators [i.e. digital music distributors] will scope that up and put it into the commercial stream. But most of those kids have no idea how publishing works or what data needs to be provided to register something.”
Under U.S. copyright law, musical compositions are covered by a compulsory mechanical license, allowing anyone who wants to make mechanical use of the composition, such as by recording it or incorporating a recording of it into a streaming catalog, to do so upon sending a notice of intent to the owner of the publishing rights and paying a statutory royalty.
Where the publishing data is known, available, and complete, either directly from publishers or via service providers like Music Reports, streaming services (or their agents) can send NOIs directly to the rights owners. Where the information isn’t known or directly available, streaming services can search the Copyright Office’s registration database to try to find the composition and its rights owners. If the data are not in the database streaming services and other mechanical users must file the NOIs with the Copyright Office itself.
Until a few months ago, however, that process was time-consuming, labor-intensive, and costly. NOIs had to be filed individually, on paper, along with a fee of $2 per song in addition to a $75 upfront filing fee. The Copyright Office then needed to process those submissions by hand before they could be searched. As a result, some users skipped the NOI step altogether, along with paying the required royalties.
“There are two separate issues,” Colitre explained. “First, you need to have the publishing data. The second issue is, you need to match the publishing data to recordings.”
Music Reports has metadata on more than 7o million recordings in its own Sondex database, but only about 16 million of those are fully mapped to the underlying compositions, “and we probably have the biggest matched database in the world,” Colitre said.
The Copyright Office began trying to update its internal systems to be able to accept NOIs electronically as early as 2004, working with Music Reports and others. But progress was slow and in 2010 the Office’s IT budget was severely cut, effectively killing the project. Since then, however, a series of high-profile class action lawsuits against Spotify and other streaming services over non-payment of mechanical royalties refocused attention on the issue and helped restart the upgrade project.
The Copyright Office’s new system allows NOIs to be submitted in bulk, via spreadsheet, and lowers the processing fee to 10 cents per song, greatly reducing the cost of complying.
“It means we’re going to be able to do this much more efficiently and at much greater scale,” Colitre said.
And scale is critical.
“There are 72 million recordings in the commercial ecosystem, maybe 125 million if you count mashups,” Colitre said. “That doesn’t mean there are 72 million compositions, because Jimmy Hendrix’s [cover of] All Along the Watchtower is probably on 17 different albums. But it’s a lot.”
As a practical matter, a mere 2 percent of those songs generate 98 percent of plays, according to Colitre, but the problem is growing, at an accelerating pace, due to the explosion in new recordings.
“You’ve got half a million new recordings coming into the system every month,” Colitre said. “That’s triple what it was a few years ago and it will probably be more a year from now, because the growth is accelerating.”
Most of those new recordings are coming into the system without proper publishing data attached, creating huge potential liability down the road.
“We have pretty good data on the head of the snake. But the long tail of content creators — the kids in the garage — don’t know anything about publishing is or what you need to do to register something,” Colitre said. “It’s a huge education challenge. You can come up with whatever configuration of pipes you want, but if you don’t solve that education piece you’re not solving the problem.”