Commercial product but brilliant reality hack.(celemony.com)
celemony.com
Commercial product but brilliant reality hack.
http://www.celemony.com/cms/index.php?id=dna&L=0
13 comments
Why would the age of celebrity end? High quality output becomes much easier, but the authorship is still difficult (and requires talent). Photoshop and Illustrator definitely changed the game in the design world, but there are still clear differences in talent.
Ok, What I was getting at here is that ordinary people will see in a very real way just how easily stars like Britney can be manufactured. Photoshop has already come a long way in exposing to everyone in a visceral and easy to understand way the simplicity with which images can be altered. Everyone kind of knew this before hand but its now becoming ingrained.
I wasn't suggesting that everyone will suddenly be a virtuoso, I'm just hoping that as people in the future are presented with manufactured stars the response will change to "big deal, I know a guy who can do that on his mac with a girl down the street. They're really good. Here, listen...".
Of course, this won't fix the publics fascination with giving people who can't handle it celebrity status and too much money and using the inevitable train wreck as entertainment. At least we'll quit pretending its music.
I wasn't suggesting that everyone will suddenly be a virtuoso, I'm just hoping that as people in the future are presented with manufactured stars the response will change to "big deal, I know a guy who can do that on his mac with a girl down the street. They're really good. Here, listen...".
Of course, this won't fix the publics fascination with giving people who can't handle it celebrity status and too much money and using the inevitable train wreck as entertainment. At least we'll quit pretending its music.
"could the age of celebrity finally be over?"
Thats like saying that the piano would kill all virtuosos, because all the notes are already there, as opposed to a violin where the player has to press the string on the right spot.
Just getting the pitch right is not gonna make the music, not by a long shot.
Thats like saying that the piano would kill all virtuosos, because all the notes are already there, as opposed to a violin where the player has to press the string on the right spot.
Just getting the pitch right is not gonna make the music, not by a long shot.
OH. MY. GOD!!!
Since when does something that's impossible in theory can be possible in practice?! Or maybe the theory is simply wrong?
This is certainly a game-changer.
Since when does something that's impossible in theory can be possible in practice?! Or maybe the theory is simply wrong?
This is certainly a game-changer.
It's not really impossible in theory, or possible in practice. Or something.
They just changed the rules a little bit. Since one goes into it knowing that the music is being played on even tempered instruments, you've already ruled out a near infinite array of possibilities. It's just the remaining infinite array of possibilities that they have to deal with. So, you see, it's simply a matter of reducing the problem to its simplest form.
Seriously, though, if given a very complex piece of music played on oddly tempered instruments instead of a traditional piece of music played on even tempered instruments, this software would croak (unless they specifically trained it to deal with those tunings, and even then if the music is particularly complex/dense, there would be artifacts). But, since it knows the basic rules by which the music is playing, and traditional harmony is not very dense--stacked thirds are pretty wide open--it can mostly pick out the bits it needs. If you look at a spectrum analysis of a piece of music at a reasonably high resolution and sample rate--a single track with a single instrument, of course--you can quite easily pick out the individual notes. Combine this kind of analysis with the existing auto-tune functionality (up or down sampling a sound without affecting time) and you've got this tool.
I'm not saying it's not cool. It definitely is. But, it's been possible for quite some time to do this manually...the magic is in the automation, and the accuracy of it. (Though I'm betting there will be artifacts in many situations.)
I don't record music much anymore, but I'll probably pick up a copy of this just for fun.
They just changed the rules a little bit. Since one goes into it knowing that the music is being played on even tempered instruments, you've already ruled out a near infinite array of possibilities. It's just the remaining infinite array of possibilities that they have to deal with. So, you see, it's simply a matter of reducing the problem to its simplest form.
Seriously, though, if given a very complex piece of music played on oddly tempered instruments instead of a traditional piece of music played on even tempered instruments, this software would croak (unless they specifically trained it to deal with those tunings, and even then if the music is particularly complex/dense, there would be artifacts). But, since it knows the basic rules by which the music is playing, and traditional harmony is not very dense--stacked thirds are pretty wide open--it can mostly pick out the bits it needs. If you look at a spectrum analysis of a piece of music at a reasonably high resolution and sample rate--a single track with a single instrument, of course--you can quite easily pick out the individual notes. Combine this kind of analysis with the existing auto-tune functionality (up or down sampling a sound without affecting time) and you've got this tool.
I'm not saying it's not cool. It definitely is. But, it's been possible for quite some time to do this manually...the magic is in the automation, and the accuracy of it. (Though I'm betting there will be artifacts in many situations.)
I don't record music much anymore, but I'll probably pick up a copy of this just for fun.
This was a great tool, but with DNA it is absolutely awesome.
Audio processing/production is what got me into programming originally.
Audio processing/production is what got me into programming originally.
DNA is the first thing in a long time that I've seen that I really don't know how they did it! I think usually I've got some guess, even when its outside of my expertise (most hackers do) but this?
Fourier transforms and lots of 'em is as close as I can get. Its magic. I love it.
As a musician (and amateur recorder) myself, I can say that this is truly incredible. Think about it. As this technology progresses there may come a day when a single microphone is all it takes to record all tracks (and sub tracks of the individual instruments and vocals) of an entire band.
If this is the case, the cost of recording could go down significantly. Less equipment is necessary to record the multiple tracks, less time required to "get it right," which in turn means less time to bill for by the recording studio.
How do you think this technology will change the music industry?
If this is the case, the cost of recording could go down significantly. Less equipment is necessary to record the multiple tracks, less time required to "get it right," which in turn means less time to bill for by the recording studio.
How do you think this technology will change the music industry?
I'd be interested in how it represents a pitch bend or slide.
Would it be a sequence of small individual notes or one note that moves up/down the screen?
Would it be a sequence of small individual notes or one note that moves up/down the screen?
I can't decide if I should laugh or cry.
Why practice (music) anymore? But think of the possibilities...
Why practice (music) anymore? But think of the possibilities...
How does one account for the data lost via compression and interference?
I can't wait to play with this.
Seriously with stuff like this available to the average punter, could the age of celebrity finally be over? Here's hoping.