This post was sent to us by a friend who runs an indie label:
Just about a year after hiring two part time people, to do nothing else but issue DMCA takedown notices we’ve crossed the 50,000 notice milestone. The division of labor requires one person just to monitor YouTube, and another handles all DMCA compliant sites such as CyberLockers, Torrent Search Engines, etc.
The DMCA is BROKEN.
Most of the take downs are for the same title, at the same site, the same day. Day after day during the initial release period of the album (generally the first 60-90 days) it is a constant game of whack-a-mole.
We shouldn’t have to have the same title removed from a site more than once – and each time we issue a notice it takes 24 to 48 hours to remove. But, once it’s removed it is generally back on the site within a few hours.
We should not have to send a notice for the same title more than once, ever – Not to Rapidshare, not Grooveshark, not any one of the probably top 20 offending sites we track, and those are just the ones that even have a DMCA provision (The Pirate Bay for example does not, nor did Limewire to the best of my memory).
If site operators want to hide behind “how do we know what’s infringing”… Well, here’s how, we’ll let you know! If we issue you a notice, you now know… do you think the title will suddenly not be infringing the next day, when re-uploaded by the same offending person? Seriously? Does Billy in Pittsburgh suddenly own the rights to a Radiohead album (for example)?
Internet piracy apologists are quick to accuse labels and artists of wanting the government or others to become piracy police. This is simply not true. Most labels I know of have assumed the responsibility to track and issue takedown notices for themselves and on behalf of their artists (who should be focused on creating, not policing). Ironically, these same people are offended and attempt to diminish the issue when confronted with the overwhelming amount of takedowns being issued.
Keep in mind, we’re issuing DMCA takedown notices for ALBUMS not songs, entire albums are zipped as an archive and now distributed with as much ease as songs once were… let me say that again, our notices are for ALBUMS not songs…
There can be no question why album sales continue to plummet, and why digital album sales have leveled off… meanwhile, I suppose individual songs will continue to grow given the ease, convenience and low cost of a 99 cent purchases from iTunes.
The simple math says that if each of those uploaded ALBUMS was only downloaded ONCE by one other person, that is a loss of revenue of $350,000 dollars wholesale ($7 x 50,000). If each one we’re downloaded only TWICE that is a loss of $700,000 dollars in revenue a year ($350,000 x 2). This is just for ONE indie label tracking only it’s top five titles at any given time.
Yes, many will exclaim that not every illegal download is a lost sale (to the artist/label/rights holder). But, these numbers illustrate the financial impact of just ONE or TWO illegal downloads per DMCA takedown notice. I think any reasonable person would agree the number of downloads per upload is significantly more than ONE or TWO.
We only have the resources to track 5-10 titles at a time with any effectiveness. Catalog is a free for all. When adding in current titles that fall below the current top ten best sellers and a catalog that reaches back almost two decades the numbers become truly staggering.
This is why the number one agenda of the recorded music industry must be to address the illegal exploitation of artists work and closing this loophole in the DMCA, which is clearly not the intent of the law.
The Trichordist 
Another indie label owner writes :
I have taken refuge in the free for all of deep catalog where I agree, resistance is truly futile.
How can any one finance new recordings and their marketing given the present anarchy?
Even worse is trying to compete not only with free, but 80 years of more well-marketed back catalogs of material that people are used to hearing. It’s kind of hard to run out of ‘new’ music to listen to with literally millions of songs to choose from just in the popular artists’ lineups. It takes time for a listener to develop an ear for your music, as well. So you’re already behind in the game just by virtue of not getting in first. 😦 These are issues that record companies are supposed to help you with.
Joseph, not sure what your comment has to do with Robbies to be honest. There are many new and developing artists being promoted all the time. If you’ve ever turned on a radio you would hear a lot of new music and artists breaking on “Hit Radio” stations. Lady Gaga, Edelle, Katy Perry, etc all illustrate this fact. Most people like to hear and discover new music. The failure of the DMCA makes it harder to so, because much of the new investment goes to the ROI of sites operating illegally, not to the artists themselves, or the people making investments into those artists careers.
Unfortunately, the courts have consistently decided that it is not service providers’ responsibility to police their networks – see Viacom v. YouTube and UMG v. Veoh. The major media companies pay piracy monitoring companies like Peer Media and MarkMonitor tons of money to play the DMCA whack-a-mole game. These companies can automate the process to some extent, but for an indie label it’s not clear that the cost is worth it.
For example, there were some leaked emails several years ago between UMG and what is now Peer Media, suggesting that the cost to monitor a single track was (as far as I remember) $2k per track per month, and $4k per month for albums. Although UMG most likely only used this service on the most popular current titles, just imagine what the total cost would be to monitor UMG’s entire catalog of 3 million tracks or whatever it is.
YouTube does have a feature like what you’re asking for – you send them your content, they will watch for it on an ongoing basis, and give you a choice of blocking the upload or letting you share in the ad revenue. The majors have mostly opted for the latter. However, cyberlockers like Rapidshare don’t do this, and Grooveshark is a whole ‘nother story.
LimeWire claimed to have a filtering system that would block uploads, but it was a bunch of disingenuous cow droppings: it used a technology that was trivially easy to fool (for the techies out there, a simple hash algorithm). It was LimeWire’s way of claiming to be “copyright compliant” while actually being nothing of the sort. The majors weren’t amused and sued LimeWire out of existence.
That brings up a dodge that I’m sure Chris is aware of. They give a file a name that may or may not have anything to do with the content but that’s irrelevant. OK, so you have a tool that looks for file hashes of the entire upload, and inside archives and it blocks those that have already triggered DMCA notices. But then another user uploads a file that has a different hash (checksum) because either 1 bit changed or it’s an encrypted file. Youtube and some other sites are getting smart and inventing ‘signatures’ that rely on watermarking or some quality of the file that exists even at the lower bit rates (say, an approximate ratio of colors in an image’s palette) to get around the first but the second is the nature of the Internet and in fact any communication medium where anyone can ‘broadcast’ in the days of heavy encryption. (Yes I know it’s not technically broadcasting since it’s to a single other user of the network) A possible workaround for encrypted files is to get ahold of the keys that the pirate sites use. Quite often, they use the site’s URL or domain name! If they use different passwords for each file, then you would have to join the group to get passwords which is not too different in principle from how illegal use of Bit Torrent is tracked except it involves a more organic approach.
The point is that quite a bit of sites don’t really give a d*** because that’s what attracts users – the ‘free’ stuff.
Um, Joseph, what you just described is exactly what LimeWire tried to get away with and failed. Everybody else uses an acoustic fingerprinting technique, known to mathematicians as a “robust hash,” similar to what Google uses. The de facto standard technology for this outside of GoogleLand is from a vendor called Audible Magic. In fact YouTube used Audible Magic for a while until Google phased it out, but all the others (Dailymotion, Facebook, etc.) still use it.
So, today sites either use this type of technique or they don’t use anything. Nobody uses plain hashing anymore.
While I support you in theory there are some serious economic questions about your analysis that are not born out by consumer behavior. The “simple math” you refer to in order to calculate revenue losses makes a huge assumption that is simply not true. Each one of those illegal album downloads does not it fact have an actual dollar value of 7 dollars because you have no way of knowing how many people would download the album if it cost money. Record sales and consumer behavior would suggest that your “sales” downloads would be significantly less than your “free” downloads as consumer behavior changes drastically as soon as a price is attached. Significantly less people would buy the album. That is a simple economic reality that weakens the argument. Artists need to be honest and adjust their argument in light of this reality.
Sorry Joseph, that argument was addressed in the post and the “simple math” is used simply to illustrate the economic damages if even only one or two illegal downloads resulted from each upload. Considering there are 50,000 DMCA takedowns issued, and that the number of illegal downloads easily exceeds one or two per upload removed, the extended number of both illegally distributed unlicensed copies and the economic damages from it is rather large and easy to understand in this context.
Also, every illegally distributed download is not a “lost sale”, it is only a lost sale to the artist as the site illegally exploiting the artists work is being paid for access to those downloads by advertising revenue or subscriptions, or both. So the sale isn’t actually “lost” it’s just that the artists are getting ripped off.
Currently, illegally downloaded music is said to account for 95% of all music consumed online. No one is suggesting that if the illegal exploitation were magically stopped that sales would increase by 20 fold. However, If file sharing was managed to the level of retail shoplifting than reasonable people would expect to see the current numbers double or triple, which would put the recorded music industry on a trajectory that is consistent where it was headed prior to en mass unchecked illegal exploitation of artists work at the turn of the century.
There was a general working proposition at SNOCAP that if we could monetize 10% of p2p traffic, monetizing the second 10% would be exponentially easier to accomplish.
It is difficult in many instances to say that no portion of illegal sales are lost sales, because all sales are monetized. For example, Megavideo monetized sales either by advertising revenue, or subscription. All sales were monetized. There were no (or very, very few) truly lost sales. The question seems to be whether we want someone who does not pay a license fee and does not compensate film makers or artists to set the price and collect it for himself.
The people who start these sites do not want to be legitimate retailers. Yes, Megavideo now says, “you know, I was just about to launch a service that would pay artists, if you had just let me keep operating a little bit longer, I would have shown you how it was done.” If that is actually true–personally, I don’t believe it–it doesn’t explain what took so long.
The more likely explanation is that if these illegal sites can’t make money illegally, they will find some other way to do it. Sometimes it’s dating sites like Sam Yagin, sometimes its Skype, sometimes its something else–but what it has never been so far with one exception is a legitimate music or film company.
That exception was the original Napster and Shawn Fanning who went on to start SNOCAP, which was designed to solve many of the problems we still have today. That company licensed p2p technology and got all the major labels to agree to license p2p. The problem wasn’t the labels not license, the problem was that the illegal operators did not want a legal solution. I know that’s hard to believe because the myth runs counter, but I know they licensed because I did the deals and I know that the illegal sites wouldn’t take the deal because they told me so.
Digital music takes up physical space. Anyone who denies this doesn’t know what they’re talking about, or a liar. And since it is a physical, mechanical (hard drives move don’t they?) copy, a statutory mechanical royalty is due for every piece of digital music that ever changed hands. Many people are breaking the law.
That being said.
I’m somewhat baffled by the fact that YouTube and the rest can count visits and plays and charge $$ to advertisers based on this data, but are mysteriously unable to use that same data to pay copyright holders. But we know that the internet, as a whole suffers from a severe form of akrasia. So……..
Blanket licensing fees should be paid to PRO’s for songwriters & publishers. The labels should create their own collection organization and create their own formula for payouts to each label and then let the labels share that payout with the artists who have a stake in the copyright. If the sites don’t pay the fee, have them arrested.
The first thing to remember is that in pretty much every major country, a statutory mechanical license is a available to downloaders for songs. That means that all these services could get halfway legal prospectively at least if they complied, e.g., Limewire. They don’t because they don’t want to.
There are licensing agencies that cover sound recordings (SoundExchange, PPL). While the particular kind of licensing is not typically part of their remit, they are able to do a variety of licenses in limited circumstances. They have never been asked by the illegal services.
There actually many problems with blanket licensing in the “ISP licensing” model, this article outlines many of them http://www.semaphore-music.com/2MTP/gratis/CastleMitchellWhatsWrongwithISPMusicLicensing.pdf
The most fundamental problem with the ISP licensing model is the semi-Ponzi scheme aspect. That is, because the payment pool is made up of a series of fixed monthly payments, the rates payable to stakeholders (assuming they all agreed to this model) will decrease as soon as the rate of increase in the number of customers decreases or increase in the number of stakeholders is greater than the increase in the (a) rates or (b) number of subscribers.
Google makes more money by keeping the web topological rather than semantic. There’s an incentive for Youtube to not provide a structured tagging mechanism for users to categorize the content of videos. From a software perspective, it’s dirt simple. But you won’t find them doing that anytime soon because, in addition to making copyright violation easier to monitor, it would allow the outside world to index the videos. From a knowledge base standpoint, it is my opinion that Youtube is guilty of the wanton destruction of information – especially of the meta variety.
Not really needing this link published, but a good (though involved) read for understanding the difference between topological and semantic can be found at:
http://nplusonemag.com/the-stupidity-of-computers
You have to ignore the author’s worry about the singularity and use it as a history of how some of the various internet companies made their mark.
I have to wonder if some of this problem can be ameliorated by redefining “red flag” knowledge, a provision within the current DMCA that has basically been interpreted out of existence by judicial opinions.