Is Pro Tools 11 really the 'new standard for audio production'?
YouTube Review - The Pulse: You won't find me there
Mixing: Where to start? - The drum kit?
To become a hit, your song needs a BETTER TITLE!
Q: When should I normalize, and by how much?
How to record or amplify the melodica or any unfamiliar instrument
The first rule of acoustic treatment
Q: How should I set the compressor in a live stage show?
Golden Globes 2011
Recording acoustic guitar in stereo - should you use spaced or coincident mics?
There was a time, sometime before the 1970s, when singers could sing. But that idea fell out of favor, and now we have singers that can sing a bit, but can't sing in tune all of the time.
Fortunately we have plug-ins to cure any tuning problems - Auto-Tune is the classic, Melodyne is another.
But wait... before you reach for that plug-in, there's an old-fashioned technique that is better. Easier too, and it costs nothing (that's why you don't hear much about it!).
This technique is called 'punch-in', or sometimes 'drop-in'. It was developed in the days of analog multitrack tape recording. On early machines, it didn't always work too well, but machines from the mature age of analog recording can do it flawlessly.
It works like this...
Suppose everything is OK except for a couple of notes in the middle of one line. There are gaps at the beginning and the end of that line.
What you need is for the singer to repeat and re-record just that one line.
So the engineer listens to the song carefully so that he is sure where the line comes (some engineers take down timer readings, but this shouldn't really be necessary).
He backs up the tape, probably to a point in the middle of the previous chorus. He puts the tape into play, and the singer starts to sing along to his previous performance, to get in the mood.
When the gap before the offending line comes, the engineer holds down the play button (while the tape is playing) and presses the record button. This will switch the recorder into record mode. At the end of the line, the engineer either presses stop, or presses play by itself (depending on the machine) and the recorder drops out of record and back into play mode.
And that's all there is to it. If the singer still doesn't get it right, the punch-in can be done again.
There is little to go wrong as long as the engineer presses the buttons at the right time.
These days with hard disk workstations you can automate the punch in process so it can be repeated as often as necessary with total accuracy. You could do this with some tape machines too.
However, I find that once you get bogged down in fiddly little numbers and precise settings, you lose the 'flow' of the session. If you do it the manual way, if your software allows it, there is nothing to think about but the music.