Jump to content
  • Advertisement
Sign in to follow this  
Hanchord5

Vocals Rmove

This topic is 4304 days old which is more than the 365 day threshold we allow for new replies. Please post a new topic.

If you intended to correct an error in the post then please contact us.

Recommended Posts

Anyone knows how to remove (most of the) music from a song and keep only the vocals? Removing vocals is easy, this could be harder, right Thanks..........

Share this post


Link to post
Share on other sites
Advertisement
The thing about removing vocals from music is that the vocals are usually panned (positioned) to a specific area (either wide stereo or dead-center), which makes it possible for the vocals to be isolated and removed. The music is a different case, as it usually consists of multiple instruments panned to different areas of the sound field, often times with the panning shifting over time. You could try to EQ the music out, perhaps using high and low-pass filters - but I doub't you'll ever achieve satisfactory results. Perhaps you should try to just find the acapella for the track.

Share this post


Link to post
Share on other sites
Assume S = song, M = music and V = vocal. S is known, and you also know that S = M + V.

You can easily (what you said) calculate M, so you can just as easily calculate V = S - M.

Share this post


Link to post
Share on other sites
Guest Anonymous Poster
bob is right, when you can remove the voice, you can also remove the music.
and vice versa. every telephone uses this, so that you dont hear yourself speaking, but always can hear the one you're calling.

Share this post


Link to post
Share on other sites
Actually, subtracting vocal from total doesn't sound too good. In fact, it almost always sounds awful, producing devastating audio artifacts unless there is very little frequency collision, which is very unlikely considering that the human voice spans most of the audible frequency domain.

I assume that to remove the vocals, you are using a channel mixer to annihilate everything that is panned to centre. This works fine on a well-mixed track where the only centred tracks are vocal and bass (the bass can be reintroduced with a side-channel EQ) but is otherwise unreliable, particularly when the sounds has been heavily post-processed.
If a track has been intensively mastered, restoring the vocals is more or less impossible. Reverb smudges all the frequency boundaries, compression & limiting prevents you separating waves by Fourier methods (which is dodgy in the first place) and harmonic excitation makes everything look the same, from the perspective of a spectrum analysis bot.

I hate to rain on your parade, but unless you are working under exeptional circumstances, you should probably forget about it. The same message is echoed on mashup/bootleg/restoration forums all over the internet.

Regards
Admiral

Share this post


Link to post
Share on other sites
Quote:
Original post by TheAdmiral
Actually, subtracting vocal from total doesn't sound too good. In fact, it almost always sounds awful, producing devastating audio artifacts unless there is very little frequency collision, which is very unlikely considering that the human voice spans most of the audible frequency domain.

I assume that to remove the vocals, you are using a channel mixer to annihilate everything that is panned to centre. This works fine on a well-mixed track where the only centred tracks are vocal and bass (the bass can be reintroduced with a side-channel EQ) but is otherwise unreliable, particularly when the sounds has been heavily post-processed.
If a track has been intensively mastered, restoring the vocals is more or less impossible. Reverb smudges all the frequency boundaries, compression & limiting prevents you separating waves by Fourier methods (which is dodgy in the first place) and harmonic excitation makes everything look the same, from the perspective of a spectrum analysis bot.

I hate to rain on your parade, but unless you are working under exeptional circumstances, you should probably forget about it. The same message is echoed on mashup/bootleg/restoration forums all over the internet.

Regards
Admiral

Subtracting works great assuming you have a good estimate of the music. Since he said he could easily do it, I assume he does have a good estimate of the music.

The subtraction method fails only if you have a bad estimate of what you subtract.

Share this post


Link to post
Share on other sites
I suggest Fleximusic Wave Editor using Band Pass Stop Filter in that you can either do blocks or keeps the frequency range. But you have to find by trial and error in which frequency bands to pass/stop.

Share this post


Link to post
Share on other sites
Quote:
Original post by Brother Bob
Subtracting works great assuming you have a good estimate of the music. Since he said he could easily do it, I assume he does have a good estimate of the music.


Sure it works great if you have a good source, such as the mixed vocal track, isolated. However, I got the impression we were starting from a commercial track (from a CD or similar). My point was that in many cases you can extract what sounds like a good vocal track through simple channel mixing. But as good as it sounds, a lot is left behind, and attempts to subtract this waveform from the total doesn't produce nearly such great results.

Regards
Admiral

Share this post


Link to post
Share on other sites
Sign in to follow this  

  • Advertisement
×

Important Information

By using GameDev.net, you agree to our community Guidelines, Terms of Use, and Privacy Policy.

Participate in the game development conversation and more when you create an account on GameDev.net!

Sign me up!