Forums

Fancy Music Processing Product

Started by Eric Jacobsen March 21, 2008
First, I'm in no way shape or form connected with these folks, I'm a
curious bystander and perhaps potential user:

http://www.celemony.com/cms/index.php?id=dna

The claim is that they can pull notes out of a chord in recorded music
and alter the notes individually,  e.g., change chord from a major to
a minor, change the key, etc.

I'm a bit skeptical as to how well it might work in general, and I'm
also not quite clear on how this is really applied.   It appears to be
a plug-in to sequencer software, which, I think, implies that it
samples and then resynthesizes?

In any case, if it does what is claimed it's quite impressive.

Eric Jacobsen
Minister of Algorithms
Abineau Communications
http://www.ericjacobsen.org
Eric Jacobsen <eric.jacobsen@ieee.org> writes:

> First, I'm in no way shape or form connected with these folks, I'm a > curious bystander and perhaps potential user: > > http://www.celemony.com/cms/index.php?id=dna > > The claim is that they can pull notes out of a chord in recorded music > and alter the notes individually, e.g., change chord from a major to > a minor, change the key, etc. > > I'm a bit skeptical as to how well it might work in general, and I'm > also not quite clear on how this is really applied. It appears to be > a plug-in to sequencer software, which, I think, implies that it > samples and then resynthesizes? > > In any case, if it does what is claimed it's quite impressive.
I saw this and the other post on a similar product a few days ago. My question is, why not just play it the way you want it to begin with? Oh, right! That assumes you know how to PLAY! ... -- % Randy Yates % "So now it's getting late, %% Fuquay-Varina, NC % and those who hesitate %%% 919-577-9882 % got no one..." %%%% <yates@ieee.org> % 'Waterfall', *Face The Music*, ELO http://www.digitalsignallabs.com
Could this be a licensed application of Empirical Mode Decomposition and 
the Hilbert-Huang Transform - as developed at NASA Goddard?

David

"Eric Jacobsen" <eric.jacobsen@ieee.org> wrote in message 
news:bt58u3lms1tl0hrs69vij6osqme399i3j4@4ax.com...
> First, I'm in no way shape or form connected with these folks, I'm a > curious bystander and perhaps potential user: > > http://www.celemony.com/cms/index.php?id=dna > > The claim is that they can pull notes out of a chord in recorded music > and alter the notes individually, e.g., change chord from a major to > a minor, change the key, etc. > > I'm a bit skeptical as to how well it might work in general, and I'm > also not quite clear on how this is really applied. It appears to be > a plug-in to sequencer software, which, I think, implies that it > samples and then resynthesizes? > > In any case, if it does what is claimed it's quite impressive. > > Eric Jacobsen > Minister of Algorithms > Abineau Communications > http://www.ericjacobsen.org
On Mar 21, 4:19 pm, Eric Jacobsen <eric.jacob...@ieee.org> wrote:
> First, I'm in no way shape or form connected with these folks, I'm a > curious bystander and perhaps potential user: > > http://www.celemony.com/cms/index.php?id=dna > > The claim is that they can pull notes out of a chord in recorded music > and alter the notes individually, e.g., change chord from a major to > a minor, change the key, etc.
well, changing the key by shifting *all* of the notes the same amount, that can be done and has been done, even to a broadband mix of instrument notes. but yanking a note out of a chord, operating on that, and putting it back it is very new. even separating a duet (Rob Maher long ago did a paper on that) is, or at least used to be, difficult and hazardous.
> I'm a bit skeptical as to how well it might work in general, and I'm > also not quite clear on how this is really applied. It appears to be > a plug-in to sequencer software, which, I think, implies that it > samples and then resynthesizes? > > In any case, if it does what is claimed it's quite impressive. >
i am too. and from what i've seen in Electronic Musician, it *does* do pretty much what is claimed. it's in different forms, and it's not so much sequencer software, but general audio editing software (like Pro Tools) that it works with. (also Cubase and whoever else runs VST, and also AU for the Mac.) looks like they be doing to the pitch-shifting and time-scaling market sorta what Google did to the landscape of search engines. a company i used to work for as well as former competitors (like Serato) might be a little concerned about these German guys. r b-j
On Mar 21, 7:20 pm, robert bristow-johnson <r...@audioimagination.com>
wrote:
> On Mar 21, 4:19 pm, Eric Jacobsen <eric.jacob...@ieee.org> wrote: > > > First, I'm in no way shape or form connected with these folks, I'm a > > curious bystander and perhaps potential user: > > >http://www.celemony.com/cms/index.php?id=dna > > > The claim is that they can pull notes out of a chord in recorded music > > and alter the notes individually, e.g., change chord from a major to > > a minor, change the key, etc. > > well, changing the key by shifting *all* of the notes the same amount, > that can be done and has been done, even to a broadband mix of > instrument notes. but yanking a note out of a chord, operating on > that, and putting it back it is very new. even separating a duet (Rob > Maher long ago did a paper on that) is, or at least used to be, > difficult and hazardous. > > > I'm a bit skeptical as to how well it might work in general, and I'm > > also not quite clear on how this is really applied. It appears to be > > a plug-in to sequencer software, which, I think, implies that it > > samples and then resynthesizes? > > > In any case, if it does what is claimed it's quite impressive. > > i am too. and from what i've seen in Electronic Musician, it *does* > do pretty much what is claimed. it's in different forms, and it's not > so much sequencer software, but general audio editing software (like > Pro Tools) that it works with. (also Cubase and whoever else runs > VST, and also AU for the Mac.) > > looks like they be doing to the pitch-shifting and time-scaling market > sorta what Google did to the landscape of search engines. a company i > used to work for as well as former competitors (like Serato) might be > a little concerned about these German guys. > > r b-j
There was a good video demo of this by Neubaker at Frankfurt MusikMesse showing it operating in real time. One of the caveats that he pointed out was that it works best on musical phrases that are played on one instrument - mixing various instruments and/or vocals apparently works but not optimally. That suggests their algorithm may be looking for similar harmonic structures and that dissimilar structures may confuse it. Eric
> well, changing the key by shifting *all* of the notes the same amount, > that can be done and has been done, even to a broadband mix of > instrument notes. but yanking a note out of a chord, operating on > that, and putting it back it is very new. even separating a duet (Rob > Maher long ago did a paper on that) is, or at least used to be, > difficult and hazardous. >> I'm a bit skeptical as to how well it might work in general, and I'm >> also not quite clear on how this is really applied. It appears to be >> a plug-in to sequencer software, which, I think, implies that it >> samples and then resynthesizes? >> In any case, if it does what is claimed it's quite impressive.
> i am too. and from what i've seen in Electronic Musician, it *does* > do pretty much what is claimed. it's in different forms, and it's not > so much sequencer software, but general audio editing software (like > Pro Tools) that it works with. (also Cubase and whoever else runs > VST, and also AU for the Mac.) > > looks like they be doing to the pitch-shifting and time-scaling market > sorta what Google did to the landscape of search engines. a company i > used to work for as well as former competitors (like Serato) might be > a little concerned about these German guys.
So... Their demo has a fairly fixed guitar pluck. A very definable string vibration. There are frequencies common to all chords, so they must remove humanly discernable pitch from a common timbre. The ear doesn't 'see' the masking tactics - mp3 is cheap and tiny (makes classical music sound shit too!) I'm an enthusiastic laymen (sorry), it would be good to here peoples views on comp.dsp about this. After all, the public buys into 'I can play anything, then re-arrange it' sadness. I personally think they are selling this on the basis you can load up a track then play the whole thing differently, the excitement generated from using a mouse(?) to completely change your favourite James Bond theme tune to play something completely weird is a very forceful marketing angle. (It also make me wish I could actually play an instument well) Is it really big and clever? Or is it the old divide everything by two and see what we've got trick! : )
On Mar 22, 9:07 am, Randy Yates <ya...@ieee.org> wrote:
> Eric Jacobsen <eric.jacob...@ieee.org> writes: > > First, I'm in no way shape or form connected with these folks, I'm a > > curious bystander and perhaps potential user: > > >http://www.celemony.com/cms/index.php?id=dna > > > The claim is that they can pull notes out of a chord in recorded music > > and alter the notes individually, e.g., change chord from a major to > > a minor, change the key, etc. > > > I'm a bit skeptical as to how well it might work in general, and I'm > > also not quite clear on how this is really applied. It appears to be > > a plug-in to sequencer software, which, I think, implies that it > > samples and then resynthesizes? > > > In any case, if it does what is claimed it's quite impressive. > > I saw this and the other post on a similar product a few days ago. My > question is, why not just play it the way you want it to begin with? Oh, > right! That assumes you know how to PLAY! ... > -- > % Randy Yates % "So now it's getting late, > %% Fuquay-Varina, NC % and those who hesitate > %%% 919-577-9882 % got no one..." > %%%% <ya...@ieee.org> % 'Waterfall', *Face The Music*, ELOhttp://www.digitalsignallabs.com
Well you could be a piano player for instance and want a guitar to accompany you. K.
Randy Yates wrote:
> Eric Jacobsen <eric.jacobsen@ieee.org> writes: > >> First, I'm in no way shape or form connected with these folks, I'm a >> curious bystander and perhaps potential user: >> >> http://www.celemony.com/cms/index.php?id=dna >> >> The claim is that they can pull notes out of a chord in recorded music >> and alter the notes individually, e.g., change chord from a major to >> a minor, change the key, etc. >> >> I'm a bit skeptical as to how well it might work in general, and I'm >> also not quite clear on how this is really applied. It appears to be >> a plug-in to sequencer software, which, I think, implies that it >> samples and then resynthesizes? >> >> In any case, if it does what is claimed it's quite impressive. > > I saw this and the other post on a similar product a few days ago. My > question is, why not just play it the way you want it to begin with? Oh, > right! That assumes you know how to PLAY! ...
Maybe you are the one musician who plays defect free, with the maximum of emotion, for hours and hours. For the rest, studio post processing has always included a huge amount of fixing up fluffed notes. This has generally been by fudging together different takes. If you can fix one fluffed note within a take, think how much studio time you could save. The demo fixing one fluffed note within a guitar chord is impressive, and the kind of fixup that would be useful in professional recording work. However, it remains to be seen how well this thing works with various types of music. The sample has pretty clean distinctive/percussive notes. Most music is considerably more slurred and overlapped than the demo piece. Steve
> Could this be a licensed application of Empirical Mode Decomposition and > the Hilbert-Huang Transform - as developed at NASA Goddard?
Hey that's interesting, here's another method by Mitsubishi:- "Separation of Mixed Audio Sources by Independent Subspace Analysis" http://www.merl.com/papers/docs/TR2001-31.pdf
> David > > "Eric Jacobsen" <eric.jacobsen@ieee.org> wrote in message > news:bt58u3lms1tl0hrs69vij6osqme399i3j4@4ax.com... >> First, I'm in no way shape or form connected with these folks, I'm a >> curious bystander and perhaps potential user: >> >> http://www.celemony.com/cms/index.php?id=dna >> >> The claim is that they can pull notes out of a chord in recorded music >> and alter the notes individually, e.g., change chord from a major to >> a minor, change the key, etc. >> >> I'm a bit skeptical as to how well it might work in general, and I'm >> also not quite clear on how this is really applied. It appears to be >> a plug-in to sequencer software, which, I think, implies that it >> samples and then resynthesizes? >> >> In any case, if it does what is claimed it's quite impressive. >> >> Eric Jacobsen >> Minister of Algorithms >> Abineau Communications >> http://www.ericjacobsen.org >
Steve Underwood <steveu@dis.org> writes:
> [...] > Maybe you are the one musician who plays defect free, with the maximum > of emotion, for hours and hours. For the rest, [...]
Smart-assedness noted. No, Steve, I didn't mean to imply that it wouldn't be useful there. However, that's not really the application I saw being pushed, at least not in the previous youtube video of the german and his invention. -- % Randy Yates % "She tells me that she likes me very much, %% Fuquay-Varina, NC % but when I try to touch, she makes it %%% 919-577-9882 % all too clear." %%%% <yates@ieee.org> % 'Yours Truly, 2095', *Time*, ELO http://www.digitalsignallabs.com