Stripping lyrics from mp3's making them instrumental

[Spooner]

do you guys know any programs that do this successfully?

I've tried AnalogX's, but it doesn't work.... AT ALL

 

[BD2003]

Its pretty much damn near impossible unless you have an EXACT acapella track of the voice to subtract from the original wave.

 

[Spooner]

Originally posted by: BD2003
Its pretty much damn near impossible unless you have an EXACT acapella track of the voice to subtract from the original wave.

there's gotta be something powerful enough to do it!

i refuse to belive it's "impossible"

 

[SWirth86]

Originally posted by: Spooner

Originally posted by: BD2003
Its pretty much damn near impossible unless you have an EXACT acapella track of the voice to subtract from the original wave.

there's gotta be something powerful enough to do it!

i refuse to belive it's "impossible"

I don't think it's a matter of power.....It's just that there's no way to distinguish the music from the singing.

Although, it would be wicked cool if it was possible

 

[Spooner]

then why would AnalogX advertise that it works great, then i download it... and it works like NOT AT ALL.

why bother developing something that doesn't work ?

ugh :disgust:

 

[xchangx]

The analogX program basically reverses the polarity on one channel. So, what ever is set to the center channel is taken out. The program works on songs where the vocals were all mixed in the center.

 

[LiekOMG]

Taking the singing out of an mp3? Don't think its possible. There is just no way to seperate it from the music.
However, taking the singing out of a multichannel music file would be very very easy.

 

[Spooner]

Originally posted by: DOACleric
However, taking the singing out of a multichannel music file would be very very easy.

what do you mean?

 

[Thegonagle]

It's not really possible to seperate the vocal track after a multi-track studio recording has been down-mixed to two channels for the CD. Kind of like how it's not possible to re-encode a 128K MP3 to 256K and get back the lost quality. Once the tracks have been mixed, they're staying mixed.

 

[NikPreviousAcct]

Originally posted by: Spooner

Originally posted by: DOACleric However, taking the singing out of a multichannel music file would be very very easy.

what do you mean?

Think of a multichannel music file to be a timeline running right to left on your desk in front of you. Think of it as your keyboard. On your keyboard, you've got lots of keys. True they're all corresponding to different data, but you can see that there are five obviousl "layers" of keys... the numbers, the line that starts with <tab> qwer, the line that stars with <caps lock> asdf, etc.

Think of each layer as different channels in the file. The <tab> line consists of the data that makes up the voices, the <caps> line consists of the bass, the <shift> line makes up the lead guitar, etc.

If you simply strip out all of the keys in the <tab> line, you no longer have that data available to complrise how the keyboard will function. You can't type any of the keys on that row.

Likewise, if you strip out the channel in the file that contains the voices, you're left with only music. But you need to have a program that's specifically designed to edit music - mroe specifically, multichannel files.

nik

 

[Gunbuster]

ever try getting milk out of coffee?

 

[Spooner]

this is how AnalogX says it "works":

AnalogX Vocal Remover works on the same principles that the hardware removers do - that in most instances vocals are equally mixed in both channels, and can identified and therefore removed by simply changing the phase on one channel by 180 degrees. While this won't remove vocals in all instances, it does work in many cases, and can sometimes be used to remove bass or breakbeat sections as well, which I find to be GREAT for sampling! Depending on the effects used on the vocals, sometimes the reverb or ambience of the vocals is left.

 

[LiekOMG]

Wav's, Mp3's, CD's, etc are all technically 'streamed' audio, meaning all the individual instruments and vocals that make up the file are simply recorded as a single track. When you buy a CD, its streamed music. However, when the musician actually makes the music and has it recorded in a studio, each instrument is set up on its own channel and recorded individually. This way if one instrument turns out too loud, or too soft in the final track, its easy to adjust the volume, or make any changes necessary with only affecting one instrument (or a voice). If the music was initially recorded as a stream, it would be impossible to fine tune individual aspects... just the same way it is kinda impossible to remove the voice in a song from a stream, because its already mixed together.

An example of a multichannel music file would be a MIDI. However a MIDI can't contain vocals. A module (MOD, S3M, XM, IT) can contain vocals, which can be removed if you wanted. None of these are streamed files, though

 

[stonecold3169]

with an MP3 it is technically impossible, as everything is jumbled together.

Now, depending on what kind of cd/dvd you have, and how it is encoded, it can be really easy as someone else has said, as it's typically as easy as removing one channel, leaving the rest perfectly intact. The way MP3 files are encoded, however, it would be impossible to get a "perfect" rip... maybe decent, but it would take one hell of an algorithm to do it...

 

[MercenaryForHire]

Damn, AnalogX sucks. Just kills the sound totally.

Ech.

- M4H

 

[NikPreviousAcct]

Originally posted by: Gunbuster
ever try getting milk out of coffee?

damn nice analogy