Is it possible to OBTAIN ONLY THE VOICE from a song in mp3?, Get only the voice from a song in mp3, remove the audio, get the voice Options

[mrmister]

Hello
Sorry if I am so repetitive and I explain it so much, but there is a tremendous confusion when people is asking this and they don't understand well the question.
Some people want to remove the voice from a song and get only the audio. WELL, I DON'T WANT TO DO THAT.
I want to do exactly the contrary.
I have a song in mp3 with its vocals.
I want to delete the audio, but I want to get only the vocals.
How may I do this?
Is there any program in which I can do this?
I have audacity and sound forge so far.
Thanks

[DVDdoug]

There are a couple of programs that attempt to do that.

There is a FREE plug-in for Winamp (FREE) called DSP Centercut.

GoldWave ($50 after free trial) has some Stereo Center Channel processing tools.

And, there are a couple more. But, you should be aware that just like the "old vocal removal trick" (which works by subtracting left from right) these can be fun to play with, but it's impossible to "un-mix" a song and these things never work perfectly.

This post has been edited by DVDdoug: Jul 6 2012, 05:48

[mrmister]

There are a couple of programs that attempt to do that.

There is a FREE plug-in for Winamp (FREE) called DSP Centercut.

GoldWave ($50 after free trial) has some Stereo Center Channel processing tools.

And, there are a couple more. But, you should be aware that just like the "old vocal removal trick" (which works by subtracting left from right) these can be fun to play with, but it's impossible to "un-mix" a song and these things never work perfectly.

I have the voice from an old TV commercial, what I need is to extract the voice from the music, to get that voice, and put new music on the background.
I give you the youtube link, the TV commercial is this:
(it's in Italian I'm doing an homage to Lancia Fulvia an Italian car)

http://www.youtube.com/watch?v=-ZbgFzvMJgw

So I want to get the voice and remove the music in the background.
What do you think would be better for this?
GoldWave?

[db1989]

There is a FREE plug-in for Winamp (FREE) called DSP Centercut.

Which was ported to foobar2000, and I can recommend that version.

What do you think would be better for this?

I think you’ll have to just try the different options for yourself. We can’t make reliable predictions.

[mrmister]

There is a FREE plug-in for Winamp (FREE) called DSP Centercut.

Which was ported to foobar2000, and I can recommend that version.

What do you think would be better for this?

I think you’ll have to just try the different options for yourself. We can’t make reliable predictions.

Is there any tutorial showing how to do this?
GoldWave is an audio editor... I have also Audacity and Sound Forge... could I do the same (removing the music and getting just the vocals) with Audacity?
If could you give me any clue about how to do this, I'd really appreciate it

[uart]

I think it will be impossible. You might be able to emphasize the vocal spectrally, but that's about all .

[db1989]

Is there any tutorial showing how to do this?

Install foobar2000 or Winamp; install plugin;activate plugin and its mode to isolate the centre a.k.a. cut the sides; play track; judge result; convert to new file if happy.

My implication here is that the algorithm used in the plugin is more advanced than those in most audio editors, especially Audacity, which has not been known for the stellar quality of its effects as far as I am aware.

[2Bdecided]

Yes, the algorithm in fb2k is quite advanced. You can usually hear artefacts from the algorithm though.

I suspect the stereo sound stage in an mp3 (never mind a YouTube video, which has been transcoded from whatever the user uploaded to whatever format you chose to download/capture) is going to make the artefacts far worse.

Good luck!

Cheers,
David.

[uart]

I think it will be impossible. You might be able to emphasize the vocal spectrally, but that's about all .

Sorry forgot to add: I took a look at the audio on that clip and I'm pretty sure it's mono.

[mrmister]

I think it will be impossible. You might be able to emphasize the vocal spectrally, but that's about all .

Sorry forgot to add: I took a look at the audio on that clip and I'm pretty sure it's mono.

Thanks everyone here for being so kind and helping me with this
for sure I'd need a bit more help, when I use it, but I'll install it and tell you the results or the doubts I have
Of course it is mono !!! is from the 60s
I'm doing a kind of video homage to a retro-car Lancia Fulvia... I'll show you when it be finished

[DonP]

Of course it is mono !!! is from the 60s
I'm doing a kind of video homage to a retro-car Lancia Fulvia... I'll show you when it be finished

The reason this is important is that the major trick in splitting off vocals depends on the vocal being center stage (equal in both channels) while the music has more separation. That won't work at all if the whole recording is mono.

[uart]

Yep MrMister, that was my point about being mono, it's not going to work. Perhaps we could make a new voice-over for it. I can't speak Italian so I'm just going to have to guess. Maybe something like,

This man is going for a drive. He's not driving a rally, he's just going out to get some milk. This is how Italians drive all the time.

Sorry, just a sick joke. ;p

[benski]

Just a quick FYI to summarize some things mentioned on this thread.
There is no way to extract or remove the voice in complete isolation. The audio has been mixed and cannot be unmixed. You would need access to the original multi-track studio tapes (which often don't even exist for older material, especially if the song was recorded in one take)

There are, however, a number of "tricks" that work reasonably well most of the time based on the fact that the vocals are usually panned center (but so is the kick drum, among other things ...). If the track is in mono, then this trick won't work. It also doesn't work for older stereo songs where some of the instruments are in the left and some are in the right (e.g. Credence Clearwater Revival's "Susie Q")