21 Feb
2008
21 Feb
'08
9:28 a.m.
Hi,
Is there a linux tool which can analyze a song? I want to listen only
the voice, drums or guitarpart....
regards,
Dirk
21 Feb
21 Feb
10:15 a.m.
hi.
thats simply not possible, since the frequencies of these parts are
overlapping. You can try to fade out certain areas, e.g. the bass part
(kicks, bass) by using an eq or when the voice is only a plain
monosignal while the rest is stereo you could delete the stereo-center
to get some kind of a karaoke effect. But a tool that analyses music to
simply switch off certain instruments does not exist. Also not for
windows or mac.
--
Best regards,
Sebastian.
Web: linuxaudioblog.sternenhejim.de
On Thu, 2008-02-21 at 17:27 +0100, schoappied wrote:
Hi,
Is there a linux tool which can analyze a song? I want to listen only
the voice, drums or guitarpart....
regards,
Dirk
_______________________________________________
Linux-audio-user mailing list
Linux-audio-user(a)lists.linuxaudio.org
http://lists.linuxaudio.org/mailman/listinfo/linux-audio-user
22 Feb
22 Feb
7:48 p.m.
On 21/02/2008, Sebastian Tschöpel <tschoseb(a)tu-cottbus.de> wrote:
hi.
thats simply not possible, since the frequencies of these parts are
overlapping. You can try to fade out certain areas, e.g. the bass part
(kicks, bass) by using an eq or when the voice is only a plain
monosignal while the rest is stereo you could delete the stereo-center
to get some kind of a karaoke effect. But a tool that analyses music to
simply switch off certain instruments does not exist. Also not for
windows or mac.
While I fully agree with Sebastian, I can't resist asking :
if it is not possible, then how do your ears and your brain do it, then ?
A musician can be trained to mentally separate the parts of an orchestra;
why could not a machine do the same ?
(I'm not implying it would be trivial, though ;-)
Regards.
--
Best regards,
Sebastian.
Web: linuxaudioblog.sternenhejim.de
On Thu, 2008-02-21 at 17:27 +0100, schoappied wrote:
Hi,
Is there a linux tool which can analyze a song? I want to listen only
the voice, drums or guitarpart....
regards,
Dirk
_______________________________________________
Linux-audio-user mailing list
Linux-audio-user(a)lists.linuxaudio.org
http://lists.linuxaudio.org/mailman/listinfo/linux-audio-user
_______________________________________________
Linux-audio-user mailing list
Linux-audio-user(a)lists.linuxaudio.org
http://lists.linuxaudio.org/mailman/listinfo/linux-audio-user
23 Feb
23 Feb
4:31 a.m.
On 23 Feb 2008, at 10:48 AM, Jean-Baptiste Mestelan wrote:
A musician can be trained to mentally separate the
parts of an
orchestra; why could not a machine do the same ?
(I'm not implying it would be trivial, though ;-)
it's a big leap from machine to musician!
It is hard enough to get a machine to be a good instrument for a
musician to play, let alone to analyse a sound as a musician would.
Still, what a machine does now would have been sci-fi to most people
just 20 years ago.
simon
1:40 p.m.
On Saturday 23 February 2008 01:48:38 Jean-Baptiste Mestelan wrote:
On 21/02/2008, Sebastian Tschöpel
<tschoseb(a)tu-cottbus.de> wrote:
Why not give thanks to the Creator for what He has given us. Music and our
sensory repertoire relating to it are indeed miracles.
When all our robots have positronic brains, maybe they could be so trained :-)
hi.
thats simply not possible, since the frequencies of these parts are
overlapping. You can try to fade out certain areas, e.g. the bass part
(kicks, bass) by using an eq or when the voice is only a plain
monosignal while the rest is stereo you could delete the stereo-center
to get some kind of a karaoke effect. But a tool that analyses music to
simply switch off certain instruments does not exist. Also not for
windows or mac.
While I fully agree with Sebastian, I can't resist asking :
if it is not possible, then how do your ears and your brain do it, then ?
A musician can be trained to mentally separate the parts of an orchestra;
why could not a machine do the same ?
(I'm not implying it would be trivial, though ;-)
25 Feb
25 Feb
4:48 p.m.
On Sat, Feb 23, 2008 at 12:48:38AM +0100, Jean-Baptiste Mestelan wrote:
if it is not possible, then how do your ears and your
brain do it, then ?
They can't. They just give you the illusion. What you hear is
not the original sound but a mental reconstruction of it.
This is why you can dream sounds, or why a trained musician
can 'hear' a melody when he sees it on a score.
The reconstruction aspect of human perception is probably
at the core of 'consciousness'.
Ciao,
--
FA
Laboratorio di Acustica ed Elettroacustica
Parma, Italia
Lascia la spina, cogli la rosa.
21 Feb
21 Feb
10:31 a.m.
On Thursday 21 February 2008 18:27:58 schoappied wrote:
Hi,
Is there a linux tool which can analyze a song? I want to listen only
the voice, drums or guitarpart....
Probably not possible, but Eleveyta has a vst plugin. Designed for voice
elimination and such but has a two dimensional GUI. One can "see" what is
playing at any time, select this "region" and either try to extract or mute
it. I have not tried it.
His plugins are very original. His clone-boy enabled me to use a spare vocal
track to "vocode" choirs so well that I can put them up front in a mix. So
give it a try, let me know how it works out.
Cloneensemble has a vocal extraction/elimination plugin (DX and VST) that
offers a "capestral" algorithm. Such is suited for voice recognition and
such--I read the papers he sited but cannot claim to understand them fully.
Designed for voice in the center so not as likely to do your job as the
Eleveyta plugin.
12:11 p.m.
On 21/02/2008, David Baron <d_baron(a)012.net.il> wrote:
Probably not possible, but Eleveyta has a vst plugin.
Designed for voice
elimination and such but has a two dimensional GUI. One can "see" what is
playing at any time, select this "region" and either try to extract or mute
it. I have not tried it.
It is possible to extract sources from a stereo mix that was created
using artificial panning in the studio to position individually
recorded sources in a stereo field. This paper describes one very
effective technique:
http://www.audioresearchgroup.com/papers/18.pdf
I'm not aware of any Linux software to do it though (and I haven't
tried the plugin you describe either).
Chris
11:58 a.m.
On 21 February 2008 at 17:27, schoappied <schoappied(a)gmail.com> wrote:
Is there a linux tool which can analyze a song? I want
to listen only
the voice, drums or guitarpart....
I don't think this will do all you want, but SonicVisualizer could
help you see better into the song structure.
G'luck....
--
Kevin
5914
days inactive
5918
days old
linux-audio-user@lists.linuxaudio.org
8 comments
8 participants
participants (8)
-
Chris Cannam -
David Baron -
Fons Adriaensen -
Jean-Baptiste Mestelan -
Kevin Cosgrove -
schoappied -
Sebastian Tschöpel -
simon wise