My Bizarre Run-In With Content ID

I recently uploaded a Diablo-themed Special effects video to my YouTube channel, continuing the pattern from that earlier Nerf one.  I'm pretty happy with how it turned out, but the story behind it highlights a corner of the copyright-related minefield facing content-creators on the web today.

The video is an original work made by me, starring Bill Lyon, a former-colleague and current-collaborator of mine.  All of the audio in it was either recorded by my own microphone, or came from one of two fully licensed stock libraries:  the included library that comes with Final Cut Pro X, and the "Pro Scores" music collection by Video Copilot.

When I uploaded the video to youtube it was flagged by their Content ID system as matching a track owned by AdRev.  The track they were claiming?  One of the pieces from Pro Scores!

Now it gets crazy, because AdRev isn't even the bad guy in this story...

For anyone unfamiliar with YouTube's Content ID system, it's a fully automated process whereby copyright holders can submit their work (video or audio) to Google, and it is then compared against new videos that are uploaded.  If the software detects a match, it will automatically get flagged.  What happens next depends on what the rights holder has chosen.  They may choose to block the new video entirely, or allow it to remain up.  However the most popular option has been to leave the video, but "monetize" it with advertisements, the revenue from which goes to the original copyright holder.  Naturally, this prevents the uploader of the video from monetizing it themselves.

This system has, for the most part, been an extremely useful and pragmatic Faustian bargain for YouTube and its users.  The ability to monetize infringing works has so far been sufficient to placate rights holders, and prevents the YouTube library from being eviscerated by clamping down on the free-for-all nature that made it popular in the first place. A few of my own videos, like Project Gravity, could not remain on the site without this largesse.

However, every system does have flaws.  As it turns out Video Copilot has been having issues for some time with unscrupulous third parties fraudulantly claiming Pro Scores music as their own, submitting it to Content ID, and monetizing videos that use it.  This doesn't sit well with Video Copilot, since the "infringers" being denied full rights to do what they like with their videos are the company's own customers!  As I learned from their site, Video Copilot was eventually successful in getting YouTube to shut down the fraudulant flags, but without the music "assigned" to someone in the Content ID system, there was currently nothing preventing it from happening again.

The solution?  Video Copilot partnered with AdRev, a company that handles this sort of copyright enforcement on an outsourced basis, and registered the music themselves.  Now, videos that included Pro Scores music would still get immediately flagged, but contacting Video Copilot's customer support would let you put your channel on a whitelist to be excluded from the process.

So, after a confused weekend of sending emails to both Video Copliot and AdRev (just to be safe...) my video is now up for the world to see, with full rights retained by myself.  I wish I could end this post with a grand proposal for how to avoid the headaches I went through, but everyone involved seems to be doing the best they can given the odd constraints of the situation.  (Except of course for the people who claimed the Pro Scores audio as their own.  Not cool.)  Regardless, it remains a great example of just how strange the intellectual property landscape has become.

And if you'd like to see the video in question, here it is!

Beautiful Google Ad

I actually spotted a Flash ad for this video on my webcomic several days ago, and word is that Google is going to be playing it during the Superbowl.

Simple, honest, emotional...  now that is how you do advertising.

When "Targeted" Ads Fail

I've been hearing a lot of talk lately about Google's decision to start using AdSense to deliver ads that are targeted not only based on a web page's content, but also on each user's history and interests. The tactic isn't new, only the pervasive, widespread reach of Google is.

The privacy concerns of this have been talked about endlessly, and I'm not writing to add to that discussion. My question is: privacy aside, how many companies actually have the intelligence to even use such a system effectively? Theoretically, the more information you have on someone, the better your targeting should be. Given this logic, a service like Facebook, which knows a lot about me, should have very-well-targeted ads.

Theoretically.


I mean, come on. This isn't rocket science, people.

Google Experimenting With Customizable Search Results?

Today I noticed some new icons next to my Google search results...


Seems they're part of a trial "Edit Results" program Google is testing on a random selection of people with Google Accounts over a few weeks. They set up an informational page to explain it more.  This feature only applies to people who are logged in to their account when searching, and it only affects the result list of that particular user.

You can click on the Up arrow to promote good results to the top, delete bad ones altogether, and add notes about different results.

This seems like an interesting way to approach user-feedback-driven search, without actually making Google search results user-feedback-driven.  However, if they could figure out a more configurable way to do this (split into different "Search Sessions," perhaps?) I think it could actually be very useful when trying to do research online, especially with the note-taking feature.