F’rinstance, John Fogerty played all the instruments on “Centerfield”, Paul McCartney played everything on “McCartney”, etc.
How do you do that?
Which track do you record first? My guess would be the drum track. So are you “playing” the song in your head as you drum away? Or do you have someone playing guitar/piano/whatever along with you? But if that’s the case, how do you prevent them from appearing on the recording?
They lay down different tracks, and use mixing equipment to get the finished results. From an engineering standpoint, it doesn’t matter which track gets laid down first; eventually, they’ll all be mixed together. In practice, though, it might be easier to lay down a piano or guitar track first that the artist can hear through headphones as he plays the drums.
It used to be done on tape, but thanks to sophisticated computer software, it can be done much more quickly and easily and with better results. The producer can alter any track to achieve a desired sound, then alter the finished product as he wishes instead of having to splice tape.
Also, depending on the setup, not everything has to go into the final recording. When I produce stuff, I can pick and choose which tracks I want to go into the final mix.
Really, thanks to computers, it’s not as magical as it used to be.
A friend of mine recorded a few tracks for a demo like that. He had session musicians record the song first, then listened to the recording in his headphones while he played each instrument. (He did two guitar tracks, piano, and three horn trakcs; another guy did drums.)
I’m sure there are other techniques as well. Some musicians are good enough to simply play a part without needing to hear the others.
I’ve always assumed, as a non-musician, that one would look to the drummer to keep the rhythm. So if you were laying down all the tracks yourself, it would be difficult to do the drum track trying keeping the beat to the prerecorded guitar track. Maybe I’m thinking too much about this.
Some folks just use a generic “click track” to keep rhythm if they haven’t got a drum part worked out yet. As I play guitar best (relatively speaking), I’m prone to recording the guitar track first, and then sort of filling everything else around it. To keep rhythm I just use a track that’s little more than a metronome, or snag a repetative rhythm track from somewhere (a Garageband prefab track, for instance), and then replace it with something from my head, more or less. Every once in a while I’ll be thinking of, say, a bass line first, so I record that first; or maybe the keyboard part was what I thought of first, so that’s what I record first. Unfortunately, the only real instruments I have are a few guitars; the rest must be synthesized, and as I’m a pretty crappy keyboard player, I rely on MIDI pretty heavily to get something that approximates a decent performance. Again, not surprisingly, the first thing I tend to lay down is the guitar part or parts, because that’s the only thing I’m approaching decent at doing spontaneously. I imagine other musicians take something of a similar approach: Start with what they know best.
I generally play all instruments in the music I write, although I do bring in guests from time to time to breath new life into things.
I start by using a metronome to record a “click track”, which is basically a track that does nothing but click at a constant rate. Using this as a guide, I can record any instrument as long as it is in sync with that guide. I will usually lay down a piano track first because it provides a structure and helps me keep my place in the song. Generally though, this track is what is called a “scratch track”. More than likely, it will not be included in the final song.
At this point, I will record drums and bass. If this sounds good and has a decent groove, I will start adding instruments. I generally add more tracks than I will use. When things sound full enough or even overly so, I will start mixing the tracks and removing what I don’t like.
Doing it all yourself has it’s pros and cons. You have to rely entirely on instinct because you aren’t bouncing ideas off of other musicians. You also have much more control over how you want things to sound though and it really just takes having a clear picture of what you are trying to achieve.
I always thought that too, but I’ve actually met a couple of people who did all the instruments on recordings, and in both cases they put down the guitar tracks first. Both were guitarists first and foremost, so using their strongest instrument to start things off was their preferred method.
Here’s how we do it (band in studio or solo multi-tracking):
Decide on a tempo and put it on a click track (in the headphones). This will help whoever plays the drums on the actual track, whether it be solo or band.
Go into a vocal booth with guitar. Start click track, count down (4, 3, 2, 1). Run through the song on tape at tempo.
None of what you have just done gets on the recording. It’s your rough blueprint. The drummer comes in next and plays along to the demo. Then the bass player comes in and does his thing with the demo and (real) percussion track in his headphones.
Step 3: Profit! (everyone else fills in their part, or the solo guy starts tracking more takes).
It’s like taking candy from a baby. In the late 80s, Pete Townsend released an album called ‘Scoop,’ which was just his cassette demos of all the great Who songs (with him singing all the vocals). It’s a splendid look at the meat that goes into a huge sausage.
A guy, a guitar, a click (or metronome). That’s how it starts, unless the guy with the guitar plays piano. I suppose you can guess how that starts…
I play a limited number of instruments, but I’ve recorded a few pieces with multiple instruments. Same deal: A click track, then maybe the rhythm guitar. Then some lead stuff. Then maybe vocals on top of that.
Let me first respond to this. Hogwash.
It depends on the artist and if the band is a “product” or not. If the release is a actual band, they play. It would be rare for someone like Green Day, The Cure, Modest Mouse or Cake to bring in people to lay tracks for them. Perhaps a friend might play something somewhere on the album, but the bulk will be the band.
Personally, I’d like to know where you got this information. I don’t often call for “cite” but CITE?
Now,. on the the OP.
That is how I do it. Also the use of “scratch tracks” is a studio staple food. You might lay a simple drum track along with a click track to get the basic groove, but it will be replaced later with the real drum track.
Not long ago I did a Ska song. As I was working the piece out I used a distorted keyboard line for the guitar section. It was never intended to remain in the final mix but I wanted something to work with while I got the horns figured out. A few days later the keyboard track got deleted and the real guitar track was laid.
There is a ton of freedom in the studio because you have wiggle room with tracks. It’s not like you only have four and you can only record once. You can keep re-recording, adding and deleting. With digital, you can move tracks around. So if you like the back up vocal from the first chorus, you can copy/paste it to the second and third chorus if you want. No one will EVER be able to tell because it’s back in the mix. In fact, call me lazy, but I do that all the time.
My experience is as a Hollywood studio musician in the 1970’s & 80’s. Things may have changed since then, and my experience was mostly with non-royalty groups.
With that said, there were many factors that went into deciding how to build a song. Yes, a single musician could do everything, but that would take a tremendous amount of studio time. Better, and more common, is for a good rhythm section to go into a small studio and record a solid set of tracks together. They work as a unit, think as a unit, play as a unit, and can lay down a song with a minimum of studio time.
Sometimes a scratch vocal was added just to guide things along, since a demo, a scratch mix, and a lead sheet would often go to an arranger who decided what other instruments to add and where (in the song).
Since different studios have different strengths – size, room acoustics, electronic equipment available – the studio that recorded the rhythm tracks might not be the one that added the strings (or even in the same city). And, again, although it is possible for a single violinist to record an entire section by overdubbing, I never heard of this being done, although small string sections were often multiplied into big ones.
So most of the overdubbing was done by sections, not individual instruments, although solos could always be added later.
Although I didn’t work with Motown, this is basically how they did things in the 60’s.