In Recording in Stereo (Part 1), I discussed my experiences with and feelings regarding different techniques for recording in stereo. Here, I'll talk about some practical applications.
I've always loved all types of music and have had the good fortune to be able to make stereo recordings of classical and jazz musicians using the techniques I've learned. One thing I'd never gotten to try however, was to record a full-tilt rock band using these same techniques. I'd recorded rock the traditional way using multiple mono inputs, later combined and placed in apparent locations on the soundstage during the mix. But these recordings, like the others I'd always heard, left me wanting more. While many contained great music, I never got any of the feeling I'd get when in the presence of these bands while they played; never experienced on record that "in your chest", visceral impact from the music. To be sure, I'd heard some of these played back quite loudly but this always felt like I was listening to a gigantic table radio, not the real band.
Having heard the differences between classical or jazz recordings made direct to stereo vs. those made using the traditional, close multi-mic methods, I knew there was a great deal more information that could be put into a record. I knew it was possible to capture a lot more of the feeling of being in the presence of the performance. But I'd never heard it done with rock. Many colleagues told me "you can't do it with rock". Knowing plenty of musicians who are adept at listening to their fellow band mates and balancing themselves without the assistance of an engineer, I wasn't convinced. Recently, after years of carrying the idea around in my head, I got the chance to try it out. A good friend of mine is a musician and after hearing about the idea, he expressed an interest in trying it with his band.
One of the goals of recording in stereo using the techniques I described in Recording in Stereo (Part 1), is to capture not only the sounds of the instruments but the air around the instruments as well. The idea is to give the listener more of the audible cues they would experience in the presence of the musicians, in the actual space where the performance took place. This requires an acoustic environment that will "support" the music being played since this is a primary contributor to the sound of the air around the musicians and to how they react to the interaction of the sound of their instruments with the space in which they are performing.
After a search for a proper space in which to record, we were fortunate to have a small church in upstate New York made available to us. The plan was to have the band arrive on Saturday for set up and rehearsals and to record the session the following day. On Saturday night I received the phone call informing me the connection of a space heater in the church blew the AC power. Murphy had struck. Still, I decided to drive up on Sunday and see if we could do anything to remedy the situation. On Sunday morning, with the church AC still out (all fuses were intact), the mood started to sink lower until someone pointed out a house a few hundred feet away. Talk about "the kindness of strangers", that family allowed us to connect to their AC power! So we ran several extension cords through the woods and pretty soon the session was getting under way.
We had two electric guitars, an electric bass and a set of drums set up in the front of the church. As the band warmed up, I walked around the space in order to find the best balance between the direct sound from the instruments and the ambience of the church. Near the first row of pews, I set up a mic stand with a stereo mic bar on it that would hold the mics 15" apart. I put the matched pair of Earthworks QTC1 omnidirectional microphones on the stereo bar and between the mics, I attached what a friend of mine dubbed the "Diament disk", actually my variation of the Jecklin disk, a circular absorbent baffle, to the center attachment on the stand. Mic cables fed two of the phantom powered inputs on a Metric Halo MIO 2882+dsp which served as mic preamp and A-D converter. The Firewire output from the MIO fed my PowerBook hard drive.
Signals were converted to digital at a 24 bit word length and a sampling rate of 96k. This would maintain the highest quality for possible use on the PCM tracks of a 24/96 "Digital Audio Disk"1 at some point in the future and still allow for the best possible transfer to the 16 bit, 44.1k standard used for CD.
We ended the session after about eight hours, anxious to get back to the listening room to hear the results of our efforts. The monitoring system revealed the concept had been proven. One can indeed record a rock band direct to stereo. The drums sounded BIG, filling the space with their power. The bass drum had that "in your chest" feeling I'd long been seeking. The electric bass had a snap and precision of pitch I haven't heard on a rock record before and the electric guitars, well, I've always heard it while in their presence but I've never heard the sound of an electric guitar and its amp this way on a record before. There was a "bite" anyone familiar with the real sound knows well but in my experience has never before been preserved on a recording.
In the end however, this was after all a first shot at the concept as well as being the first time I'd recorded in this particular location. Next time in this room, I'd experiment with placing the mics just slightly closer to the plane described by the front of the drum set and the guitar and bass amps that flanked it. A bit more of the direct sound in such an ambient locale would probably provide even more or the visceral impact we found so enjoyable when listening to the playback. Also, in the absence of the Saturday rehearsal time, the music, though full of good feeling, ended up being less than we believe it could have been. This only left us inspired to do this again and plans are already under way to capture performances we'll be proud to release on a distributed recording2.
So far, I've talked about a purist approach, recording an entire performance live, in real time. What about those occasions when all the musicians can't be present simultaneously or when one person plays more than a single musical part? What about those kinds of performances that can't happen "live"? The advantages offered by multitrack recording can still be applied using the technique of a stereo microphone array.
In Part 1 of this article, I mentioned a "better way" to multitrack than using close up mic positioning to achieve isolation. The distances mics are commonly placed from sound sources drastically alters what they hear. To my ears, tonality, harmonics, bass and dynamics are altered radically and this is where a lot of the work comes in later in trying to make it all sound right again. I fully understand the reasoning behind attempting to isolate everything this way but personally, I find the sonic cost too high. The alternative is to isolate in time rather than in space.
The first requirement is a good sounding space in which to make the recording. It doesn't have to be the best room in the world and doesn't have to be auditorium sized. It just has to sound good when a band plays music there. Many rooms can often be helped quite a bit by placing a liberal amount of what Harry Olsen called "functional sound absorbers" around the periphery, especially at the corners and early reflection points (where the sound will first bounce off the walls or ceiling). These can be either the commercial kind made by ASC (sold as Tube Traps) or the DIY (do it yourself) variety.
When the best part of the room for a band to set up and play has been found, the best place for the microphone array (always a stereo pair) can be determined. Listening tests will help find the place where there is a good balance between the direct sound from the instruments and the ambience of the room.
The next requirement is that you have to know ahead of time where on the stage you'll want all the musical parts to come from. This initial "mix" must be done prior to the actual recording. In a way, this is a reversal of the common record first, mix later method by which most records are made. We'll still have to mix later but that will only be to balance the musical parts, since we'll have already placed them on the stage.
Since we're going to record everything in stereo, each take will feed its own pair of stereo tracks. For each additional take, the microphone array remains where it is but the feeds from the mics are moved to a new pair of tracks. All fader pairs on the mixer are panned full left and full right.
Let's look at an example: In a tune with vocal, background vocals, two guitars, a bass and a synth, I might want the vocalist centered, the background vocals in the background (imagine that!), one guitar on the left and the other on the right. The bass might be only slightly to the right of center, but behind the vocalist. I'll place the synth on the left, behind the first guitar. To achieve this, I'd record one guitar first. The player will be standing on the left as seen from the mic array position. We want to leave room for the vocalist to stand in the center, in order to create a convincing image of the band, so the guitarist will have to imagine the vocalist is really standing next to them in order to leave enough room. This will be true for all subsequent parts; the players must not overlap with (that is, they mustn't occupy the same space as) previously recorded players on the soundstage.
When the first guitar part is done, remember, the mics remain where they are but I'll move their feeds to two new tracks. Now, I'll record the second guitar, this time with the player standing on the right from the mics' point of view. Then on a third pair of tracks, the bassist can stand just off center but a bit behind where the vocalist will stand. On a fourth pair of tracks, I'll record the centered vocalist. On a fifth pair, I might have a couple of background singers standing in the background, a good distance back so they really sound that way. This provides a more convincing illusion of stage depth than trying to accomplish it by pulling their faders down in the final mix. Finally, we might overdub a synth. Since the goal is to capture the band in an "acoustic" environment, rather than use any direct feeds, all electronic instruments, synths included, are always recorded through a good sounding instrument amp. In this case, we'll place the synth amp on the left, just behind where the first guitar was.
We can even add sampled sounds to the record, drums for example, if there isn't a real drummer. For these too, no direct feeds are used. Instead, the sounds are fed to a good pair of loudspeakers and recorded via the microphone pair. This puts the sampled sounds in the same room as the rest of the instruments and can have the added benefit of making them harder to identify as samples. The only change here is that since we are recording from two discreet sources (the speakers) instead of one, there is no real sound in the center, so I might put the mics on separate stands, spaced the same distance apart as the speakers and place each about 2 feet in front of its respective speaker. The "phantom" center will be restored upon playback.
For the mix, all that is needed is to push up pairs of faders to get the right balance. Ideally, the record level is set for the loudest sounds and not touched for the lower level parts. This makes the mix easier and also more convincingly "real". When a good pair of mics is used (I tend to favor a matched pair of Earthworks QTC1s but have heard good results with other good omnis as well3 ) I've found no desire for EQ either. Just push up the faders for the balance and everything just gels beautifully. Even if some direct feed, special effects parts are added, the fundamentals are captured in a way that makes the result quite convincing.
Wonderful things can happen when a recording is made in stereo.