Slashdot Mirror
Analyzing YouTube's Audio Fingerprinter
Al Benedetto writes "I stumbled across this article which analyzes the YouTube audio content identification system in-depth. Apparently, since YouTube's system has no transparency, the behaviors had to be determined based on dozens of trial-and-error video uploads. The author tries things like speed/pitch adjustment, the addition of background noise, as well as other audio tweaks to determine exactly what you'd need to adjust before the fingerprinter started mis-identifying material. From the article: 'When I muted the beginning of the song up until 0:30 (leaving the rest to play) the fingerprinter missed it. When I kept the beginning up until 0:30 and muted everything from 0:30 to the end, the fingerprinter caught it. That indicates that the content database only knows about something in the first 30 seconds of the song. As long as you cut that part off, you can theoretically use the remainder of the song without being detected. I don't know if all samples in the content database suffer from similar weaknesses, but it's something that merits further research.'"
There's the open-source library - libOFA - developed by Music IP (http://code.google.com/p/musicip-libofa/) which happens to create PUIDs on the first 135 seconds of audio in a track. It's used in the music-IP mixer (for mood mixes) but is also used by music database projects such as MusicBrainz.
From what I've seen, it's pretty decent audio fingerprinting, but I'm sure would be subject to the same limitations- if you remove the first 30 seconds of a clip- it would produce a very different fingerprint.
There's no reason to believe youtube isn't using this library or a derivative. There's also no reason to believe this result isn't intended. If the first 30 seconds of a song are missing- maybe that makes youtube confident that it could be considered fairuse.
Either way, I could imagine creating a fingerprint based on different sections of a song has the same problems doing an MD5 hash would- each fingerprint would be entirely different. If you don't just compare bit-to-bit, it'll be impossible to catch ALL permutations. And the fact is, that's a lot of computing power anyhow.
Belief? Hope? Preference?The Existential Vortex
I thought the purpose (however misguided it may be) was to prevent people from uploading copyrighted songs/music videos and re-mixing them. So if I only use portions of the song that aren't in the first 30s I'm home free? That seems silly, the system must still be under refinement or is only there to stop the most blatant offenders.
It's a good thing no one at Youtube reads Slashdot. Otherwise they might come up with a fix! So, everyone keep this a secret! SHHHH!
Give a man a fire and he'll be warm for a day. But light a man on fire and he'll be warm for the rest of his life.
Here's an idea. Start out the video with a useless narrative for the first thirty seconds "blah blah blah skip until :30 and ignore this intro blah blah" then start the music. That way everybody is happy. All google employees are too elitist to read slashdot, right?
Take the cheese to sickbay, the doctor should see it as soon as possible - B'Elanna Torres, "Learning Curve"
This seems like a good time to pump my own open source project: pHash. pHash is a perceptual hashing library that computes hashes for audio, video and image files, with text and PDF hashing coming soon. We use an algorithm similar to YouTube's audio fingerprinting method but we do not only take into account the first 30 seconds. Although, it's impossible to tell from this basic test whether their algorithm truly only looks at the first 30 seconds, or if the algorithm considers them to be different audio files. If the song is only 1 minute in duration, and 30 seconds is blank, is that really the same audio file as the full 1 minute version? At some point the audio files are not really the same anymore, although the perceptual hashes should be somewhat close to each other. Please give pHash a try. We could use some feedback from the OSS community and would appreciate it greatly.
That cool tech like this is being used to prevent "piracy" instead of something more useful.
Web2.0: I love when people Flickr my cuil and digg my boingboing until my google is reddit and I start to yahoo
Then a car commercial parody I made (arguably one of my better videos) was taken down because I used an unlicensed song. That pissed me off. I couldn't easily go back and re-edit the video to remove the song, as the source media had long since been archived in a shoebox somewhere. And I couldn't simply re-upload the video, as it got identified and taken down every time. I needed to find a way to outsmart the fingerprinter. I was angry and I had a lot of free time. Not a good combination.
The guy who wrote TFA is upset that his largely unviewed videos didn't pass an automated test.
My beef with the system is that when culurally significant videos such as the Chinese "Caonima" get taken down because the song violates some copyright of a company I've never heard of on a song I've never in a million years think of buying.
Hope that link works, I had to copy it from Google since I can't even access Youtube anymore here in China.
Why exactly does it merit any research? This is not riddle posed by Nature — people devised this device (ha-ha), and know all the answers perfectly already, they just don't want to tell you. You are not advancing scientific progress by figuring out somebody's scheme.
You may be advancing your own knowledge and skills, but calling it "research" has no more merit, than paparazzis' "research" into celebrities' lives...
In Soviet Washington the swamp drains you.
And who fingerprints the analyzers who analyze the analyzers?
Karma: Excellent. 15 moderator points expire sometime.
An unfortunate result. The last 30 seconds of most songs are not usually as interesting as the first 30 seconds.
I wonder if he tried mangling the first 30 seconds at all. For example, keep the first 5 seconds, mess up the 6th and 7th seconds, and then continue on. Or perhaps adding in a base line that would be hard to hear. Or something at the high end of the audio frequency spectrum, to annoy all those teenagers while I listen to my free music in peace.
Music Genome
Youtube uses Audible Magic's audio fingerprinting technology, which is based on this patent by MuscleFish: http://patft.uspto.gov/netacgi/nph-Parser?Sect1=PTO1&Sect2=HITOFF&d=PALL&p=1&u=%2Fnetahtml%2FPTO%2Fsrchnum.htm&r=1&f=G&l=50&s1=5918223.PN.&OS=PN/5918223&RS=PN/5918223
imeem have been doing this for the last few years, and they don't use audible magic, they used the Snocap fingerprint system which apparently was good enough for them to buy Snocap. Their business model has always been built around using the content identification system to make sure the right people get paid for audio played on the site.
imeem is primarily used by people uploading and sharing audio, so using an audio fingerprinting system seems more appropriate than youtube relying on an audio fingerprinter for video content.