Thanks to Covid-19, we have been getting most of our dose of culture online, as we take in everything from virtual art exhibitions to live-streamed concerts.
Classical musicians have also been riding the virtual wave, performing live performances on social media, but something has been dimming their spotlight: copyright bots.
Classical musicians and organisations have been experiencing disruptions and even shutdowns during their live-streamed performances on Facebook and YouTube, as bots or content identification software flag their music as copyrighted material, a report from The Washington Post reveals.
How do they detect copyrighted material?
Powered by algorithms, the automated bots scour the internet for unauthorised or illegal use of copyrighted video and audio clips. They do this by analysing the content and comparing them to reference audio files sent by record companies and performing rights societies. Online content deemed in breach of copyright laws is then taken down or muted; you may have seen this in YouTube clips that won’t play due to a copyright claim by a company or when Instagram videos featuring pop songs suddenly disappear.
But classical music has proven to be a tricky area. The compositions of Mozart, Bach and Brahms, for example, exist in the public domain, which means their use is unrestricted.
However, if a musician releases an album with his or her own renditions of Beethoven, for example, with the record label Naxos, the audio would then be available for bots to use as reference to take down similar-sounding clips.
The automated bots simply do not have the ‘trained ears’ to differentiate between separate recordings of classical pieces.
When classical musicians play these works, they add their own flourishes or nuances to the piece, and certain sections may match copyrighted clips of recordings by other musicians or renditions of the piece, even though neither have ownership to the centuries-old original material, written by composers and musicians who are long gone.
But here's how they sometimes they get it wrong
In the case provided by The Washington Post, a man named Adrian Spence, the artistic director of chamber music ensemble Camerata Pacifica, broadcasted a previous performance of Mozart's Trio in E flat (K. 498), which was flagged by bots for containing a clip with "audio owned by Naxos of America".
Facebook suspended Camerata Pacifica’s access to live-streaming, while Spence disputed and cleared the claim.
In another instance, YouTube also blocked a live stream of a recorded performance of Carl Nielsen's Wind Quintet, Op. 43 by the ensemble, citing five copyright claims from different record companies.
“I have no protection for my own produced material,” Spence told the American publication. “If you want to put a copyright claim against me, I’m happy to take the time to write back to you and say, ‘This is an erroneous claim and here’s why.’ But when you’re immediately blocking videos or streams, that’s negatively impacting our very mission in a time where this now has become mission critical.”
So what does this mean going forward?
Spence’s example raises the issues of content recognition software, which relies on algorithms and not humans to decide on these breaches. Even the record label does not necessarily have control on who or what these bots take down and report to social media sites. Automation does not see the intentions of the musicians conducting these live-streams either, and cannot consider how these broadcasts are ways for them connect to audiences amid the pandemic.
Social media platforms such as YouTube and Facebook have been trying to refine their technology, according to the report. Measures include updates to music and video policies, including giving users earlier notice that their content may be flagged for copyright misuse.
However, while social media sites may be updating their systems, they also use third-party platforms, which may not necessarily make the same adjustments that take into consideration the complex aspects of classical music.
When it comes to the battle against the bots, there is still a long way to go.