Slashdot Mirror
New Programs Fight GooTube Copyright Battle
PreacherTom writes "The specter hanging over GooTube for the past several months has been the issue of copyright infringement: will lawsuits eventually kill the $1.5 billion deal? In response, a vanguard of software developers is aiming to turn the tide on filtering copyrighted material — and their handiwork is expected to hit the market in the coming months. One example would be Audible Magic, a 'fingerprinting program' for video released a few days ago that promises to use peculiarities of recording and editing to tag and identify forbidden material." From the article: "Other outfits promise releases in the next few months as well, as they expect the video authentication market to be many times larger than the market for software that safeguards music copyrights. Just how much money is there in such filtering software? The market is at its inception, so estimates are hard to come by. But revenue from user-generated content sites should reach $850 million by 2010, up from $80.6 million this year, according to In-Stat. Software makers are eager to tap into such growth rates."
Is it wrong that I searched for GooTube on Google and tried going to gootube.com?
What about the use of copyright works as part of a composition or parody? There are many legal uses of copyrighted works that are not directly controlled by the copyright holder.
It ain't possible to make such software.
There's also an interesting side market of copyright infringement mercenaries--- in the classical music world, there are people who collect program books from concerts all over the world and meticulously double check whether the ensembles had purchased the rights to perform certain copyrighted works.
Frankly, I see this as proof of the effectiveness of the free market. Copyright becomes self-inforced, and the govt doesn't always need to be involved.
My point is that there is a will, so somebody will find a way...
www.jmagar.com
-
They'll meet at the gloryhole tonight.
You insert noise when uploading, remove said noise when you have retrieved it.
Next!
The Kruger Dunning explains most post on
There's a limit on how many videos of kittens and home jackass stunts people are interested in watching. And this tech seems like it could make AMVs impossible (probably a good thing).
Although, a youtube's worth of amateur porn would be unstoppable... I hope they're hosted somewhere without obscenity laws.
They just simply don't care whether or not Youtube can filter every piece of audio.
My prediction is they will sue for the simple fact they did commit copyright violations by allowing the clips to even exist. When Youtube eventually decides to settle, Universal will continue with the lawsuit simply to make an example out of them. Afterwards Universal will purchase Youtube to simply bury them.
Universal has done this before with mp3.com and I think they will do it again with Youtube.
In case anyone else is as confused/grossed at as I was when reading the summary... GooTube is apparently just a fancy pants way of referring to YouTube as newly acquired by Google.
Ugh, please don't coin this term to refer to Google buying YouTube. It sounds like some Elmer's glue product, or a porn site.
Got a great clip of Jay Leno? Can't post it to YouTube 'cause it will be auto-tagged?
Post a still frame followed by a message "for the rest of this clip, go to www.piratevideo.kn/abc123/JayLeno."
Google may delete it but it won't be automatically blocked.
There are many legal uses for copyrighted works.. including parody and criticism..
then there's the inevitability of false positives which ALWAYS happen with search algorithms.. or are you going to trust these particular coders to make absolutely bug free code which accounts for every single inevitability
With this said.. let's apply your principle to law enforcement.
We'll have robotic systems disable your car if you speed or violate any traffic law, with your only option being to send a letter snail mail to the state to request reactivation (I think that about approximates reporting a false positive to a webmaster).
what? your wife is in labor with contractions 3 mins apart? TOO BAD!
what? that guy behind you is shooting at you?.. well I guess you die then.. can't be driving unsafely now!
what? you happen to be more interested in keeping that guy from rear-ending you so you ran the stop sign? I'm sorry but you should have considered the possibility a dead car is less convenient than whiplash!
VLC FOR MAC IS DYING! IF YOU DEVELOP, PLEASE SAVE IT!!
fingerprints are a great idea... except so far experience have proved they just dont work. its not just sample/pitch etc, they are completely vulnerable to just lossy compression to the point that most digital fingerprints are lost simply by decrompressing a file and then recompressing it.
Its been a good few years since this was last contemplated with any real intention, and it was dropped for CD rootkits (ROFL).
Not only that, but how do they intend to actually add the fingerprints to files already in the wild
there's a hottie in accounting that i wouldn't mind showing my gootube to.
Loss in the digital/Analog conversion? Sure - but are you goign to be able to notice after GooTube reduces the image to 190x240 and compresses the living fsck out of it, and then blow it up to 240x320?
Personally, I think that if the media conglomerates succeed in destroying Web 2.0, people should simply abandon computers and IT. Slashdot will be remembered as a realm of relatively free speech. People will talk about "the good old days" of viewing video online on those "computers" while they raise vegetables and do the laundry. People should buy art made by their neighbours and relatives. IT'll be crap, at first, but as people practice, the work will improve. Neighbourhood live theatre. Let the cities turn to dust.
RS
Shoes for Industry. Shoes for the Dead.
The RIAA and MPAA got the DMCA 512 provisions passed so they could stomp on alleged copyright violations without bothering with going to court. In return, they threw the bone of safe harbor to ISPs, a bone which the ISPs actually already had thanks to the Netcom and some other decisions. Now alleged copyright violations happen fast and furious and the copyright owners find their own law means they _can't_ sue the ISPs provided that the ISPs respect takedown notices and have a policy against violations. Why should the ISPs go through any further effort to prevent violations?
The copyright owners made their bed, now they can damn well lie in it.
This is what all those people with copyright claims were waiting for. There would have been no point sueing youtube before, They had no money. Now youtube is owned by a group with very deep pockets it's time to go in for the kill.
because at the end of the day, What would copyright owners rather do? To send a C&D letter and have the content taken down? Or to have the video site constantly infringe on copyright which would mean the copyright owners can take them to court over and over again?
There is billions to be made from copyright infringment on the internet now. It's created a world of law-suites.
God Be Gone
Or lack of?
If you strip all the content that people care about, they will stop coming.
---- Booth was a patriot ----
You're confusing fingerprints with watermarks. An watermark is metadata stored in a signal using steganography, where the stego-system is intended to be resilient to minor distortions of the signal. A fingerprint is a hash value computed from a signal and used to access metadata stored separately. The thinking is that acoustic fingerprints can be made more robust to common audio codecs than watermarks, and they also satisfy audiophiles who won't stand for intentional signal degradation.
Stay away from this company. They called the president of our University to tell him we had illegal file-sharing on campus, and we *HAD* to buy their product before the RIAA/MPAA started suing us. Naturally our president freaked out. I'd stay away from a company that spreads such FUD.
It's not line its possible to do acoustic fingerprinting...
"It ain't a war against drugs.it's a war against personal freedom" --Bill Hicks
but hasn't Google been successful enough at holding litigation at bay for copyright infringements? I'm pretty certain that they have a plan that falls in line with 'organizing the world's information' as a slogan.
If Google comes out with a copyright holder protection scheme for YouTube that works, then they have all that they need to continue on until 'all the world's information is organized' by Google. That's not exactly a monopoly, but it certainly makes a big dent in the competition. I'd certainly be trying to do this if I was Google. Its smart, even if it might cause some short term pain.
Support NYCountryLawyer RIAA vs People
correct me if i am wrong, but i do not understand how this will not prevent videos you have made and then edited in premiere or whatever you choose.
For Youtube to compare just 1 video submitted to every work out there would be computationally prohibitive and for every work submitted computationally impossible, not to mention that they also have to have the works in memory to compare in the first place.
These algorithms are a nonstarter for Google.
Nice try, but they got rid of those little "loopholes" with the passing of the DMCA. There's no exception for the traditional rights to Fair Use.
Once they manage to put some form of DRM -- no matter how trivial -- on all major media distribution schemes, they will have effectively eliminated Fair Use, except for the anointed few that the Copyright Office deems worthy of receiving exceptions. And at least in my world, anything that you have to receive regular permission from an authority in order to do -- permission which can be denied at any time -- is hardly a right.
"Ladies and gentlemen, my killbot features Lotus Notes and a machine gun. It is the finest available."
Wow you should go work for Bose or somebody like that - hell you should work for NASA.
When I was at school noise was random and irremovable because, without reference to the original you couldn't tell what was signal and what was noise.
I take my hat off to you, removing an unpredictable distortion is an impressive feat - it would have to be pretty much random and unpredictable otherwise the fingerprinting software would just spot your filtered signal and correct for it.
"A slew of specialized software makers are releasing new kinds of technology that promise nearly 100% reliability in detecting copyrighted works."
Great. Now how about 100% reliability in determining whether an exemption such as fair use or education applies to use of the work? Because that's more important.
"98+% accuracy and virtually no false positive identifications"
That sounds like a mixed up quote of statistics from an inept marketing guru. Does that mean they call 2% "virtually no false positives" or does the program return an answer "I don't know what that is" for 2% of the time, and the false-positive data has been conveniently omitted? I'm going to err towards a worst-case scenario for the sake of argument.
Gracenote contains 69,462,367 songs in their database, and from those songs alone we could expect to see up to 1,389,247 songs incorrectly labelled. Multiply that by the number of distributed copies worldwide...
This message was scanned by European governments and contains no terrorism.
I fully expect that in a few years, devices like those TBCs will be unlawful to purchase if you're not a video professional; like scanners that can receive the 800MHz cell band. Video hardware that strips DRM, even "accidentally," will become contraband, like hacked Satellite TV descrambler cards.
I have heard that in the Soviet Union, every photocopier was serially numbered and registered. I could easily see a future in this country where that is the case for any device capable of removing the DRM from content. How else are you going to keep people from just buying 'professional' gear? They'll serialize them, register them to a list of approved owners, maybe toss a hefty tax and right-to-inspect on them, too.
Call me paranoid, but it's not hard to extrapolate an endgame like that from the Macrovision laws, and proposed Analog Hole legislation. Coupled with the tendency of our government to try and turn the screws when a law is demonstrated to be ineffective, versus taking a step back and reconsidering why it doesn't work (which might be an admission of failure); I think we could be filling out BATF forms in order to buy a time-base corrector before we know it.
Time to buy your "pre-ban" equipment now...
"Ladies and gentlemen, my killbot features Lotus Notes and a machine gun. It is the finest available."
Google's main business is the search and advertising market. Now take this into consideration when the copyright holders will be supplying them 'fingerprints' to all of their copyrighted content. It would cost Google a ton of money if they wanted to tag all of this content by themselves, but instead the copyright holders will be doing this for them, and will be happy to do it also. Granted I don't know how much money they have recouped from search results/advertising from their acquisition of Youtube, but I am sure that they spent more then a passing thought on this before they decided to buy it. So if you figure that they can get away with it long enough to recoup the cost of Youtube, and still acquire all of the data from these 'digital fingerprints', then they still have one heck of a deal going for them, provided they can avoid a costly settlement.
For fuck's sake... stop saying Gootube.
Don't be so snide. The idea's kernel is probably more accurately described as using some property of the input data (eg, a hash code) or even a decoration that can sit in the header that will provide a seed to some mutually agreed on PRNG that can be used to add and then subtract distortions to the signal.
The description by the GP isn't pedantic by any means, but he does have a point. I think.
You'd end up with the kind of running battles that go on between the ISPs and bit-torrent. If it's a regular pattern, then it will be identifiable and anything encoded or carrying some kind of obfuscation technology will be easy to block at the point of upload, if were to find a way of creating a removable wrapper so that it got through the fingerprinting process then what's to stop the fingerprinter taking a peek inside as part of the process. These people aren't idiots, the will adpat their technology. To keep up with the filterers you'd probably have to make it into an open-source project to enable enough people to work on it round-the-clock - the MPAA have already employed hackers to do their dirty work, so doubtless they'd find ways of infiltrating the project. Best thing would be to put the effort into creating your own content in the first place, that's what this whole 2.0 thing is about or just stick to using stuff released as part of the creative commons initatives.
If I upload a copyrighted image file to ImageShack or Photobucket without permission, does that make ImageShack or Photobucket liable for that copyrighted image? If I upload copyrighted content to a Geocities homepage without permission, does that make Geocities liable for that copyrighted content?
IANAL but from my understanding of DMCA Safe Harbour provisions, those who provide hosting are not legally liable for content on their servers unless they fail to comply with a DMCA takedown or something.
So what makes Google Video and YouTube different?
Why are they "on the hook" and being told that they need to activly remove copyrighted content?
I've said it before. The whole reason it is worthwhile is the ability to get clips of news shows and other similar things that I never watch and never record for myself. The amateur stuff on the site, with a few minor exceptions, that does not violate copyright also does not get me to look.
Dog is my co-pilot.
I'd like to see some sort of standard for identifying files as able to be redistributed.
I maintain a couple mirror sites and before I mirror anything, I read the Terms and Conditions, EULAs, whatever, to make sure that redistribution is implied to the point where I feel comfortable mirroring it, or expressly given.
While this wouldn't stop redistribution of copyrighted content, it'd make the lives of people that only want to legimately redistribute content much easier.
I don't know how such a system would work - it'd need to be embedded in all file types and work on all platforms. The only way I can think of that wouldn't involve serious changes would be standard filename naming conventions, but that is so low tech and lame.
I have heard that in the Soviet Union, every photocopier was serially numbered and registered.
You mean, unlike several brands of laser printers?
Using copyright protection to disable copyright protection? Priceless.
-Clio
Karma: Bad (mostly from not giving a fuck)
Blog: http://clintjcl.wordpress.com
The Audible Magic tagging is live on Grouper.com a tiny version of YouTube, as is everyone else in the field. Uploads are being stopped in their tracks, so to speak. Grouper forums discussion and reaction --> http://forums.grouper.com/showthread.php?t=3699 It's the beginning of the no-fun period of online video.