Hacker News Reader

Hacker News

Comment by jackvalentine | Hacker News Reader

Full thread

jackvalentine

I presume this will have the same faults as YouTube's Content ID:

- Can't identify when it is fair use.

- Can't identify when the user has licensed the content outside of the system for use.

View on HN

Or even:

- Can't identify when the user is the actual copyright holder who the accuser licensed the content from.

These automated systems are always going to have lots of false positives because they only know as much about the copyright status of a work as they have been told. The data entry process that drives them is inevitably going to be incomplete and error prone.

Copyright enforcement in the digital age is a hard problem that I've not seen a clear solution for. The legal system around copyright that exists today is not scalable enough to handle the volume of works being produced and distributed. It's not clear what sort of system could ever hope to handle the rate of disputes coming from a site like Youtube in a judicious and economical fashion.

Youtube and Soundcloud's content ID engines effectively killed the genre of mashups. There was a pretty vibrant community with a lot of new artists on both, with rooms on plug.dj and even turntable.fm back in the day. Then the content ID engines hit and one by one every mashup artist (with the exception of girltalk) had their work taken down by poor fingerprinting. It was such a massacre we took to making http://mashup.fm which used youtube-dl to cache everything locally, but by then it was too late. All the artists stopped creating content because they didn't want to be permabanned from these huge services.

I worry what by-catch Microsoft's engine will leave in its wake.

Warning for those with sound turned on. The website automatically plays music at full volume.

I hate this. There really should be a setting in browser so that I don't allow sound from any website unless I explicitly click the speaker icon in tab bar. Annoying as hell.

If you use Firefox, go to about:config and set "media.autoplay.enabled" to false. A few web sites throw a fit when they can't autoplay, but that's a price I'm willing to pay.

It's probably one of the worst faux pars you can make in UX. I'd appreciate it if they wouldn't announce to everyone in the bathroom what I'm browsing

That’s actually a default in the latest nightly. Pretty nice IMO.

It'll be a default on Firefox 55. Or you can try the current Nightly.

YES!!! Thank you! :)

EDIT: though it's not exactly what I had in mind... I want content to autoplay as normal, just the sound should be muted until I unmute it...

You missed the biggest one:

- Can't identify content that has been cropped or had some sort of filter added.

It's amazing how many blatantly copyrighted videos (tv shows and movies) like this still appear on youtube, with ads.

Sure it can. The number of tricks you have to do to a video to circumvent the content ID system makes the video difficult to watch. For example, you have to stretch the audio by some percentage or else the audio fingerprinting will ID it regardless of your filters.

The early circumventers did, but the newer ones don't seem to mess with the audio as much. I suspect they either went a bit too far with the initial modifications or content-id had to be dialed back a bit.

For example, you have to stretch the audio by some percentage or else the audio fingerprinting will ID it regardless of your filters.

A few years ago, sped-up music was uploaded to YouTube in an attempt to bypass content ID, and with the realisation that this actually made it sound better, a whole new genre --- Nightcore --- was born.

...or so the story goes, but I've seen this explanation a few times before and it does make sense.

“Either he’s playing classical music at 78 RPM, or I’m still dreaming.”

http://www.gocomics.com/calvinandhobbes/1987/02/01

It looks like pirates have found enough ways around Content ID, since a lot of movies are available with the original + "full movie" in the headline... That's how I became fan of Air Crash Investigation and Mythbusters, btw.

After the live YouTube feature added there was a lot of channel with the Simpsons, I saw yesterday that they flipped horizontally the episode in order to don't be tracked.

1) Fair use is incredibly hard to identify, especially as the intent in the center of all of it.

2) This is easy once you go multi-site. We [0] are actually doing that across many sites.

[0] https://pex.com

1) Duh. Doesn't stop companies like Microsoft and YouTube saying "we've solved the copyright problem!" though.

2) This pex thing you've linked doesn't seem to cover the use-case I had in mind. I'm working on an analysis of XYZ Hollywood movie with a number of people working remotely (for argument's sake, with the consent of the publisher). I stick it in my OneDrive and share a link to them and they download it.

MS blocks it because I'm "sharing copyrighted materials" and they don't have the context in which I'm doing it legally.

If all you need to do is transfer the file, couldn't you just use something like WinRAR to archive the file, possibly with a password?

Pirates already do the same thing.

Why should I? I thought we were supposed to be making life easier by giving our lives to the six or seven companies that dominate our internet lives.

Sure, but you're using _their_ free platform to host _your_ content that _may_ be fair use. They have the right to be as legally cautious as they want on stuff appearing on their platform.

You have the right to put your stuff on your own platform without such oversight if you want and directly deal with the legality.

Reply on news.ycombinator.com