The only thing that impresses me about these is the amount of time and effort spent in collecting the pieces. Even there, I suspect it may be less work than I thought, considering the new automated tools that can seek out a keyword in the audio for you. But watching them, I just feel... "meh". I'm actually more impressed by the large multitrack musical performances that synch up many you tube clips, like the Gotye thread in the Editing forum.
[Mark Suszko]"I suspect it may be less work than I thought, considering the new automated tools that can seek out a keyword in the audio for you. "
Cynical, cynical, cynical. You guys are depressing the shit out of me.
Look, audio tools that can sweep the entire web, or even just YouTube - that's NSA stuff. Not even Viacom can do that. Software tools that normal buy can search the audio from what's IN YOUR BIN. You'd have to get the clips IN the bin first.
And you'd have to program the software to look for one word or phrase at a time, and you wouldn't know the phrases to tell it to look for. Even the phrase "I like big butts and I cannot lie" is a massive task. You'd have to have the clips that contain those words in your bin, ready to go. Not all of those clips are even on YouTube at all.
I can also tell you from having constructed demos for the exact voice software that we're talking about that it doesn't work GREAT. Merely very good. It took some work to define phrases it could reliably find just to reliably pull off a 90-second demo.
So you may prefer vertical editing over horizontal editing, fine. But this is a level of sustained complexity that most people in the COW will never even attempt.
I enjoyed it. It was difficult to make out some of the audio because of the quick cuts and background music/noise, but I appreciated it. I would imagine this project is more likely a result of crowd sourcing than it is software searching.