Comping a vocal track with Ableton Live



Although Albeton Live is not traditionally used for vocal-centric music I have begun to find it one of the better pieces of software I have used for exactly that purpose. One of the reasons for this is the ease with which you can record multiple takes and then create a composite track from the best parts.

I’m sure there are many ways to perform this task but this post is about the method that I found worked best for me.


When I create a song using Ableton Live I tend to use both the Session and Arrangement view to create a version for performing live and a version to record. I start in the Session view as this gives me the greatest flexibility to experiment while the overall arrangement of the song is still in flux. With Live, rather that recording the whole song from end to end as I would have done with more tradition recording software such as ProTools or Ardour, I will record loopable sections and create a Scene for each. For example, if the verse alternates between a couple of chords, I will create a sample with just those two chords and load that sample into a Scene called “Verse 1”. I will then go on to create Scenes for all the other verses, choruses and instrumental sections. These Scenes will also include clips for bass and drums etc as appropriate.

One of the great things about Ableton Live is that none of the clips in a Scene need to be the same length. The drum pattern may be one bar long, the chord progression two bars long and the bass line eight bars long. Live will simply keep looping the clips, and you can ensure that they are all kept synchronised by setting the global quantisation to one bar.

Once these backing Scenes have been created it is time to lay down the main vocal track for each. I start by creating an audio track for the vocal and making sure the track is armed (the red record button at the bottom of the channel strip is lit.) I also turn monitoring to off on the vocal track as I do direct hardware monitoring - your setup might be different.

Next I will launch the Scene for the first verse. I’ll let this roll for a little while and warm up a little by signing over the top. This is also a good opportunity to check the level of the signal that is coming from your microphone and adjusting accordingly. Make sure that you test singing your loudest and that the signal never exceeds -1 or -2 db.

Once I’m ready I’ll just click the record button on the empty clip in the vocal track that goes with the Scene I am listening to. Because the Scene will just loop indefinitely, I’m free to sing the verse over and over without ever touching the keyboard to rewind. I like this method of focusing on one piece of the song at a time as it lets me really settle into a particular section before moving on. I’ll normally do quite a few takes of verse before I stop, sometimes ten or more, but then again I’m not the world’s best singer…

I’ll then go on and repeat this process for all the verses and choruses straight away as I can create the composite of all the tracks at any time outside of the studio. I also think there is enormous benefit in leaving a bit of time between recording and listening back. When something is too fresh in your mind it has been normalised by your brain somehow and you just can’t hear it objectively.

Now it’s time to build that composite track by grabbing the best bits from all the material you have recorded. Time to grab a glass of single malt whiskey, settle into a comfy chair and listen to all the takes. By launching a Scene you can hear all the takes that were recorded, but you’ll have limited facilities to edit them this way so I will also hit the transport record to build a copy of the song in the Arrangement view. As I listen I’ll try to make notes about which of the takes I thought sounded best so I can go back to that one quickly. Sometimes there might even be a small part of a particular take that is good, even if the rest of that take isn’t so great.

Once I’ve got a copy of all the takes in the Arrangement view I will swap to that view add a new audio track called something like “Composite Vocal”. If you expand the tracks for the original recordings and the composite vocal you will be able to see the waveform for the recordings. I like to zoom in enough to see some detail in the waveform but still far enough out that I can see where each take begins and ends.
 
While playing the song in the Arrangement view you can press Commmand-F to get the display to follow the play head. This will give you a good visual cue as to which take your am listening to at any given time. If I’m lucky, one of the takes will be good and I can use it in its entirety. Sometimes however I will have to pick bits and pieces out of the various takes and stitch them together.

When you start this process, all the various takes recorded are represented as a single clip. To divide that clip up into separate sections that you can move around, click on the waveform at the point in the clip you want the cut to be and press Command-E. This will divide the original clip into two clips. Chances are you will need to make a second cut to remove the specific section of audio you want, so just click on the waveform at the end of the section you want and press Command-E again. This new clip can now be renamed (by right clicking on it) to something like “First Verse”. Now all you need to do is click and drag this clip onto the empty “Composite Vocal” track and position it in the correct place.

If you cut smaller sections of audio from multiple takes and then positioned them in the composite track to make a full verse, you may want to consolidate those sections back into a single clip to make it easier to move the verse around later on. Do to this simply highlight the group of clips and press Command-J. This new single clip can then be renamed as desired. It is also possible to highlight past the recorded audio to add some blank space at the beginning or end of the clip if they is required for easy quantisation of a loop.

Well, there you go, that’s my process for recording a composite vocal track. I’m sure there are many other ways to skin this particular cat but so far I have found this process very easy and straightforward.

 

  1. dangarstu posted this
Blog comments powered by Disqus