When recording rock music in a studio, it’s common for the vocals and each instrument to be recorded on a separate track. In fact, the band doesn’t even need to play together; often the parts are recorded separately and then overdubbed or mixed together.
I’m curious as to how recordings of an orchestra playing classical music are done. Presumably the entire orchestra plays together, but is each instrument mic’d separately? Or is each section mic’d separately? Or (for a stereo recording) do they just put two microphones in front of the entire orchestra and record it that way? Is there any mixing done after the recording like with rock music, or is that frowned upon? For example, if the producer thinks that, say, the clarinets aren’t coming across loud enough, would that be something that would be fixed in post-production by increasing their volume in the mix (assuming they have their own microphone), or would that be fixed during recording by telling the conductor and having the orchestra re-record the entire piece?
They play it all the way through then re-record short parts of the piece to fix a squeaking oboe, flat violin, etc. With digital editing they can fix a single-note.
There are various methods, but with modern multi-track recording, individual instruments are miked and recorded to separate tracks, all in sync. For instruments in groups like violins, more than one player may be on one mike/track. Since violins typicaly share a music stand in pairs, mikes are often done the same way, or in groups of four with overhead mikes.
Separate tracks allow for mix control later and substitution as DanBlather says.
There are purists who insist on a single stereo mike in front of an orchestra, and for some live recordings, that is done, but this relies on the orchestra and conductor to control the balance live and very little can be fixed later. Also, room acoustics are a larger factor this way, and cannot easily be altered later.
It also depends on the budget. Individual miking/tracking is more expensive than direct-to-stereo.
Really? That’d be a hell of a lot of mikes. The recordings I’ve done have involved microphones scattered throughout and around the orchestra (frequently on high stands pointing downwards) but not as many as a 1:1 or 2:1 mike-to-musician ratio. You want to hear a certain homogeneity of sound, not a bunch of individual instruments.
(Choirs are even more diverse - usually one to two mikes per section, out in front. I’ve heard the results of recordings where the mike placement was done badly and you can pick out individuals. It sounds awful.)
Depending on what the balance issues are, some can be fixed in the mix and some require re-doing a section. Repeats are usually done in short sections depending on where the best places to blend the splice in are.
And to record a CD can take anywhere from a few hours to a few days, depending on the issues involved.
Who said anything about a 2:1 mike-to-musician ratio? Are you thinking that each instrument would be recorded in stereo with 2 mikes? Not usually, although a drum set may be recorded with a half-dozen mikes and a piano is often in stereo. A section, like a trumpet section, may have only one mike. The stereo mix is typically created by panning at mix time, not with two mikes on everything.
There is a trade-off, and it’s a matter of taste. The more “homogenous” the sound is, the less control you have over it later. Echo/room ambience can easily be added, but not as easily removed. Combining instruments on one mike makes it hard to separate them later unless they are of different pitch ranges or play at different times. Engineers and producers have different techniques and different reasons behind them.
Well, in some cases I think you would want to focus on individual instruments. For example, what got me thinking about this question was Strauss’s “Tales from the Vienna Woods”. It features a very soft zither solo. Now, the zither isn’t a particularly loud instrument, especially the way it’s played in this piece, so I was wondering if for the recording they give it its own microphone. (Come to think of it, I wonder how the instrument is even audible when the piece is played before a live audience!)
It’s entirely possible the zither was recorded at different time than the rest of the orchestra, except in truly live recordings, of course. It also could have been placed in an acoustic isolation booth. “Gobos,” or baffles, are often placed inbetween instruments or groups to help isolate the sound. A baffle might be a few feet high, allowing players to see over them, but blocking some of the sound so it doesn’t leak into an adjacent mike. Isolation booths have windows for the same purpose.