This is just mental.
The video doesn’t seem to be loading for me, but I can get a vague idea from the description.
Does this program do anything to reverse engineer the audio if it can distinguish individual tracks and notes? I once downloaded an experimental program that turned basic, non-layered audio samples into MIDI files in an attempt to import them to Finale to cheat my way out of transposing piano sheet music. It didn’t work because the drum beat threw off the whole system, but I’ve always been fascinated with the idea of algorithms breaking down audio components after they’re mixed.
Am I reading too much into this? I’ve got the .WMV downloading at an abysmal rate - that seems to be on my end right now - so I’ll probably be able to answer the question in an hour. It seems like the program would almost have to break it down into individual tracks to retune only individual instruments.
Part of me is looking at this and thinking, “This is so. Fucking. Cool.” I can see how in some of my work I could’ve really used this well. The bit where they alter the chords using a MIDI controller is phenomenal. It’s not just a tech demo; this sounds really, really cool. Insanely cool tech.
The rest of me is thinking, “This is an absolute abomination.”
GyRo567 – it’s exactly what it says it is; it can take a recording of a chord (or multiple notes otherwise played at the same time) and from that extract the individual notes into samples, which you can then alter each note individually the same as any individual note sample. Absolutely insane, as Sebmojo says.
The guy who came up with this is going to be rich, because every studio in the world is going to want this tech.
MIDI polyphony separation has been around for a while, as it turns out. What this does, however, is break down a waveform, which was previously thought to be impossible, or at least so resource-intensive that it was beyond the processing power of even pro workstations.
With this, a chord recorded as a wav. file can be separated into its individual notes. It doesn’t work exactly like that, but the approximation is close enough to the real thing as to make little difference to the human ear. And you can not only shift the pitch, but the timing and shape of each note. And, with a sufficiently beefy computer, it can be done pretty much in real time, as though the pieces of the track were separate loops being laid down in a sequencer.
Given that MIDI gives you each individual note separately to begin with, this is somewhat easy to believe. ;)
That is badass! They can breakdown the chord to individual notes and change EACH pitch and then throw it back together. Wow.
I’m curious to know how well it’ll actually work once it’s out on the market – five bucks says it won’t work anywhere near as flawlessly as their demo video suggests :)
If it does work, this software is completely fucking revolutionary. You could be a pretty crappy musician but have a good musical sense, and take what you’ve recorded, and fix it up in post-production into something worth listening to.
A lot of the CDDJ models have pitch shifted timing correction. It usually works via 1 second samples being stretched and interpolated, but I go watch video now.
ok, the FM parsing is cool.
Oh, pfft. This is old news. :)
If some company could build this into some kind of effects stomp box, sort of like the PitchRider box singers use…BILLIONS OF $ TO BE MADE!
I’ve used Regular Melodyne with Sonar, so I don’t doubt that this will work when they release it.
What happens when Melodyne becomes self-aware?
Very cool technology.
Ummmm, pro-tools does alot of this already, and makes your favorite crappy singers sound good…even live.
Nothing has been able to go into a chord and extract notes, or go into a mix and extract elements. Melodyne is light years beyond what anybody else is doing in this field.
Ummmm…no. This is nothing like pro-tools.
Wow. That’s amazing.
It’ll be interesting to see how soon this gets used as a guitar cheat to make chords that aren’t physically possible without alternate tunings, tunings that the guitar player in question may not be able to play in reliably.
So what I’m saying is… sign me up!
The Close Encounters of the Third Kind theme becomes our national anthem?
It will also be interesting for people like me who only play piano and keyboards. Up until now, I’ve been basically playing the fake string sounds (or whatever else works - older keyboards have some really strange sounds that sound great with effects) through guitar amplifiers because I can’t play stringed instruments. If I can use a real instrument to get a few reference recordings, it seems like I can use my keyboard to get more life-like guitar sounds.
It still won’t be perfect analog quality of course, but it seems like I’ll have a lot more capability than I currently do. Either way, I’m still geeking out in anticipation. This is definitely something I’ll buy if it turns out half as good as they advertise.