Trends in recording come and go, but acoustic music is a constant that can be relied on to sustain and challenge all engineers. From miking to mixing, successfully capturing an acoustic ensemble in the studio always puts one's engineering skills to the test.
Having engineered a wide range of acoustic groups, I can say that the techniques necessary for a successful outcome can vary greatly from project to project, based on factors such as musical style, group composition, and budget. I'll delve into specific examples of certain projects that I've worked on, which will show a variety of engineering approaches. But first I'll give an overview of some of the key issues and challenges you'll face in almost any acoustic-recording situation.
When setting up for an acoustic session, my first concern is always for the comfort and relaxation of the players. Most acoustic musicians need to play near to each other, so that they can hear and see subtle musical cues and can feel supported by and involved with the ensemble. Engineers, on the other hand, prefer to separate sound sources in order to retain maximum control over the individual instruments in the mix.
After more than 20 years of recording, I've learned that putting the musicians' needs first always increases my chances of capturing an inspired, connected performance. Certainly, technical compromises may result, but ultimately there is no point in pushing for a technically “perfect” recording that ends up feeling stiff or unmusical. I like to start with a tight circular arrangement in an acoustically appropriate room. My objective is to create an environment that is as similar to an informal living-room jam session as possible.
With this goal in mind, I don't automatically make musicians play with headphones on sessions of this type. My experience has shown that musicianship thrives when the musicians in an acoustic ensemble find their comfort zone and play at an appropriate level for the room. More often than not, the elusive qualities of blend, acoustic balance, and solo volume go out the window when a group disconnects from the room and becomes dependent on headphone monitoring.
On the other hand, if hearing problems in the studio can't be solved acoustically — by repositioning performers, changing the room's characteristics, or suggesting a different playing dynamic — I'll gladly hand out headphones to anyone who wants them. For example, a kit drummer or percussionist will rarely be able to hear an unamplified acoustic bassist or guitarist clearly in the controlled acoustics of my studio. In that case, headphones will enhance the recording by allowing the drummer to play at his or her desired performance volume.
Using quality microphones and giving careful attention to placement are essential in a tracking session, especially when recording natural acoustic sounds. Besides capturing accurate, pleasing sounds on the intended instruments, it's important to consider how every mic on an acoustic session deals with off-axis sound, also known as bleed or leakage. Typically, there will be noticeable coloration in the timbre of signals picked up at the sides and rear of a unidirectional microphone, and this is a particularly crucial factor when you have a number of instruments in the same room.
Acoustic ensembles come in many sizes and varieties, and you have to adjust your technique to fit the style and vibe of a given group. For example, a mandolin-guitar-bass lineup performing old-time music in the style of the popular O Brother Where Art Thou soundtrack will need to be tracked and mixed differently than the same combo recording high-tech “new-acoustic” music or modern bluegrass. Similarly, the production values of a singer-songwriter combo will vary drastically depending on whether they are aiming for pop-radio airplay or pursuing careers as Chicago-blues revivalists.
I've found that in general, older genres require special attention to minimalist miking techniques and capturing a good overall blend. Conversely, contemporary acoustic-music styles may rely heavily on overdubbing, separation, effects, and other production values that are standard practice in the rock world. Therefore, the best advice I can give as far as recording a specific group is to do your homework and listen to some recorded examples before venturing into unfamiliar production territory.
TO PUNCH OR NOT TO PUNCH
When recording any live ensemble, questions about fixing parts will inevitably arise. In the case of acoustic groups, there is no set answer that an engineer can give, especially when recording everyone in the same room. If the question is “Can I rerecord that solo?” it is useful to remind the group that, because of the leakage that is inherent in the situation, it will likely be faster to record a fresh take of the entire song with the whole group, rather than to punch-in a replacement solo. For something such as a missed guitar chord, bass note, or clarinet squeak, my answer is usually, “I'll try.”
It's most efficient to do a punch-in fix immediately after a take, before the musicians change their mic positions. That way, if a recorded “ghost” of the mistake is still audible on other tracks or if the punch-in or punch-out doesn't work, the group will be most amenable to doing another take right away. Taking detailed notes about mic selection, placement, preamps, and so on will also make it possible to attempt fixes at a later date, when you can focus your concentration on repairing flaws in an otherwise satisfactory take.
Overdubbing is often a great way to deal with vocals, solos, and tricky passages that might otherwise slow the momentum of an ensemble date. Scratch tracks of the aforementioned can be performed and recorded along with the ensemble, as long as the musician in question is acoustically isolated. But normally that entails everyone using headphones, which can make players hypercritical or can skew the group dynamics. In my experience, it is always best to persuade players in this situation to slip one headphone cup off their ear, so that at least half of their hearing is trained on the balance in the room.
When it comes time to overdub replacement tracks or to add entirely new musical layers onto an acoustic music production, don't forget about the importance of ambience. It can be distracting to hear a natural, roomy ensemble with one foreground vocal or instrument that is obviously overlaid or isolated in a different (usually deader) room.
If your intent is to obtain a unified blend of all the mix elements, make your overdub merge into the whole using mic placement rather than added effects. Record the overdub in the same space that the ensemble used, rather than in an iso booth. And don't be afraid to add extra room sound to the track by using a more distant mic position, using an omnidirectional or bidirectional (figure-8) mic, or recording a second room mic. If you can get the track to fit in with the overall ambience right off the bat, that's one less challenge during mixdown.
For those musicians who are working alone or who are layering many instruments one by one, similar principles can be used to create a colorful ensemble sound on a track-by-track basis. Incorporate a room ambience you like into your foundation track — be it percussion, acoustic guitar, or whatever — and make it a point to match or complement that room sound on each successive overdub.
Again, using a separate room mic for each track is an effective technique that allows for many creative possibilities in the mix. Excellent examples of this approach can be heard in the recent records of Tom Waits. When working as a session guitarist on Waits's Blood Money (Anti Records, 2002), I was delighted to find that my guitar, mandolin, and banjo parts were being picked up simultaneously by any combination of mics set up around the room, including an old boom box that was wired directly into the console.
STANCH THE BLEEDING
In live-ensemble recording scenarios, leakage is frequently a fact of life. It's always best to deal with excessive bleed at the source, by using proper mic placement and pickup pattern control. Getting a tight, focused sound when tracking guarantees that when you're mixing, raising a single fader will boost the desired instrument without distorting or washing out any other source in the room.
After mixing a few acoustic sessions, the ear begins to discriminate between good bleed (which adds depth, liveness, and richness of tone to the ensemble) and bad bleed (which is usually a concentration of muddy midrange frequencies, indistinct bass, or phasey high end).
The most important rules in this regard are (1) don't automatically reach for the EQ control just because you hear leakage on a track, and (2) don't make leakage even worse by adding too much compression.
If you feel relatively confident that your tracks and overall ensemble character are in the desired ballpark, I recommend that you start your mix by listening to all the tracks together without EQ or effects. That will help you to avoid the common mistake of analyzing and tweaking each track separately — a common practice that can work to the detriment of overall warmth and clarity. Once you have the ensemble sound established, it will be easier to evaluate and refine the effect of EQ and bleed characteristics in your mix by muting tracks one by one, rather than listening to individual channels isolated from the whole.
TALES FROM THE STUDIO
To provide some additional perspective, the following are descriptions of four acoustic-music dates that I recently engineered. As you'll see, the engineering strategies that I used varied greatly depending on the musical style and circumstances.
I invited this roots-oriented, harmony-singing country band to Guerrilla Recording to help in the testing for my EM review of the Royer Labs R-122 active ribbon microphone (see the January 2003 issue). My goal on this session was to use the figure-8 ribbon mics in a Blumlein configuration, placed head-to-head with the left and right mics oriented about 90 degrees off-axis from each other. (For a further explanation of Blumlein, XY, and other stereo-miking techniques, see “More Than the Sum” in the June 2003 issue of EM; text available at www.emusician.com).
Recording was done in the purist mode, with the mic preamps connected directly to a 2-track DAT. In this setup, there was no option for equalization, reverb, or panning before going to tape; all of these recording esthetics had to be addressed solely with mic placement. The goal I was anticipating was a smooth and natural old-school sound, with three-dimensional depth, rich room ambiance, and clear imaging.
Ultimately I was able to achieve the desired result, but only after a lot of trial and error. Because at least three of the group members sing and switch around between guitar, mandolin, and accordion, each of the songs required a lengthy sound check. During these sound checks (which doubled as rehearsals) plenty of adjustment was needed to create the desired balance relative to the mic pair.
This acoustical balancing entailed physically moving the performers, asking them to change their singing or instrument volume, and accounting for whether they played standing or sitting. Then there were the usual performance issues: maintaining the balance and dynamics, keeping everyone stationary, and getting a satisfactory performance. There's no punching-in or overdubbing in this format!
An additional issue was the room itself, which became a major factor in the overall sound of the recording. Initially I set up the group in my drum room, figuring that the wood floor and hard surfaces there would contribute a lively quality to this ribbon-mic recording. But upon playback, it became apparent that the distance of the performers from the mic pair — at least four feet as I remember — contributed too much room sound. Moving the entire group and the mic pair to the larger and deader carpeted lounge produced a better sound, with less room coloration and more detail.
Although it was a lot of work, and all the repositioning bogged down the flow of the session, everyone was pleased with the results. The group released these recordings as their first demo, and I would not hesitate to try this approach again for bluegrass, old-timey, blues, and other roots-oriented and ethnic styles as well. Of course, two-mic stereo recording is also a standard technique for small classical and jazz ensembles in concert.
Because there is no mixing after the fact, this method can save both time and money. And this live approach to studio recording is a refreshing alternative to the norm of manipulated multitrack recordings. Of course, a good-sounding room and a quality mic pair are essential ingredients for the success of this technique, as are a well-rehearsed group and plenty of patience on everyone's part.
Carla Kihlstedt and Two Foot Yard
Room sound also played a significant part in the recorded esthetic of Two Foot Yard, violinist-vocalist Carla Kihlstedt's trio. This group, which mostly tracked violin, cello, and drums live, ran the gamut from delicate chamber music to raw rock numbers and unpredictable improvisation.
At the outset of the recording process, I decided to set up this three-piece group in the Guerrilla Recording drum room, where the strings and drums would sound their best. I also felt that the physical proximity of this arrangement would aid their intuitive communication. But the thorny issue of recording drum kit and violin together in a live, bright room remained.
Fortunately Kihlstedt is a forceful violinist with magnificent tone, and drummer Shahzad Ismaily tends to play relatively quietly. The Royer R-122 ribbon mic, which I used on the violin, sounded great, and I was able to place it pretty close to the source (within a foot above the bow in most cases). Its tight pattern yielded significant drum rejection when positioned with its off-axis side fins aimed at the kit.
The use and careful placement of cardioid condenser mics on the cello, combined with a bit of baffling, also contributed to the success of this method. On medium-intensity numbers, the drum leakage onto the violin and cello mics contributed an interesting, roomy sound, which I hear as one of the CD's most intriguing facets.
In some cases, the off-mic drum coloration on the Royer was a bit too midrange-heavy or boxy, as it can be on a dedicated drum-room mic. In these instances, a judicious midrange cut between 350 and 750 Hz tamed the problem without negatively affecting the string timbre. As expected, for Two Foot Yard's more aggressive rock songs, isolation of the violin and cello was required to keep those instruments from being overwhelmed by pounding snare and cymbal wash. On these pieces, a separate drum-room mic was used in the mix to capture the room sound and maintain consistency in the drum tones.
For the classically influenced pieces, on violin and cello only, I turned the XY drum overhead mics around to face the strings, and printed their stereo signal as an ambient pair to fill out the close-miked string sound. On one song, this process was repeated three times to create a faux violin trio!
For the first of the three scored parts, Kihlstedt stood in her usual position in the room, on the right relative to the stereo mic pair. For the next track I moved her, along with her close mic, to the center position, and then she went over to the left side for the final pass (see Fig. 1). With careful balancing and panning of the three close-mic tracks, and the three pairs of stereo ambiance, a very credible approximation of a live trio was created.
After the basic tracking was done to 16-track analog tape, overdubs were added to most of the compositions. Because very little of the string sound bled onto the drum mics, some fixing was able to be done on the original violin and cello tracks. Kihlstedt's vocals were not cut live, except in the case of songs in which the feel of live performance was essential, or when the timing of the violin and vocal was so intricately interlocked that it would have been impossible to overdub.
In the final assessment, the decision to cut as many tracks as possible in the same room paid off for all involved. Although the CD is very stylistically eclectic, it has a unified organic feel and a unique sonic signature. These would have been difficult to achieve had greater separation been pursued from the outset.
When I was originally contacted by Bon Singer — leader of this vocal-oriented world-music group — to record their CD, several scenarios were discussed. For cost-effectiveness and authentic concert-hall ambiance, we determined that location recording was the best option. After some searching around, several tracking days were booked at Temple Beth-El in Richmond, California.
In the initial planning session, several factors had to be addressed. In addition to my production concerns and the physical setup, notes on mic and preamp selection, additional gear, and other considerations also had to be relayed to remote specialist Karen Stackpole of Stray Dog Recording. After the initial session, Karen would be working alone to bring in the equipment and record the basic tracking sessions to eight tracks of ADAT.
My strategy was to arrange the group's six female singers in a tight half-circle and record them on a stereo mic pair, with one spot mic for soloists if needed. Although I have recorded vocal combos of this size with individual mics for each performer, I have always found that the blend and balance of the voices is more pleasing when captured with a phase-coherent coincident XY pair. Ya Elah also has three instrumental accompanists who play vielle (the ancestor to the viola), oud, and hand percussion. With close mics on these instruments, and a stereo pair of room mics to capture the Temple's ample reverberation, all eight ADAT tracks were spoken for.
Temple Beth-El offered a large multipurpose room with a raised stage, uncarpeted floors, and no fixed seating. The openness of the space allowed us to move the group around and explore some different setup possibilities. Initially, various group arrangements — with the singers on the carpeted stage and the musicians down on the floor — resulted in hearing problems between the two factions. Putting the entire group on the stage was also problematic.
Rather than settle for a flawed setup or add the complications of a monitoring system, I got creative and scanned the room for more possibilities. We ended up arranging everyone on the bare floor at the back of the room, in front of a sliding, segmented room divider. That area of the room was much more live, which solved everyone's hearing difficulties and put the singers more at ease. Adding a few carpets under the vocalist's half-circle and slightly opening the room divider behind the singers eliminated most of the close reflections, making the space more conducive to favorable recording.
With those arrangements set and with the group well rehearsed, the first recording session ran smoothly. A few trial takes were necessary to perfect the acoustic balance of the singers, and strips of tape were then placed in front of their feet to keep them in position. The musicians, who were seated facing the choir, also had to be kept as close as possible to their mics, since leakage from the robust vocals was significant. Once those details were tightened up, it was simply a matter of running pieces in the hope of capturing that magical take.
Rough mixes created at the first session sounded great to Karen Stackpole and to me. And although the project is currently on hold, it's an exciting prospect to think about mixing the tracks, considering their big-room sound and natural reverb. Obviously the preplanning, gear transport, and coordination of remote sessions is labor intensive when compared with the ease of booking a studio. But this approach was necessary to make to the project work within the given budget. More importantly, it gave the singers the chance to shine in a space that was acoustically supportive and appropriate to the music.
Recording a jazz session entails many elements that are similar to the other types of acoustic sessions discussed here. It is essential that everyone is comfortable, close together, and connected visually and musically so that the communication can flow unimpeded. A less obvious, but equally important point for any acoustic music with drums is that the ensemble has to play at a level that works for that particular room.
Although the music of the five-piece group Collective Amnesia is decidedly modern, I envisioned a warm, classic, Blue Note sound to mirror their standard all-acoustic lineup of drums, bass, piano, trumpet, and sax. We chose to record the project at Bay Records in Berkeley, California, a wood-floored studio with a fine Yamaha grand piano, and I brought a selection of my favorite tube condenser mics. Recording was done to 16 tracks of ADAT, in order to keep tape costs low and allow flexibility in the choice of a mixdown facility.
The members of the group set up in a circle, fairly close to one another. Because pianist Leonard Thompson plays with a soft touch, special attention had to be paid to minimizing leakage of the drums and horns into the omnidirectional Schoeps mics that I'd placed inside the Yamaha grand. Rolling the piano about ten feet away from the rest of the group helped, as did some carefully placed blankets and baffles for the piano and drums. I achieved additional separation by positioning the instrument so that the slightly opened lid blocked any direct sound from the other instruments.
The horns were close-miked, and the bass had one mic on the tailpiece (see the sidebar “Upright Advice” for details of the miking method). To control drum bleed, I asked drummer Bryan Bowman to play more quietly, and made sure that he had the right headphone mix to accomplish that. Once the group found its balance in the room, the overall dynamics, ensemble balance, and solo levels snapped into focus, with very gratifying recorded results. When listening to playbacks, I remember that I rarely had to move any faders. But as soon as the drums got too loud, the mix became a struggle, with the bass and piano getting drowned out and the cymbals becoming diffused and washy.
Because of scheduling difficulties, I was unable to be present for the mix of this project. But I did master the Collective Amnesia CD and was very pleased to hear that the band, producer Justin Morell, and mix engineer Jon Evans remained faithful to the straightforward ensemble sounds that were captured during tracking.
Myles Boisenis a guitarist, producer, composer, and head engineer/instructor at Guerrilla Recording and The Headless Buddha Mastering Lab in Oakland, California. He can be reached email@example.com.
RECORDING SINGING GUITARISTS
Sooner or later, anyone who records acoustic music has to deal with the challenge of recording a vocalist who plays acoustic guitar. For tracking purposes, seasoned musicians are usually able to overdub to a scratch guitar or vocal track. There are those instances, however, in which the performer's needs — perhaps influenced by inexperience or irregular timing — dictate that the vocal and guitar tracks be recorded simultaneously. On such occasions, it is good to have a few miking tricks up your sleeve.
Although it is possible to mic a singing guitarist with a single microphone, let's start this discussion with a minimal two-mic setup. As always, leakage and phase cancellation between the two sound sources is a crucial factor. Most engineers have dealt with this common problem before. Typically, the vocal or guitar will sound fine alone. But once the second mic is brought into the mix, unpredictable tonal distortion dominates the sound, changing as the faders are moved.
To minimize those problems, start by miking the guitar using either an omnidirectional mic by the bridge or sound hole, or a figure-8 mic in front of the neck joint aimed at the sound hole, with the off-axis side or “null” of the pickup pattern carefully oriented at the singer's mouth. In an acoustically dead room, the figure-8 guitar mic should do a very good job of rejecting vocal sound, especially with a quiet vocalist. Conversely, the omni mic will not reject much vocal sound, but since it has no proximity effect, it can be placed very close to the guitar without boominess.
To pick up the vocals, start by using your favorite vocal mic, positioned a bit higher and farther away from the mouth than usual (see Fig. A). Trying to get the vocalist to sing with his or her chin up toward the mic, and head turned somewhat to the left (for a right-handed guitarist). Such placement increases the distance between the two mics, and prevents the singer from projecting toward the guitar mic. When placed properly, the vocal mic will add sparkly guitar harmonics to the mix, and midrange boxiness from the guitar will be minimal.
If you want a more elaborate setup, try using a pickup-equipped acoustic and recording its output to supplement the signal picked up by the guitar mic. Or try replacing the omni or figure-8 guitar mic with a pair of cardioids in an XY configuration. With such a pair, you'll need to experiment in order to maximize rejection of the voice.
Here's a trick to increase separation when recording the acoustic bass, particularly in a jazz or a rock setting and with a drum kit. Select a dynamic mic that has good low-end response (my favorite is the Sennheiser MD 441 because of its supercardioid pattern and onboard EQ), wrap it in a towel or a piece of foam being careful not to block the mic's grille or acoustic vents, and wedge it in the tailpiece of the bass.
The mic should be pointing up and positioned with its diaphragm just below the opening between the legs of the bridge (see Fig. B). This ingenious technique, which was shared by veteran German bassist Peter Kowald during his recording with Damon Smith of the bass-duo CD Mirrors — Broken But No Dust (Balance Point Acoustics, 2001), yields the most favorable ratio of bass to room sound. In addition, the bassist can move freely without any change in timbre or level.
BANDS ON THE WEB
Features CD and MP3 downloads, as well as performance schedules.
Information about Carla Kihlstedt and Two Foot Yard. Order the CD Two Foot Yard (Tzadik, 2003).
Information, gig schedules, photos, and MP3 downloads are available.
Ya Elah concert information, contacts, links, bios, and more.