An announcement from the startup Dot Blockchain Media a couple of weeks ago heralds an exciting new development in the potential for blockchain technology in the media industry. Dot Blockchain has partnerships with the New York-based recording artist STOLAR and the indie distributor FUGA. These partnerships enabled Dot Blockchain to take STOLAR’s track “Forget And Feel” and place an entry for it on a blockchain.
But Dot Blockchain has taken an important step beyond that: it has embedded the blockchain address for the track into the track itself, using audio watermarking technology from Digimarc. This means that anyone who comes upon the file can use the identifier to read metadata, rights holder, licensing, and transaction information from the blockchain.
Unlike metadata stored in file headers — which anyone can alter or delete, unless it’s protected by some sort of DRM scheme — the identifier is inextricably bound to the file and very difficult to remove without marring the audio. The data it leads to can be changed on the blockchain, so that anyone who reads the address by using Digimarc’s watermark detection app can see the data. And the data can change as necessary, but only those with permission (such as the artist or distributor) can make changes.
This is the vision for watermarking and blockchain technology that I laid out in the whitepaper I published last April, with support from Digimarc. I don’t claim credit for coming up with the idea, but it’s great to see the actual implementation.
Some startups’ visions include the idea of digital content files themselves existing on blockchains, even files that are licensed or purchased by users. I don’t subscribe to this idea, for reasons that go beyond the current limits in blockchain efficiency and scalability that make it impractical. Instead, there are many good reasons to put metadata, licensing information, and transactions on blockchains instead of monolithic databases maintained by single entities; see the whitepaper for more on that.
The music industry has largely ignored watermarking for a long time. The one mainstream use of watermarking that I’m aware of is not very sophisticated: some labels use it to embed simple identifiers of retailers (e.g., Apple or Amazon) into files they send to those services, so that allegedly infringing files found “in the wild” can at least be traced back to a retailer without implicating an individual user.
Instead, many applications currently exist for music fingerprinting. In a fingerprinting scheme, an algorithm makes an educated guess at the identity of a music track by calculating some salient characteristics of the file (the fingerprint) and looking those up in a database of fingerprints to see if there is a match. Although the leading fingerprinting schemes are decently accurate, the limitation with fingerprinting is that every instance of a given music track has the same fingerprint; while in contrast, multiple files containing the same music track can have different data embedded as watermarks if the files are to be used for different purposes.
Thus, Dot Blockchain Media’s use of watermarking for music is among the most significant and promising ever. Dot Blockchain is not a solution provider but a developer of technology architecture designed for others to build solutions on, organized as a public benefit corporation (think Kickstarter). Kudos to Dot Blockchain for taking this step, and let’s hope that others take up the mantle and continue the experimentation.
If only this solved an actual problem. Who’s IP did they use for the watermarking?
Read the whitepaper (this isn’t about piracy).
Digimarc’s software and IP.
[…] We use invisible watermarking because of its robustness (test it here) and security. Though this technology is centralized by design, it is not incompatible with a blockchain service, as proven by this initiative in the music industry : Dot Blockchain Media Makes Blockchain Plus Watermarking a Reality. […]
[…] Nous utilisons un filigrane invisible en raison de sa robustesse (à tester ici) et de sa sécurité. Bien que cette technologie soit centralisée par conception, elle n’est pas incompatible avec un service blockchain, comme le prouve cette initiative dans l’industrie de la musique: « Dot Blockchain Media Makes Blockchain Plus Watermarking a Reality. » […]
This technology isn’t easy to succeed because the limit in size of the data collected from the blockchain which is actually stored into the music file.
Take into account that if I hash a file I start from the name, size, owner and timestamp and obtain a string which obviously can be too easy inserted into the audio file. Problem is when a third party comes to me with a file having just 1/100 seconds be it more or less than the initial file size. I hash it again and award a “owner certificate”. Problem is who can really pretend to be creator of the music file ? The one who firstly asked me to stamp his file ?
And third scenario is if the evil party cuts a 1/100 seconds then add instead another 1/100 second in the beginning or the end of the file. Would the singer be able to easily identify which file is of himself ? I doubt that….
Everything comes to a situation of “who came first to the copyright technology enabler” actually ….
Mind you that I also work on such a technology through a start-up, actually using the Etherem blockchain to develop my services for my customers (I am based in Europe).
That’s all true. Blockchain doesn’t solve the problem of verifying rights ownership any more than existing techniques do, though existing techniques could apply to blockchain applications. For example, the music publishing industry currently uses Interested Party Identifiers (IPIs) to identify songwriters; you have to register to get an IPI. ISNIs (International Standard Name Identifiers) are becoming popular for various types of creators too. Registry of copyright information on a blockchain could require one of those identifiers.