Jump to content

Splitting stereo mix/Removing vocals


Logicno8

Recommended Posts

I just ran upon this tutorial which sounded interesting by title "Removing Vocals from a Song" .....

 

Now that one was done in ProTools but I found the same example for Logic as well here.....

 

Now I have few thoughts and few questions about this. Firstly, I don't see why would anyone try to remove or attenuate something which is impossible to remove from the final mix (such as lead vocal) since it is embedded with the track. Karaoke ? Ok. One of the guys (PT guy) said it was for Karaoke.

So, I tried it right away to see if it works and of course, by the laws of audio it doesn't just as I thought. If you use this on a ac.guitar/vocal track...hmm it may kill some vocals in but If you try this with a full band mix it will kill EVERYTHING in the mix, make it sound out of phase and demolish the track all the way. So you can't do it with anything except small acoustic mixes. I even tried to scoop out vocal frequency and of course I scooped out most important sonic content of the song. That was desperate try to make this happen :D

 

So....question 1 : Is it just me or you guys as well don't see a chance to partially (like 70 percent or more) remove lead vocal from the stereo mix and still retain decent sound ?

 

2....I see ProTools (and some other DAW's as well) can split stereo mix file into 2 different separate mono tracks. And I really never felt like doing this but I tried anyway and I thought since Logic does not represent stereo files as 2 wave forms on top of each other that I could maybe use SPLIT option as my bounce file type to get some commercial mixes splited and I couldn't make it happen.

How to do this in Logic ?

I bounced out a Nickelback's song which has predominant guitar riff on the left and clean on the right. I was expecting to get L & R sounding files but I ended up having 2 same mono files. Not one for each side. Can anybody explain this to me ?

 

Thanks a lot.

1435488114_StereoLampR.jpg.e89415e95ae00d6e6b0c9e9fcbdb27f7.jpg

Split.jpg.d3052d7096112264e4179bbbc6f03f80.jpg

Link to comment
Share on other sites

Why would you need to that?

 

Pro Tools will convert the tracks for you when you export them over. It has to do that cos it doesn't work with interleaved files, there's no need to give yourself any extra work to do.

 

I know. You're completely right but I don't understand what's the difference between SPLIT and INTERLEAVED in Logic ? If I bounce as I usually do, Interleaved I'll get one stereo file. If I choose SPLIT, I'll get the same but twice. 2 files with exact sam panning an everything. I thought I should get L&R channels of my stereo mix. That's what confusing me. I don't get the difference.

And by the way I like the way Logic represents stereo files. As an uneven big wave form (as up and down are stereo thus uneven because of the wide stereo arrangements). I never needed to split stereo mixes but I saw this ProTools guy's tutorial and he did some phase reversing trick and than alined 2 channels of a stereo mix to reduce phase problems (you can watch that first video link in my post above).

Link to comment
Share on other sites

Ok, I watched the video in your link to see what it was all about.

 

I've never bounced a split file before but thought I would give it a try to see what happens, the manual suggests it should be straightforward enough. All I did was select Bounce then change from Interleaved to Split just like your screenshot & sure enough I've ended up with a file marked L & a file marked R in the audio bin, you can clearly see in the waveform they are different. I don't know how you've managed to get 2 left channels, open a new project & try it again.

 

I'm only on my Macbook now but tomorrow I'll try it on the main computer to see if I can remove the vocal using the technique on the video, at this point I think it's probably do-able.

9285193_Screenshot2011-03-01at00_56_59.png.33754c78c9b74d89985f63999248e2f6.png

906960597_Screenshot2011-03-01at00_50_21.png.994d16d732cb83e42f3bf016f4282370.png

Link to comment
Share on other sites

Thank you for trying to help me. I never used SPLIT option neither but wanted to try vocal trick now and I'm not sure how you did it but I tried exporting my mix from arrange window, I tried mp3 files of famous bands and my own bounced wav, aiff and mp3 files. I think I failed to split them :-( The best example would be to have something that is obviously panned hard left/hard right (as in my first example with Nickelback's song) and THAN try to listen for the sounds (cymbals, guitars, fx, whatever) and if hi hat (for example) is dominant from the L file and ride cymbal from the R file, you made it. I didn't, have no idea why....
Link to comment
Share on other sites

Hey Logicno8, I can confirm for you that your vocal removal trick can be done in Logic.

 

I used the track South of Heaven by Slayer cos at the start there's clearly different things happening on either side of the stereo image.

 

You must make sure that after you bounce to a split file you select the two split tracks in the audio bin then Edit>Disconnect Selected Split Stereo File.

 

Now you will have a left & a right track, following the technique in the video you linked I inverted the phase of the left track using the gain plugin & the vocal almost completely disappeared. I found that the vocal went just by inverting the phase but if you want to offset it use the Sample Editor to move the anchor point.

 

Finally to bring back the low end I put a low pass set to about 120 on the original track & I'm left with a karaoke version of Slayers South of Heaven :)

1230395429_Screenshot2011-03-01at17_09_47.png.795d0204b32137167b3bb0eaeea3090c.png

641016996_Screenshot2011-03-01at17_09_02.png.486384254ac20c28bf145a40f972bfaa.png

Link to comment
Share on other sites

I'm glad to hear that you succeeded. Now...I didn't know trick with Audio Bin/Edit/Disconnect Selected Split Stereo File. I engaged that and I think I made it. I say "I think" because I'm not sure is this what I should get. I used a song S.E.X by Nickelback because it starts with a bass and a heavy distorted riff on the left side. So here you can see a snapshot of my 3 regions. Grey one is an original stereo file (actual Nickelbac's song). Red regions are Left and Right , top to bottom. You can clearly tell the difference just by looking at this wave forms but what I'm not sure about is WHAT should I HEAR. I put them on 2 mono tracks (since it's split stereo file). Than I hear left with head blasting guitar riff in the middle and the other (right) with guitar not from the far left as I expected but ...weird, like far away, phasey...Take a listen. I posted mp3 in attachment and first you will hear is the original, than left channel playing the same, than right playing the same. Judge by yourself if this is what I should end up with. Is this normal to hear it this way. I guess if I was to pan them hard left, hard right that I would hear what's in original stereo file.

S.E.jpg.7eb1933ca038d823579817f9e0cf17f0.jpg

Sex.mp3.zip

Link to comment
Share on other sites

...Take a listen.…Judge by yourself if this is what I should end up with. Is this normal to hear it this way. I guess if I was to pan them hard left, hard right that I would hear what's in original stereo file.

 

That sounds right to me, the distant guitar you're hearing in the right channel is just how the guitar has been placed in the stereo image. You will get this yourself if you use the direction mixer on some of your tracks - remember the pan control on a stereo track only turns the left/right signals up or down whereas the direction mixer actually moves the instrument about.

 

Try panning your mono tracks hard left & right, you will hear the track exactly how you hear it normally. Also if you pan the normal track hard left & right you will hear what you have in your mono tracks, only coming from one monitor instead of both.

 

Have you managed to get rid of the vocal yet then?

Link to comment
Share on other sites

...Take a listen.…Judge by yourself if this is what I should end up with. Is this normal to hear it this way. I guess if I was to pan them hard left, hard right that I would hear what's in original stereo file.

 

That sounds right to me, the distant guitar you're hearing in the right channel is just how the guitar has been placed in the stereo image. You will get this yourself if you use the direction mixer on some of your tracks - remember the pan control on a stereo track only turns the left/right signals up or down whereas the direction mixer actually moves the instrument about.

 

Try panning your mono tracks hard left & right, you will hear the track exactly how you hear it normally. Also if you pan the normal track hard left & right you will hear what you have in your mono tracks, only coming from one monitor instead of both.

 

Have you managed to get rid of the vocal yet then?

 

I didn't...I split stereo mix, than disconnected those L&R files but when I inverted phase polarity on the left channel 90% of everything disappeared.

Do you place your L&R channels to STEREO or MONO tracks ? Uh...

I followed your workflow and what was suggested in videos but with serious wide productions such as Nickelback that's not possible. I'll try it with Slayer's song you used and let you know how it worked :)

Link to comment
Share on other sites

When I did mine I left the L & R tracks in mono panned to the centre, I lost alot of the bass & drums but when I put the low pass on the original they came back again.

 

I've attached a snippet of mine to let you hear whats happening, it starts off with the mono tracks panned hard left & right. It sounds a bit weird because the left has already been phase reversed & the filtered original is running underneath. I pan the left track to the centre followed by the right track to the centre.

 

The vocal is all but gone along with any width the track started with.

SoH Vocals going.mp3.zip

Link to comment
Share on other sites

Okay, Eric thanks for providing this thread. I read every post 3 times and still I think I'm missing something. David stated that first thing we need to do in order to make this happen is to make 2 mono channel strips and to put these stereo mp3 (or whatever it happened to be) files onto these tracks. Ok, I did it. Now we need mono instance of GAIN plug-in on one of the channels. I did it. Now I pan them hard left and hard right (I have Gain plug-in on the left panned track). When I use PHASE INVERT option I can still hear everything but out of phase. Which seems perfectly logical to me since I have 2 identical signals panned L & R and not actual y L & R side of a mix. So I don't lose anything it just became out of phase and sounds disgusting like everything out of phase THAT much. If I pan them dead center both, obviously they cancel each other out, I get silence. Again, perfectly expectable. Now, I'm really not sure how you guys do that but I told you...I followed what David suggested on that thread. So to wrap it up 1. I add 2 MONO channel strips. 2. I put mono GAIN plug-in on first track. 3. I import song I want to work on and copy it to both channels. Now what...

1.jpg.53014fa533a87b76f542fc77feed853c.jpg

Link to comment
Share on other sites

ohhh....I see where my mistake lies...I need to make 2 channels to be LEFT ONLY - RIGHT ONLY which I get to by press+hold on that circle button. I see, I see... Okay, still not impressive results but at least I know where I've been mistaking.

But ruari...you did a great job with Slayer, I listened and although you almost lost snare vocals were gone as well :D

Link to comment
Share on other sites

  • 1 year later...

I accidentally deleted my demo vocals recently and only had a version of it with gtrs in the background, then I read through this thread and managed to get a pretty clean acapella back using this split stereo inversion method! Yay! THANKS heaps guys! 8)

Just sharing what I've done:

 

1. Bounce both original (music with vocals) & instrumental as split stereo files

2. Pull them into arrange

3. Mute either side of both, eg. mute L original & L inst / mute L original & R inst (works either way)

4. Select the Inst track that is not muted and go to Sample Editor > Function > Invert

5. Playback the unmuted tracks (eg. L original & L inst) to check, music should be mostly removed by now

6. Bounce/export the acapella & it's done!

 

Cheers!

Link to comment
Share on other sites

Hi crimmynull, and welcome to Logic Pro Help. Thanks for reporting your experience. However if I'm understanding you correctly, you have both the following audio files, where the music is the same?

 

• Music + Vocals

• Music

 

If that's correct then to get Vocals becomes a simple mathematical equation: Vocals = (Music+Vocals) - Music. Meaning you mix both audio files, but on the Music channel strip you insert a Gain plug-in and invert the phase of both the left and the right channel.

Link to comment
Share on other sites

Hi crimmynull, and welcome to Logic Pro Help. Thanks for reporting your experience. However if I'm understanding you correctly, you have both the following audio files, where the music is the same?

 

• Music + Vocals

• Music

 

If that's correct then to get Vocals becomes a simple mathematical equation: Vocals = (Music+Vocals) - Music. Meaning you mix both audio files, but on the Music channel strip you insert a Gain plug-in and invert the phase of both the left and the right channel.

 

Thanks for warm welcome!

Yeap I had both files where the music is the same.

I did try the gain plugin method, but for some reason the result wasn't as clean. :roll:

Link to comment
Share on other sites

Archived

This topic is now archived and is closed to further replies.

×
×
  • Create New...