Fallin' In Love (Lady) question/discovery

[GuyOnTheBeach]

In my excessive amount of spare time I was messing about with this track in Audacity and noticed that if you cancel out the middle channel (by inverting the left and mixing it against the right) you get a fairly clean instrumental, which only really removes the vocals, one of the guitars and a little bit of the drum machine, however it is in mono. I tried to get a little more technical and used a (very old, primitive) plugin which would let me keep the results in stereo, however the result is pretty poor with a lot of artifact. It is also interesting to note that if you keep the middle channel and just remove the L+R channels after the intro you get what is pretty much a basic track (drum machine, guitar, bass and vocals), however again with this plugin the results are phasey and full of artifacts.

I was wondering if anyone on here had better software for such things and had ever tried this?, if so have better results been achieved?

[seltaeb1012002]

You could import the original into Izotope RX2 and manually scrub out the vocal. Takes forever, but the results are way more satisfying than the typical Center Channel Extraction.

[GuyOnTheBeach]

You could import the original into Izotope RX2 and manually scrub out the vocal. Takes forever, but the results are way more satisfying than the typical Center Channel Extraction.

That's an idea actually, I've used it a couple of times and have been amazed by the results (I've only done it for simple things, such as seperating the "instruments" on the original Dr Who theme, the two guitars on the intro for California Dreamin' and isolating Brian's vocals on the Oh Lord demo).
I will give this a go tomorrow though!

[Stephen W. Desper]

In my excessive amount of spare time I was messing about with this track in Audacity and noticed that if you cancel out the middle channel (by inverting the left and mixing it against the right) you get a fairly clean instrumental, which only really removes the vocals, one of the guitars and a little bit of the drum machine, however it is in mono. I tried to get a little more technical and used a (very old, primitive) plugin which would let me keep the results in stereo, however the result is pretty poor with a lot of artifact. It is also interesting to note that if you keep the middle channel and just remove the L+R channels after the intro you get what is pretty much a basic track (drum machine, guitar, bass and vocals), however again with this plugin the results are phasey and full of artifacts.

I was wondering if anyone on here had better software for such things and had ever tried this?, if so have better results been achieved?

COMMENT:  Of Course subtracting the left from the right channels of a stereo signal will remove anything that is in the virtual center. Usually this will be lead vocal, perhaps lead guitar, and bass guitar and kick drum. This is a fairly standard outcome with most songs, with the exception of some of mine where effects processing will not remove center information.

I am offering you this patching suggestion. (1) From the left channel send a duplicate of that signal to Aux 1. (2) From the right channel send a duplicate of that signal to Aux 2. (3) Send Aux 1 into an equalizer (outboard or on the console) (4) Send Aux 2 into an equalizer. (5) Now combine or rather subtract EQed Aux 1 from EQed Aux 2 -- with all EQs zero-ed. (6) Send this "mono" signal to a third EQ. We'll call it L-R.EQ. (7) Mix back into the original signal to remove the vocals.

Play around with the EQ settings for Aux 1 and Aux 2. You will find you can remove various elements. If you use a steep cutof for the third EQ (L-R.EQ), or use a high-pass filter in place of the EQ, set so that only the upper frequencies pass through, you will be able to remove the center vocals, but allow the bass and kick to pass on through.

Otherwise one of the oldest vocal removing devices is the Thompson Vocal Eliminator. When I was a kid this thing was sold (as an analog device) in the back of Popular Electronics Magazine -- "to make you sound like a star!" Today's digital version is quite good. At the website, check out all the DEMOs by placing the mouse over each topic.

Thompson Vocal Eliminator VE-4 >>> http://www.ltsound.com/eliminator_ppc.htm

Don't know anything about the on-line download version of a vocal Eliminator, but here is one.

Instant Karaoke Maker >>>  http://www.make-your-own-karaoke.com/?gclid=COLFuNXC4q0CFQoj7Aod7Xd53g

Most of what these devices do can be had using the patching techniques 1-7 above.

Have fun and good listening,

~Stephen W. Desper

[GuyOnTheBeach]

In my excessive amount of spare time I was messing about with this track in Audacity and noticed that if you cancel out the middle channel (by inverting the left and mixing it against the right) you get a fairly clean instrumental, which only really removes the vocals, one of the guitars and a little bit of the drum machine, however it is in mono. I tried to get a little more technical and used a (very old, primitive) plugin which would let me keep the results in stereo, however the result is pretty poor with a lot of artifact. It is also interesting to note that if you keep the middle channel and just remove the L+R channels after the intro you get what is pretty much a basic track (drum machine, guitar, bass and vocals), however again with this plugin the results are phasey and full of artifacts.

I was wondering if anyone on here had better software for such things and had ever tried this?, if so have better results been achieved?

COMMENT:  Of Course subtracting the left from the right channels of a stereo signal will remove anything that is in the virtual center. Usually this will be lead vocal, perhaps lead guitar, and bass guitar and kick drum. This is a fairly standard outcome with most songs, with the exception of some of mine where effects processing will not remove center information.

I am offering you this patching suggestion. (1) From the left channel send a duplicate of that signal to Aux 1. (2) From the right channel send a duplicate of that signal to Aux 2. (3) Send Aux 1 into an equalizer (outboard or on the console) (4) Send Aux 2 into an equalizer. (5) Now combine or rather subtract EQed Aux 1 from EQed Aux 2 -- with all EQs zero-ed. (6) Send this "mono" signal to a third EQ. We'll call it L-R.EQ. (7) Mix back into the original signal to remove the vocals.

Play around with the EQ settings for Aux 1 and Aux 2. You will find you can remove various elements. If you use a steep cutof for the third EQ (L-R.EQ), or use a high-pass filter in place of the EQ, set so that only the upper frequencies pass through, you will be able to remove the center vocals, but allow the bass and kick to pass on through.

Otherwise one of the oldest vocal removing devices is the Thompson Vocal Eliminator. When I was a kid this thing was sold (as an analog device) in the back of Popular Electronics Magazine -- "to make you sound like a star!" Today's digital version is quite good. At the website, check out all the DEMOs by placing the mouse over each topic.

Thompson Vocal Eliminator VE-4 >>> http://www.ltsound.com/eliminator_ppc.htm

Don't know anything about the on-line download version of a vocal Eliminator, but here is one.

Instant Karaoke Maker >>>  http://www.make-your-own-karaoke.com/?gclid=COLFuNXC4q0CFQoj7Aod7Xd53g

Most of what these devices do can be had using the patching techniques 1-7 above.

Have fun and good listening,

~Stephen W. Desper

Thank you for the tips, Stephen. I look forwards to trying them out, it will be interesting to see what results I get!