How to Navigate Fair Use with Modern Automated Copyright Detection
In case you’ve missed it, videomaking is the new black.
That includes vlogs, supercuts, video essays and basically any video under 10 minutes that is shared online for free by an individual with low-key means of productions.
Video essay, the genre that particularly matters to our community of learners has bloomed this last year thanks to the steady work of Tony Zhou, Kevin B. Lee and a few other essayists that have been trying to find innovative ways of analyzing film concepts, all the while keeping it accessible and entertaining.
This has lead to more and more film lovers crossing the bridge and starting making their own videos. But if YouTube or Vimeo let you share your work for free, they also have guidelines you need to follow and that include navigating the thin red line between fair use and copyright infringement.
That aspect concerns all the storytellers using pre-existing material (whether images, sound or both) to tell their story.
Here is the broad definition of fair use: (in US copyright law) the doctrine that brief excerpts of copyright material may, under certain circumstances, be quoted verbatim for purposes such as criticism, news reporting, teaching, and research, without the need for permission from or payment to the copyright holder.
The only problem? The robots in charge of scanning the million of minutes of online content uploaded each day might not agree with your view of what is fair use or not.
Here is what Zhou said when asked how he navigates fair use with modern automated copyright detection:
“I believe what I create fall under fair use.
Before I even started the channel, I spent a few days just extensively reading upon U.S. Copyright Law, I believe it’s section 107. There are a lot of nice resources out on the Internet as well, other videos, other videomakers, other YouTubers who do similar things.
The entire editing style of Every Frame a Painting was actually specifically designed to avoid the copyright bugs.
As far as we can tell, what they do is that they scan a file for the waveform, the sound. They then match that waveform with their very extensive database. They use the sound to catch it, and then they compare the image to make sure. If both of those things get hit, then you get a copyright flag.
Out of 18 videos on the channel, I would say a quarter of them have gone one copyright flag, two of them have escalated to take downs, but I’ve always got them reinstated.
But there’s a very real difference between what I think is fair use and what the robot think is fair use.
By the way if there’s anyone listening who wants to make videos, the general trick I would recommend you to do is don’t use more than ten seconds of any clip.”
To be clear, if you are planning to make videos using copyrighted material, what Zhou’s says is probably not enough to be well informed, but what he did to educate himself is likely to be what you should do too.
Onward to videomaking…
You can listen to the full episode below:[soundcloud url=”https://api.soundcloud.com/tracks/207533029″ params=”auto_play=false&hide_related=false&show_comments=true&show_user=true&show_reposts=false&visual=true” width=”100%” height=”300″ iframe=”true” /]