Q: Using phase inversion to extract an instrumental when you have the acapella

Have an urge to learn, or a calling to teach? Want to share some useful Youtube videos? Do it here!
Post Reply
User avatar
EnochLight
Moderator
Posts: 8405
Joined: 17 Jan 2015
Location: Imladris

03 Jun 2018

I know it's popular to use phase inversion/center channel cancelling to extract acapellas from a mixdown, but what if you already have the acapella as well as the final mixdown, but want to extract an instrumental?

Has anyone done this successfully using just Reason?

I found a service called PhonicMind, which I'm pretty sure uses some sort of phase cancellation to split instrumentals and acapellas. There must be a way to do this in Reason easily...

Anyone?

Here's what PhonicMind does (upload any finished song with vocals and see/hear the magic): https://phonicmind.com/
Win 10 | Ableton Live 11 Suite |  Reason 12 | i7 3770k @ 3.5 Ghz | 16 GB RAM | RME Babyface Pro | Akai MPC Live 2 & Akai Force | Roland System 8, MX1, TB3 | Dreadbox Typhon | Korg Minilogue XD

User avatar
selig
RE Developer
Posts: 11685
Joined: 15 Jan 2015
Location: The NorthWoods, CT, USA

03 Jun 2018

EnochLight wrote:I know it's popular to use phase inversion/center channel cancelling to extract acapellas from a mixdown, but what if you already have the acapella as well as the final mixdown, but want to extract an instrumental?

Has anyone done this successfully using just Reason?

I found a service called PhonicMind, which I'm pretty sure uses some sort of phase cancellation to split instrumentals and acapellas. There must be a way to do this in Reason easily...

Anyone?

Here's what PhonicMind does (upload any finished song with vocals and see/hear the magic): https://phonicmind.com/
Vocal removers have been around for decades.

Vocal extractors are only possible if you have the track already, and you subtract it from the track+vocal version.

Vocal extractors typically split the mix with a crossover so that bass and drums aren’t affected so much, then they assume vocals are in the center only and with little or no stereo reverb/fx.

It’s an inexact science to be sure, and until something along the lines of Melodyne level tech is introduced it will likely remain that way.



Sent from some crappy device using Tapatalk
Selig Audio, LLC

User avatar
EnochLight
Moderator
Posts: 8405
Joined: 17 Jan 2015
Location: Imladris

03 Jun 2018

selig wrote:
03 Jun 2018
Vocal removers have been around for decades.

Vocal extractors are only possible if you have the track already, and you subtract it from the track+vocal version.

Vocal extractors typically split the mix with a crossover so that bass and drums aren’t affected so much, then they assume vocals are in the center only and with little or no stereo reverb/fx.

It’s an inexact science to be sure, and until something along the lines of Melodyne level tech is introduced it will likely remain that way.



Sent from some crappy device using Tapatalk
Right, so - I currently have:
  • Complete track mixdown (vocals + instruments)
  • Vocal acapella (just vocal track)
I want:
  • Instruments
There's no way to invert the (vocals + instruments) mixdown against the (just vocal track) acapella to end up with just the instruments? If not, then it looks like I'm resigned to using that online service. :S
Win 10 | Ableton Live 11 Suite |  Reason 12 | i7 3770k @ 3.5 Ghz | 16 GB RAM | RME Babyface Pro | Akai MPC Live 2 & Akai Force | Roland System 8, MX1, TB3 | Dreadbox Typhon | Korg Minilogue XD

User avatar
selig
RE Developer
Posts: 11685
Joined: 15 Jan 2015
Location: The NorthWoods, CT, USA

03 Jun 2018

EnochLight wrote:
03 Jun 2018
selig wrote:
03 Jun 2018
Vocal removers have been around for decades.

Vocal extractors are only possible if you have the track already, and you subtract it from the track+vocal version.

Vocal extractors typically split the mix with a crossover so that bass and drums aren’t affected so much, then they assume vocals are in the center only and with little or no stereo reverb/fx.

It’s an inexact science to be sure, and until something along the lines of Melodyne level tech is introduced it will likely remain that way.



Sent from some crappy device using Tapatalk
Right, so - I currently have:
  • Complete track mixdown (vocals + instruments)
  • Vocal acapella (just vocal track)
I want:
  • Instruments
There's no way to invert the (vocals + instruments) mixdown against the (just vocal track) acapella to end up with just the instruments? If not, then it looks like I'm resigned to using that online service. :S
Have you tried to invert one of the two tracks in Reason? Not sure what you mean by "there's no way to invert…" - sure there is, using a Mix Channel (sorry if I'm misunderstanding you).
Selig Audio, LLC

User avatar
selig
RE Developer
Posts: 11685
Joined: 15 Jan 2015
Location: The NorthWoods, CT, USA

03 Jun 2018

EnochLight wrote:
03 Jun 2018
selig wrote:
03 Jun 2018
Vocal removers have been around for decades.

Vocal extractors are only possible if you have the track already, and you subtract it from the track+vocal version.

Vocal extractors typically split the mix with a crossover so that bass and drums aren’t affected so much, then they assume vocals are in the center only and with little or no stereo reverb/fx.

It’s an inexact science to be sure, and until something along the lines of Melodyne level tech is introduced it will likely remain that way.



Sent from some crappy device using Tapatalk
Right, so - I currently have:
  • Complete track mixdown (vocals + instruments)
  • Vocal acapella (just vocal track)
I want:
  • Instruments
There's no way to invert the (vocals + instruments) mixdown against the (just vocal track) acapella to end up with just the instruments? If not, then it looks like I'm resigned to using that online service. :S
One more thing - there are many reasons why even this simple inversion trick may not work.

First thing is random FX like reverb being different from the full track to the vocal.
Second thing is any master inserts using dynamics, since they would react differently to the entire mix compared to just the vocals.
Third thing would be if levels were adjusted on either one after the export (which is the one thing you COULD compensate for when doing the invert trick).

Finally, you can try to vocal eliminator trick I mentioned previously, possibly combining it with the inversion trick OR using the vocal to side-chain the inversion trick.

How important is this? How much $$ is it worth and how "perfect" do you require the results?

(sorry, I was confused by your original post because you mentioned both the online option which uses just a finished mix, AND the fact you had both a finished mix and the vocal stem - missed the last part with my initial post)
Selig Audio, LLC

User avatar
EnochLight
Moderator
Posts: 8405
Joined: 17 Jan 2015
Location: Imladris

03 Jun 2018

OK, thanks. It's not really important - was just trying to get it to work well in Reason. The online service is of course an option - it's not expensive for just 1 track.

Thanks!
Win 10 | Ableton Live 11 Suite |  Reason 12 | i7 3770k @ 3.5 Ghz | 16 GB RAM | RME Babyface Pro | Akai MPC Live 2 & Akai Force | Roland System 8, MX1, TB3 | Dreadbox Typhon | Korg Minilogue XD

User avatar
selig
RE Developer
Posts: 11685
Joined: 15 Jan 2015
Location: The NorthWoods, CT, USA

03 Jun 2018

EnochLight wrote:
03 Jun 2018
OK, thanks. It's not really important - was just trying to get it to work well in Reason. The online service is of course an option - it's not expensive for just 1 track.

Thanks!
So what happens when you do the inversion trick?
Selig Audio, LLC

User avatar
miscend
Posts: 1955
Joined: 09 Feb 2015

03 Jun 2018

There’s software and even VSTs that can extract the individual parts from songs including the vocals and instruments. Phase inversion tricks are a thing of the past. TRAX and XTRAX.
https://audionamix.com

User avatar
selig
RE Developer
Posts: 11685
Joined: 15 Jan 2015
Location: The NorthWoods, CT, USA

03 Jun 2018

miscend wrote:There’s software that can extract the individual parts from songs including the vocals and instruments. Phase inversion tricks are a thing of the past. TRAX and XTRAX.
https://audionamix.com
there you go, figured this was still a ways off - don’t keep up with this stuff mainly because I have no use for it in my current work. Thanks for the heads up - I don’t know how well it works, but technology is marching on!


Sent from some crappy device using Tapatalk
Selig Audio, LLC

User avatar
EnochLight
Moderator
Posts: 8405
Joined: 17 Jan 2015
Location: Imladris

03 Jun 2018

miscend wrote:
03 Jun 2018
There’s software and even VSTs that can extract the individual parts from songs including the vocals and instruments. Phase inversion tricks are a thing of the past. TRAX and XTRAX.
https://audionamix.com
Yep, well aware, in my OP I posted an online version option. ;)
Win 10 | Ableton Live 11 Suite |  Reason 12 | i7 3770k @ 3.5 Ghz | 16 GB RAM | RME Babyface Pro | Akai MPC Live 2 & Akai Force | Roland System 8, MX1, TB3 | Dreadbox Typhon | Korg Minilogue XD

User avatar
EnochLight
Moderator
Posts: 8405
Joined: 17 Jan 2015
Location: Imladris

03 Jun 2018

selig wrote:
03 Jun 2018
EnochLight wrote:
03 Jun 2018
OK, thanks. It's not really important - was just trying to get it to work well in Reason. The online service is of course an option - it's not expensive for just 1 track.

Thanks!
So what happens when you do the inversion trick?
I’m having a hard time setting it up.
Win 10 | Ableton Live 11 Suite |  Reason 12 | i7 3770k @ 3.5 Ghz | 16 GB RAM | RME Babyface Pro | Akai MPC Live 2 & Akai Force | Roland System 8, MX1, TB3 | Dreadbox Typhon | Korg Minilogue XD

User avatar
selig
RE Developer
Posts: 11685
Joined: 15 Jan 2015
Location: The NorthWoods, CT, USA

03 Jun 2018

EnochLight wrote:
selig wrote:
03 Jun 2018
So what happens when you do the inversion trick?
I’m having a hard time setting it up.
All you do is import both tracks and invert one!



Sent from some crappy device using Tapatalk
Selig Audio, LLC

jimmyklane
Posts: 740
Joined: 16 Apr 2018

03 Jun 2018

selig wrote:
03 Jun 2018
EnochLight wrote:
I’m having a hard time setting it up.
All you do is import both tracks and invert one!



Sent from some crappy device using Tapatalk
Exactly. Using the vocal and the complete mix WITH the vocal means the vocal should null with the exception of the stereo information (i.e. reverb, chorus, stereo delay, etc)
DAW: Reason 12

SAMPLERS: Akai MPC 2000, E-mu SP1200, E-Mu e5000Ultra, Ensoniq EPS 16+, Akai S950, Maschine

SYNTHS: Mostly classic Polysynths and more modern Monosynths. All are mostly food for my samplers!

www.soundcloud.com/jimmyklane

User avatar
EnochLight
Moderator
Posts: 8405
Joined: 17 Jan 2015
Location: Imladris

03 Jun 2018

selig wrote:
03 Jun 2018
EnochLight wrote:
I’m having a hard time setting it up.
All you do is import both tracks and invert one!
Hmm.. that's what I thought, but it's not working. I've tried inverting each one, yet the vocals are still present in the mix for some reason. :( Oh well, thanks for trying!
Win 10 | Ableton Live 11 Suite |  Reason 12 | i7 3770k @ 3.5 Ghz | 16 GB RAM | RME Babyface Pro | Akai MPC Live 2 & Akai Force | Roland System 8, MX1, TB3 | Dreadbox Typhon | Korg Minilogue XD

User avatar
moneykube
Posts: 3447
Joined: 15 Jan 2015

03 Jun 2018

what if voice is stereo???
https://soundcloud.com/moneykube-qube/s ... d-playlist
Proud Member Of The Awesome League Of Perpetuals

User avatar
selig
RE Developer
Posts: 11685
Joined: 15 Jan 2015
Location: The NorthWoods, CT, USA

03 Jun 2018

EnochLight wrote:
selig wrote:
03 Jun 2018


All you do is import both tracks and invert one!
Hmm.. that's what I thought, but it's not working. I've tried inverting each one, yet the vocals are still present in the mix for some reason. [emoji20] Oh well, thanks for trying!
There are two main things that could be happening, in addition to the things I’ve already mentioned.

One is timing - if the timing isn’t sample accurately align, even identical recordings won’t cancel perfectly.

The second is level - if the levels are exactly the same, even identical recordings won’t cancel perfectly.

Start with checking timing (because it’s easier to hear), nudging one track forward or backwards and listen for the telltale ‘phasing’ either increasing in pitch (getting closer) or decreasing in pitch (getter further).

Once you get the timing as tight as possible, if you’re still not getting the cancellation you expect, adjust the level of one track up or down slightly and see if the cancellation increases or decreases.

There can still be the problems I mentioned earlier - if there is reverb remaining despite the dry signals canceling perfectly, it’s the reverbs not being “printed” before exporting. Same for any residual FX you may hear - sorry, nothing further can be done if this is the case. If the cancellation comes and goes dynamically, then there is master compression on one or both tracks - and since the files were exported with master dynamics, again there is nothing further you can do.

In short, files need to be exported with this explicit task in mind or you may have some or all of the issues above!

It’s a good learning experience to at least try to solve these things, though it’s probably not something you’ll want to do too many times if you encounter any of the above issues! Let us know how it works out and what you discover about the files. [emoji3]

[also note, after doing the inversion trick and you still have some residual vocals, you can try the old vocal remover trick: use a stereo imager to split the track around 200 Hz or so, take the high side and make it mono with one channel polarity reversed (to get rid of the center channel).]


Sent from some crappy device using Tapatalk
Selig Audio, LLC

Post Reply
  • Information
  • Who is online

    Users browsing this forum: No registered users and 6 guests