News for hackers. Stuff that matters.
Re:Experimenting with the YouTube audio fingerprinting system by mdd on Sunday April 19, 2009 @11:22AM (#569472)
We've been hearing lately that "the kids" prefer the degraded sound quality of MP3 files: http://news.ycombinator.com/item?id=511615

Anyone want to bet that, a year or two from now, we'll be reading that "the kids" prefer listening to sped-up music?

Also, I object to The Waitresses being labelled "one hit wonders". The author apparently forgot about the ubiquitous holiday song "Christmas Wrapping".

Re:Experimenting with the YouTube audio fingerprinting system by alx on Sunday April 19, 2009 @07:20AM (#569334)
Really interesting reading, to learn how Youtube build its signature, and the get the works-for-all solution: increase your copyrighted music speed by 5%, ingenius :)
Re:Experimenting with the YouTube audio fingerprinting system by kwamenum86 on Sunday April 19, 2009 @12:30PM (#569523)
"Apparently they don't really care about repeat infringers"

I would be careful about operating under this assumption. Google's habit of locking out accounts across all of their services is well-documented. You wouldn't want that to happen over something silly.

Re:Experimenting with the YouTube audio fingerprinting system by zandorg on Sunday April 19, 2009 @03:01PM (#569651)
As I keep saying on HN, if people hosted 9mb clips on their own webserver, with Flash video streaming technology, there would be no way of censoring it.

Not quite off-topic, but related is: Can you currently stream Flash video off an Apache webserver? If not, would a startup writing a plugin for it be a good opportunity?

Re:Experimenting with the YouTube audio fingerprinting system by quilby on Sunday April 19, 2009 @11:06AM (#569453)
Does anyone know if YouTube actually removes clips containing parts of copyrighted films? I doubt it.

http://en.wikipedia.org/wiki/Digital_video_fingerprinting

Re:Experimenting with the YouTube audio fingerprinting system by queensnake on Sunday April 19, 2009 @12:31PM (#569525)
Eh - what's the bet some do-gooder Googler reads HN and lowers the thresholds / throws in some more types of checking, and it's an arms race, and this person will have just helped them to improve. Still, /someone/ had to do it.
Re:Experimenting with the YouTube audio fingerprinting system by zemaj on Sunday April 19, 2009 @10:12AM (#569416)
Great read! It would be nice if someone tested if 5% increases in video content also get through. I wonder if this will lead to a new generation of piracy on YouTube,
Re:Experimenting with the YouTube audio fingerprinting system by kwamenum86 on Sunday April 19, 2009 @12:25PM (#569518)
A 165 second chunk from the end of the song passed as well. One may only need a slightly altered mix of the song to circumvent the audio filter.
Re:Experimenting with the YouTube audio fingerprinting system by eru on Monday April 20, 2009 @12:22PM (#570937)
How about only speeding up the first 30s for 5% and then slowly going back to the natural pace?
Re:Experimenting with the YouTube audio fingerprinting system by dejb on Sunday April 19, 2009 @08:23PM (#569983)
Wow he shelled out $349 for Adobe Audition 3. That is dedication.
Re:Experimenting with the YouTube audio fingerprinting system by norova on Sunday April 19, 2009 @10:28AM (#569427)
Fantastic read, thanks for sharing!