Can a compressor correct vocals that are at different levels?
Is there such a thing as a loudspeaker that doesn't sound like a loudspeaker?
Do you get a sore throat when you sing?
How much mastering does a Pink Floyd soundalike band need?
Create an amazing trance riser in 7 steps
Will.i.am is giving up music to learn computer programming
Another amazing bargain keyboard instrument
How loud should the bass instrument be?
How much difference does mastering really make? [with audio]
Now at last you can replace your nearfield monitors with proper main monitors!
I discussed double tracking in some detail in a previous article and, if you're not familiar with this technique from the 1960s, I recommend taking a look first. But when you understand what the technique is and what it can do, the next step is learning how to double track efficiently. Even the Beatles found it tedious to double track their vocals. But now, we have modern tools and methods that didn't exist decades ago.
Let's take the case of double tracking a vocal, and suppose that you already have one perfect vocal take. In the 'olden days' of recording, when tape was king, the simplest way to double track would be to roll the tape again and record a complete second take. Since studio time used to be extremely expensive, there was a good deal of tolerance for inaccuracies. In many cases these supposed inaccuracies would add to the texture and actually make the result more interesting.
So you can do that! Simply click back to the beginning of the song, create a new track and record the vocal again. If you're going to be a purist about it, then you should try as hard as possible to be satisfied with your first attempt. Listen back a few times - what might seem like an inaccuracy, or even a sloppy moment, on first listening might actually be worth keeping. Judging that kind of thing is part of the fine art of production.
But suppose you do want as much accuracy as possible, perhaps to the point that the average listener wouldn't be aware of any double tracking, other than enhanced richness of texture. There are two parts to the solution...
Where a complete take will probably be riddled with inaccuracies, minor and major, it is much easier to achieve perfection working in short sections. A singer can listen to a section of one or two lines a few times and quickly learn the pacing and inflections, then duplicate it for the double track. When one section is perfected, you can move on to the next. Thus the double track can be created piece by piece working through the song.
Working in short sections is great. But there is an even easier way...
Looping is a technique from the history of film sound. Not so long ago it was common practice, and even standard practice, to re-record all of the dialog of a movie. So the actors would speak their words during shooting, then record everything again in the sound studio. The advantages of working this way are 1. That the director can concentrate on getting the visuals right during shooting. 2. The dialog can be recorded to perfection in both performance and sound quality. 3. Separating the dialog from sound effects makes foreign language dubbing easier.
In the days when film sound was recorded on film rather than on tape, a projector would be loaded with a loop of film covering a couple of lines of dialog. At the same time a 'sound camera' would be loaded with a loop of exactly the same length. The actor could watch the loop through several times and hear the original location dialog, then have several goes at repeating the performance. Actors would become very skilled at doing this and they could work their way through all of their lines fairly quickly (I was told by an industry insider that Oliver Reed worked better and faster when he was drunk!) Of course it was very labor-intensive to make up the loops, but that's the way things were in those days. Let me now skip back to the present day...
If you delve into the menu system of your DAW you will probably find that it has a loop recording function. The way it works on Pro Tools, as an example, is that you highlight the section you want to loop over, and hit the record button. Pro Tools will go over and over that section as many times as you like until the singer produces the perfect take. What's more, you can easily access all of the previous takes if you wish. Stopping when perfection , or passably near-perfection, is reached is much faster than making loads of takes then trying to sort through them.
It helps tremendously if the section you highlight has a pre-roll and a post-roll around the section you want to double track, otherwise the singer has no time to breathe between takes. It also helps to highlight a whole number of bars, usually in multiples of 2, 4 or 8, so that the rhythm is continuous, without a glitch each time the loop restarts.
I've been treating it as obvious that the singer must be able to hear their first track when they record the double, otherwise they have no reference point and cannot hope to double track with any degree of accuracy. If it isn't obvious, then one attempt at doing it without monitoring the original will soon show you the way to go. But there is a problem...
When a singer, or any instrumental performer, double tracks their own performance, monitoring both versions in their headphones, they can't easily tell which is the original recording and which is the version they are actually recording now. That's something you won't fully understand until you try it. It's a weird feeling not being sure that a note you are singing is actually you live right now, or something you did half an hour ago. Non-performing engineers should try this, so that they know how the musicians they are recording feel!
There are two alternative solutions to this. One is for the singer to slide one earpiece of their headphones back a little, so that they can hear their voice coming through the air, as well as the original take in the headphones.
The other is to pan the original take halfway to one side, and the new recording halfway to the other, just for the foldback. The singer will then easily be able to tell their new performance from the original.
This is an issue that didn't affect musicians or engineers in the early days of double tracking. But suppose that you want to Auto-Tune the original take? No problem, if that's really what you want to do. But should you then also Auto-Tune the double track?
I'd say try it! An Auto-Tuned double track doesn't sound at all like a natural double track, but it has an interesting texture in its own right. And you might try changing the parameters for the double - maybe alter the speed of pitch correction, or apply an overall pitch shift of a few cents up or down.
In summary, there's loads of fun to be had with double tracking. And it is a pure studio technique that makes your recording totally different to a live performance. Happy double tracking!