Will Snocap make music sampling work?

The New York Times has another interesting article today "Putting the Napster Genie Back in the Bottle". The story is about Shawn Fanning's new company Snocap. You might remember that Shawn was the creative genius behind Napster. Snocap is almost the antithesis of Napster.

Snocap technology creates a "digital fingerprint" of every song file it finds on the Internet. It does this by analyzing the song file as a binary file of ones and zeros and comparing them to other known files. It doesn't matter how the file name is changed, or if the file was shortened, the Snocap technology can uncover copyrighted material and either block it from download, or charge the user a fee agreed to by the copyright owners. Music owners can go to Snocap to "claim" their music and decide if they want to block it, or establish terms for sale. But for this to work, the P2P networks must also agree to employ the Snocap technology, and so far only one file sharing service has signed up.

The technology works but making it a business success will be very difficult. Jonathan Spalter, former CEO of Snocap says "They have a shot, but it's a nine-bank billiard shot and they have only one stroke of the cue to get it right," Mr. Spalter said. "You need to get consensus from a firmament of major labels, independent labels, the publishers, courts, legislature, the peer-to-peer to companies, retailers and other actors. This is the ultimate, purest form of herding cats."

The article quotes Wayne Rosso,  CEO of Mashboxx. "Mr. Rosso says that this tests the theory that people use peer-to-peer, or P2P, file-sharing networks to try out songs, and that they will pay for music they like. "It is the great hypocrisy leveler," he said. "We will see if people use P2P to sample music or not."

When I was at Napster I proposed the idea of a free sampling service and a paid subscription service. The idea was that the free file sharing service would share low bit rate and therefore lower sound quality songs for sampling. We would block the transfer of high quality, high bit rate files. If the user liked the "free sample" song and wanted to own a high quality version, complete with liner notes, lyrics, and cover art, they could purchase it from Napster. The Snocap technology will allow the concept to be tested, and see if consumers will make it commercially viable.

I wrote an earlier post "Napster - the inside story". It is sad to remember this all could have been avoided. "We made one last effort to convince the labels that they should do a deal with us. A little known underground product called Gnutella had just surfaced. It was a P2P file sharing program that required no central server and no company to operate it. Gnutella was an open source program and we were already seeing new variants of it emerge. We told the record labels that we (Napster) had a loyal audience of over 50M users. We had servers that could control distribution. If they didn’t do a deal with us and put us out of business then Gnutella and its derivatives would become unstoppable. There would be no company to sue and no server to shut down. If we worked together now we could convert the market to a paying subscription or per download model. If we didn’t do a deal chaos would ensue. They didn’t believe us and didn’t really understand what this Gnutella threat was. "

We are still in the very early stages of the digital music era. There will be new technology and new business models in the future that will effective capitalize on the enormous market for music.

Comments

[codey]

This can be so easily worked around that it is stupid how Shawn can get away with it. Adding either random values in obscure fields of the ID3 tag or padding every file with a second or two of silence or random inaudible noise would beat the fingerprinting easily.

[DonDodge]

Snocap digital fingerprinting of files is not easily thwarted. Shawn is definitely not stupid. I worked with him at Napster and can tell you he thinks outside the box.

My guess is that the digital fingerprint is done in something like 10 second intervals across the entire file. These clips are compared to the master file clips for matches. If the matches exceed some threshold...it is a copy and will be blocked. ID3 tags are useless for identifying copyrighted files, and are probably ignored.

[Avikk Ghose]

Snocap's technology is impressive, but it's unlikely that there will be substantial market demand for digital content that incorporates multiple restrictions such as various DRM systems, price points, and file formats.

With respect to desktop and mobile consumption environments, Snocap's model would be significantly more compelling if:

a) Snocap-enabled content services could distribute media files with no DRM systems
b) Snocap-enabled content services could offer standardized, consumer-friendly pricing and usage tiers
c) the availability of unauthorized p2p file-sharing services was curtailed or nonexistent

Can Snocap overcome some of these (and other) challenges? If so, this concept could be extremely successful.