Last night, my audio production class took a field trip to local studio Glenn Sound for a brief introduction to miking technique: which microhpones to use for a particular sound, where and how to place them, and so on. It wasn't completely foreign to me--I've recorded in probably a dozen studios, but always on the clock and paying by the hour, and audio engineers tend to shrug off poorly-phrased technical questions in favor of showing you the end result.
One of the demonstrations showed us how different distances between microphone and source can give you different sounds. Most rock and pop music engineers tend to use close miking, in which microphones are placed very close to specific parts of an instrument or amplifier. Each microphone might pick up a very different sound--more treble or bass, a stronger attack, and so on. The engineers then blend these parts and add effects to get a specific sound--sometimes similar to the sound of an instrument just playing in the room, sometimes totally different. This technique gives an engineer a lot of control, but requires some extra work in the mixing stage. Classical engineers, in contrast, tend to place microphones far from an instrument in an attempt to capture its natural character, as well as some of the ambient room noise (caused by the sound echoing off walls and other surfaces).
The demonstration featured a piano. The prof set up one set of mics right above the open top of the piano, and another about 12 feet away in the room. (They were two stereo pairs, the closer pair in an X/Y configuration, the farther pair in a Blumlein Array.) As expected, the closer pair had more treble frequencies, a much stronger transient response, and no room noise. A non-engineer would describe it as crisp, clean, and perhaps a little sterile. The farther pair picked up much more close echo and reverb from the walls of the room, leading to a "richer" or "warmer" tone. (Audio professionals scoff at these inexact metaphorical words, but most of us aren't professionals, so I'll stick with this bad habit for now.)
By taking the two tracks together and changing their relative volumes, you can change the tone of the piano recording. For cleaner, brighter sound, you turn up the close mic pair. For a bigger, warmer sound, you turn up the farther pair.
However, because the mic pairs were twelve feet apart, there was a twelve millisecond delay between the two signals. (Useful fact: sound travels about 1 foot per millisecond.) It wasn't quite as bad as hearing each note twice. Instead, each key sounded like the player was holding it down for a little longer than he actually was. No problem--using ProTools, the engineer clicked the room-mic track backward 12 milliseconds, and behold: the two different tones occurred simultaneously.
The engineer thought it sounded better, and the class nodded and agreed. But to me, the delay sounded better. When the delay was eliminated, the overall piano tone lost a lot of what made it interesting--the ambient noise of the room mic seemed to be obscured by the louder attack and additional treble that were picked up by the close mic.
Thirty years ago, when many classic rock recordings were made, correcting this lag would have been far more complicated. I imagine the engineer would have had to take the second track, send it through an outboard unit that added 12 milliseconds of delay, then bounced the result back to a single track. With a drum kit, maybe--you can't have the same drum sound hitting at drastically different times, or you'll end up with clickity clackety mud, which makes the rest of the band sound off time. (Although Led Zeppelin drummer John Bonham could sound incredible recorded in a stairwell.) But with this piano, it sounded natural enough without the correction that the engineer probably would have let it pass.
The point: digital workstations allow you an incredible amount of flexibility. You can correct almost anything in the mix. But just because you can doesn't mean you should. Experimentation's great, but restraint can lead to a more traditional--and, to some ears, better--end result.