Friday, June 20, 2014

Producing "Layered Tracks" Using Gain Matching.

Have you experienced this problem when producing your layered vocal (string, brass, etc.) tracks?

You know you are going to layer back-ground vocals, three parts, two deep on a part, using only one singer.  A total of twelve tracks.

As an example, here's a screenshot of a session based on my "Stevie TV" song, "Sexy Mistake."  The scratch (muted) vocal is at the top.  The next twelve tracks are the proposed back-ground parts.  The bottom tracks are sub mixes of the instrumental tracks for recording.



If you record the multiple vocal tracks as you would a solo vocal, and use a normal routing, when you balance the first back-ground vocal against the track, it sounds balanced; but as you add the additional vocals, unless you remix the vocals after you record each part, the back-ground vocals get too loud relative to everything else.  By the time there are twelve vocals, the singer (and you) can't hear what they are singing; nor the track.  They can't match or blend the vocals and the intonation goes to hell.

Likewise, if you record the first back-ground vocal where they might actually sit in the mix when all twelve vocals are layered, the first parts are too soft and you can't really judge the performance as you record them.

There is a simple solution to this problem that with just a bit of preplanning, works perfectly to keep everything in balance throughout the recording process without having to remix the vocals after every pass.

I don't know what the technical term might be, but I call it "gain matching."  It works as follows:
  • You start by creating in the DAW, the number of vocal tracks you will have when all layers are present; in this case, twelve tracks, as was done above.
  • Next, you set the input of all twelve tracks to be the same source: your single microphone.  An easy way to do this is to highlight each track name, then select the input while holding down both the "option" and "shift" keys.

  • Then, you set the outputs of all the tracks to a single aux track.  The aux track will control the volume of all the combined vocals relative to the other elements.  This makes the overall volume of the back-ground vocals easier to control.
  • Next, you put all twelve tracks into record mode (again, use the "shift-option" key command while record enabling the highlighted tracks as a shortcut). 

  • At this point, all twelve tracks will be passing sound from the single microphone to the aux tracks (labeled BG Vocals above), summing their combined volume together.  The energy of this summed audio will be matched to what the twelve unique tracks will create when all the unique vocals have been recorded.  You then set the level of the aux track where it sounds best to you, routing the mix to the singer as well from a send off the back-ground vocal aux track. 
  • After you record the first vocal, you take half the tracks out of record mode for the second pass.  Now you will be hearing the first part (six tracks), balanced with the second part (six tracks), and the net volume of the combined tracks will be the same volume as you heard before.
  • Then, as they say on the shampoo bottle, "Lather Rinse Repeat."
If hard disk space is an issue, and you don't want to physically record all the vocals as you go, you can use "input mode" for the vocal tracks instead of record mode - leaving only the actual track you want to record in record mode. 

I prefer not to do this as I have plenty of disk space and it requires a few more key clicks along the way to keep everything balanced and a slightly more complicated routing (for another blog).

Like this post?  Check out my other music tech posts!










No comments:

Post a Comment