YouTube

<sarcastic>Lemme know how that works out for ya?</sarcastic>

From what I've read, copyright owners have to submit a sample of the song/movie/tv show to Youtube for the software to compare against Youtube's collection.

Submit a sample of every item that might have a violation. :roll:

And how is any software going to compare Eric Clapton's (or Brittney Spears, or whoever) performance of a song against someone's bedroom performance of the same song?

<sarcastic>Lemme know how that works out for ya?</sarcastic>

From what I've read, copyright owners have to submit a sample of the song/movie/tv show to Youtube for the software to compare against Youtube's collection.

Submit a sample of every item that might have a violation. :roll:

And how is any software going to compare Eric Clapton's (or Brittney Spears, or whoever) performance of a song against someone's bedroom performance of the same song?
I'm wondering the same thing. That doesn't make any sense at all. Stupid, stupid.

http://youtube.com/profile?user=TheEarthBlues

They better not remove me. ;)

I suppose it had to happen - I mean, people enjoy watching bands and artists perform on youtube, and it doesn't cost anything to watch.

It'll go off-line altogether, and then come back as a subscription or pay-per-view site.....

:evil: :evil: :evil:

Vic

And how is any software going to compare Eric Clapton's (or Brittney Spears, or whoever) performance of a song against someone's bedroom performance of the same song?

I recently asked the same question of a company selling this type software, as we are considering such tools for use in cataloging unidentified cuts of the same tune -- e.g., live versions versus known studio cuts. The demo worked pretty well, but I would expect that, as that's the definition of a "demo" (working, that is). The answer I got on accuracy was surprising, claiming a high correct postive hit rate for same artist, same song regardless of live versus studio performance. I suspect it works on cross correlations (integrals of sliding time comparisons between reference and test sample) plus comparisons of sampled FFT (something like spectral analyses) stats. A strong correlation on one or both methods probably is good enough to nail greater than 80% of same arist covers. Further voice analyses may help improve this.

For different artist versions, it is even possible to identify a fair percentage of "covers" fthrough iltering and analyzing (ignoring voice, for example). The trick is eliminating the false positives. It seems this technology is working well enough to correctly identify a fair number of the copies. And in the biz and legal worlds, catching even 30 or 40% of the copies without snagging false positives is a good proposition for dishing out lawsuits and scaring people can't be so easily ID'ed.

I guess close might be good enough for entertainment industry lawyers. :roll:

A lot of the peer-to-peer lawsuits seem to be proceeding without even being close (suing grannies / pre-schoolers who don't own computers, "john doe" lawsuits, etc).

And what effect have all those lawsuits had? Not much in reducing sharing.
It has resulted in "better" sharing methods, though.

My prediction: more on-line video sites located in non USA-influenced countries, and even more copying/sharing than before

I'm gonna start downloading videos from youtube and stuff and putting the files on a disc or something incase they get taken down.

I don't see why not...some of the vids on there are precious to me.

I suppose it had to happen - I mean, people enjoy watching bands and artists perform on youtube, and it doesn't cost anything to watch.

It'll go off-line altogether, and then come back as a subscription or pay-per-view site.....

:evil: :evil: :evil:

Vic

I like that. I agree. They will sure come back as a subscription or pay-per-view site.....

-diatonick