Recording a Jazz Combo in the Kevin Brown Live Room

A couple of weeks ago, I had the distinct pleasure of tracking a jazz combo consisting of Western Michigan University jazz department graduate students Meghan Staghl (vocals), Mark Niskanen (keyboards), and Jordan Otto (drums) as well as local pro Henry Rensch (bass). They were performing an intimate setting of

“Watch What Happens,” a standard first released by Tony Bennett in 1965.Hummel 2

For this session, I kept everyone together in the Kevin Brown Live Room. When drums or upright bass are involved, some groups prefer the sound of having them isolated, since drums tend to be loud in the room and will bleed into other microphones while the comparatively quiet upright bass will often be covered by bleed from other instruments, even if the microphones are close to it. Jazz ensembles, on the other hand, tend to prefer having everyone in one room. Because jazz is heavily improvised, visual communication between musicians is very important to the way they interact with each other. Having the drummer and bass player off in separate rooms can restrict or eliminate that ability.

Every sound source in this ensemble was acoustic except for the keyboards. Mark brought along his own vintage Fender Rhodes, which was recorded direct. We also decided to run it through a small tube guitar combo amp miced with a Shure SM57 to get a couple of different tones to work with. Having the extra input allowed the Rhodes to be mixed in stereo, since the output from vintage keyboards like this one tend to be mono. Jordan’s jazz kit consisted of a non-ported kick drum, a snare, two toms, a high hat, and a ride cymbal. I decided to experiment by micing the kick drum with a Cascade Fathead ribbon, while the overheads were handled by a spaced pair of AKG 451Bs. For the snare, I used a Miktek C5. Since Jordan was planning to use brushes for most of the song, I wanted the small diaphragm condenser’s ability to capture high-frequency details. Henry had a pickup on his bass which was taken direct, and I added a Neumann TLM 49 mic on the f-hole and a Royer 121 on the bridge for detailed acoustic pickup of both the body resonance and string noise. Meghan has a rich alto voice, which I thought would be complemented by the clean, bright Mojave MA-300 in cardioid run through an Avalon AD 2022 tube preamp for a little high-end coloration.

Hummel 6My first objective was to try Jordan’s kit in front of each of the walls facing the control room, since one is covered in reflective wood paneling and the other in absorptive acoustic insulation. For our first setup, I had Jordan place his kit in front of the wood-paneled wall and then spread the musicians into the corners of the room to reduce bleed as much as possible. Meghan was on Jordan’s left, Henry was on his right, and Mark sat directly across the room from him. Since Mark’s Rhodes was being DI’d and going through a close-miced amp, he could sit anywhere without it affecting the tone of his instrument. Henry and Meghan, however, were being picked up by figure-eight patterned mics that are prone to capturing a lot of room sound, so extra care had to be taken to place them where the null point (on the side of the microphone) could face the drums. I situated Henry between Mark and Jordan, since bass players on stage will most often stand between the keyboard player and drummer, and are usually the most comfortable in that arrangement.

A few problems immediately came to light in this placement. First, the drums sounded like they were being heard at the end of a long tunnel. Lowering the overheads until they were only a foot or so above the cymbals helped considerably, but didn’t completely alleviate the washy room sound. Jordan also switched to a less resonant ride cymbal which complimented the room sound much better. Since the soloed drum mics were sounding good at this point, it seemed that most of the reverberant sound was coming from bleed in the other musician’s microphones. In particular, the Royer 121 turned out to be picking up more drums than bass due to its rear lobe sensitivity. Removing it from the mix cleaned up the drums, but it also removed some clarity from the bass, so a new position would have to be found for Henry where better isolation could be obtained. Mark’s Rhodes was too loud in the room because of the amp, and since the musicians could all hear it fine in their headphones, I decided to turn the amp down at the expense of having a worse signal to noise ratio from it.

Hummel 1I decided between takes to try a Cascade Fathead on Meghan’s voice, since it seemed a little thin with the MA-300. The musicians all agreed immediately that it had a “meatier” sound, and decided to go with it. Her voice was now a tad too colored, especially in the top end, but switching from the Avalon AD 2022 tube preamp to a Millenia HV-3C solid state preamp cleaned it up nicely and also added some extra clean gain to make up for the ribbon mic’s low output.

At this point we shifted to a new orientation, starting with Jordan against the acoustically treated wall. Mark moved to stay directly across from Jordan, but this time Henry was on Jordan’s left and Meghan was on his right. I spent a good amount of time carefully positioning the bass and vocal mics to obtain maximum rejection of the drums, and I also decided to tuck Mark’s amp into the corner behind him, facing the wall, to reduce bleed even further. At this point I was pretty happy with all of the sounds except for the drums. While the treated wall saw a small improvement in room splash over the wooden paneled wall, the same issues with bleed affecting the drum tone still arose. Even though the soloed drum mics sounded great, as soon as the bass, vocal, and keyboard amp mics were added in the drum sound became unfocused, and the natural reverb tail was too long to be appropriate for the intimate song that they were performing.

For the next and final orientation, I tried out a theory explained to me by Bryan Heany of Western Sound Studios. When dealing with bleed that is unavoidable, something must be done to make that bleed useful instead of destructive. Since audio travels rather slowly (about one foot per millisecond), and humans can hear timing differences as small as ten milliseconds, a distance as small as ten feet between sources means that the delay between direct sound from one microphone and bleed into the other microphones will be audible, making the sound come across as smeared or more reverberant. Since my musicians were between fifteen and twenty feet away from each other, I decided to try and compact them as much as possible in the center of the floor. In doing so, I sacrificed some level of isolation (the problematic Royer 121 proved to be unusable after all), but the resultant drum bleed was much less distracting, less “roomy,” and generally served the aesthetic of the track better. More importantly, all four musicians said that they felt more comfortable in this arrangement than they had in any of the others because the tighter proximity aided in communication between them. That’s the most important thing of all and, to me, says that my experiment was successful.

Hummel 5In the future, I would argue that it is possible to get a good sounding recording with a full ensemble together in the Kevin Brown Live Room with no isolation between group members. It certainly requires careful placement of musicians and gear in order to ensure that bleed doesn’t get out of hand, but the reward is that the musicians feel more comfortable and, in return, deliver a more natural performance. While the sound of a large room wasn’t especially suited to the recording that I was doing on this session, a different ensemble may enjoy the sound of the room, making placement decisions significantly easier. As for drum placements, the acoustically treated wall seemed to work better at controlling excess bleed into other microphones than the wood paneled wall, although the sound difference in the drum mics themselves was rather small. As it turns out, for maximum reduction of splashy bleed, placing the drummer as close as possible to sources that are more likely to suffer from high amounts of bleed can actually reduce the distracting effects of acoustic delay when mixing, despite how counterintuitive it seems.

-Zach Hummel

2015 Overneath Creative Collective Intern