The band I’m currently working with, The Mercy Kit, recorded and produced its first album One Big Soul from July 2019 through March 2020. It is currently on most streaming services - Spotify, Amazon, Tidal, Apple Tunes, etc. This is a blog detailing how most of it was created.
3 of the tunes on the album had been produced between 2017-19 and are not within the scope of this discussion. We used what we learned from our experience with those 3 to record 9 more tunes only 7 of which made the cut this time. The 2 that were cut from this album have been pushed back to be part of the second album which we are hoping to start work on by the end of 2020.
RECORDING
Environment
We recorded the 9 tracks in my home studio, a 220 sq ft building at the back of my property, dimensions of 9’ x 11’ with a vaulted ceiling, walls treated with acoustic panels and foam traps in the corners, laminated wood floors with spot carpets when needed. The drummer placed a carpet under his kit and I used a couple of throw carpets under the mic stand when recording vocals to help prevent rumbling of the stand.
We used Cubase 10 on an iMac, an Apollo x8p with Unison preamps and a UAD API Console strip in the insert with a HPF set for applicable instruments - i.e. snare, vocals, guitar.
The x8p has 8 XLR/line inputs each of which can be configured with UAD software preamps. For the drum tracks I patched in a MOTU 8pre to the x8p via lightpipe to get one more XLR insert and preamp.
Every session after the initial scratch tracks was recorded in sequence over existing tracks. The first tracks were comprised of the composer (Matt Heller or I) laying down the tempo and the structure of the tune recording a scratch vocal and a scratch acoustic rhythm guitar part using a click track. In the subsequent overdubbing only the drummer used the click throughout the entire song. The other players used the drum tracks as timekeeper on their overdub sessions except for lead ins and in a couple of passages in songs where the beat was temporarily played as legato.
Scratch vocal was recorded on a SM58 and the scratch guitar signal came from the output of the acoustic guitar’s electronic jack out direct into the Apollo.
Each signal went through a UAD Apollo x8p using Unison preamps with an API Console as the insert.
We bounced out each song along with the click track for the drummer, bass, and the lead guitarist to rehearse with and compose their parts.
One Big Soul
- Sanctuary (M Heller)
- **The Devil’s Highway (M Heller)
- Hurricane Nora (M Heller)
- **Sasabe Air Show (M Heller)
- Salvation (M Heller)
- **The Dover Test (M Heller)
- Open Road (M Heller)
- Unintended Consequences (Kevin McCarthy)
- Defective Heart (M Heller)
- Streamline (M Heller)
**not in scope of this discussion
Lead Vocal: Matt Heller on all tunes except Unintended Consequences which was Kevin McCarthy
Backup Vocals: Kevin McCarthy except Unintended Consequences which were Matt Heller and Kevin McCarthy
Acoustic Guitar: Matt Heller
Electric Guitar (Rhythm): Matt Heller and Tim Peck
Electric Guitar (Lead): Tim Peck
Lap Steel: Tim Peck
Piano and Organ: Kevin McCarthy
Bass: Kevin McCarthy
Drums and Percussion: Bill Stader
All songs arranged by The Mercy Kit
Recording, Mixing, and Mastering at kevinmccarthystudios in Richmond, CA
Recording Engineer: Kevin McCarthy
Mixing and Mastering: Kevin McCarthy
Microphones:
DRUMS:
Kick - open
Mic: D112 dynamic mic -> UAD Apollo x8p input
Placement: halfway into the shell pointed roughly at the beater.
Snare - above and below
Mics: SM57 dynamic mics -> UAD Apollo x8p inputs
Placement: one above coming in from the side at ~25° pointed to the middle of the snare halfway over the shell one below pointed straight up under the snares 6" below them and set at 180° to the other mic signal on record.
Record: HPF at 50Hz
HiHat -
Mic: AKG P170 small condensor mic -> UAD Apollo x8p input
Placement: pointed down from above, halfway between the edge and the bell, 6" above with the hat in open position
Record: HPF at 50Hz
Toms - rack and tom
Mics: Audix DClamp clip on mics -> UAD Apollo x8p inputs
Placement: on the high edge of the shell pointed towards the middle of each head
Overheads -
Mics: Rode NT5 small condensors -> UAD Apollo x8p inputs
Placement: Spaced Pair over each of the far crash cymbals ~20" above the bell of each
Room -
Mic: AT2020 -> MOTU 8pre input connected to the x8p via lightpipe
Placement: set directly in front of the kit about 6’ back at the height level of the drummer’s eyes pointed directly at the kit
The drummer set up his kit as he found most comfortable making sure that the perimeter of the kit was centered in the room as much as possible the edges equidistant from the walls. He tuned each drum beforehand.
To achieve the best placements, we recorded several alternatives from each mic setup separately to listen back for comparison. I would highly recommend this approach as there was no way to discern in real time how the signal was tracking and I wanted to keep the separate placements to serve as a blind shootout for the drummer when we listened back to get his objective feedback. For instance, in the case of the kick just a couple of inches of distance and direction made all the difference in the sound. And the snare mics were adjusted several times as well.
Mic gain was set to hit at about -6dB at the loudest to allow for the best amount of headroom. This parameter was also used for recording all of the following instruments.
Other than the HPF for the instruments that didn’t reach below 50Hz (vocal, guitar, cymbals, etc.) and the phase correction on the bottom mic for the snare, I didn’t employ any other processing on input.
Each mic had its own dedicated track including the L and R overhead mics as separate mono tracks. I find that to be better than recording the two mics into one stereo track. The monitoring for the drummer was mixed to give the drummer a stereo image that matched how he heard the kit as much as possible and included the scratch and click tracks in at the level he liked. Even though it was impossible to isolate the monitoring from the sound in the room, it still made a difference.
During recording, the drummer monitored the mix through headphones from a Behringer Powerplay Pro-8 headphone amp. We set a 5 bar preamble for each recording session, the last 4 with the click enabled to lead into the intro. For the drums we only recorded separate takes instead of using punch-ins to avoid the messiness of consolidating 9 separate tracks later. Since the drummer was very familiar with each tune from having performed them live for the past 3 years this worked out fine.
OVERDUBBING:
Once I had the 9 drum tracks the perspective I used from this point was one of just trying to use the sessions as a means to collect as many raw tracks as I could for the editing and mixing phases.
From each session, I separately bounced out each of the recorded tracks without processing as a raw file to be passed to the mixing phase.
I then created stems to be used as backing for subsequent tracking. These went through a rough insert chain with sends to sound reasonable.
A typical set of stems would be:
kick
drum kit
bass
rhythm guitars
keyboards
lead guitar
By the end of all of the tracking sessions each song had a collection of raw unprocessed tracks to be used in the mixdown stage.
Each tracking session was kept in its own dated directory (YYYY-MM-DD to perserve chronological ordering) and then further divided into instrument then into song. The raw tracks and the processed stems were bounced into separate directories under the song so that they could be located and accessed easily through the hierarchy that was built over time. Each time I would create a new Cubase project using a premade template to hold the previous tracks and the ones being recorded then bounced out to what I needed.
For efficiency among the various songs since they were all more or less the same players, instruments, and setups, I created presets for the stem mixes from the first tune that could be quickly employed for subsequent songs.
At the end of drums tracking each instrument in the kit had its own raw track, center panned, fader set to 0.
kick
snare above
snare below
hihat
rack tom
floor tom
overhead L
overhead R
room
BASS:
I played bass on all of the tunes. On some of the tunes I performed and recorded with the drummer to capture the performance vibe that we’ve established between us but in a couple cases I later went back once I started mixing and recorded new takes as when I isolated the sound of the bass it didn’t quite work in the end to my liking and I’m glad I did. I used both a Music Man Sting-Ray 5-string and a G&L M2500 both with round wound strings on different tracks depending upon what I liked. For each I went DI into the board split to a GK 800 amp to record two tracks that I later consolidated to one blending the two prior to mixing. I didn’t route the amp through speakers as the room was just too small to handle the long bass waves and it wasn’t necessary in the end anyway. I found it worked better to blend in amp modeling in the box at mixtime.
At the end of bass tracking each source had its own raw track, center panned, fader set to 0.
bass DI
bass GK
ACOUSTIC GUITAR:
We recorded this live in the room auditioning several different setups with various mic combinations again recording samples of variations to determine the best one. In the end the best sound came from using the Rode NT5s in a spaced pair arrangment - one mic pointed towards the 12 fret (where the neck meets the body) and the other pointed towards the widest part of the body both evenly 18" away from the guitar. The guitarist moved a little bit during performance but generally we got a good sound that didn’t need any phase correction later. A few finger squeaks popped out during a couple of the songs that I was able to tame later with Izotope RX7.
Mics: Rode NT5s - 1 at the 12th fret, 1 at the body
Record: HPF at 50Hz
At the end of acoustic guitar tracking each source had its own raw track, center panned, fader set to 0.
Ac gtr neck
Ac gtr body
ELECTRIC GUITAR - rhythm:
We recorded this using both DI post-pedals and an SM57 pointed close to the speaker ~1" away pointed at the cone halfway between the middle and the edge.
Mic: FC-387 8" to the speaker, SM57 8" to the speaker
Record: HPF at 50Hz
At the end of electric rhythm guitar tracking each source had its own raw track, center panned, fader set to 0.
Elec gtr DI
Elec gtr amp
ELECTRIC GUITAR - lead, supporting, lap steel:
We first tried using a combination of DI post-pedal and a miced amp but in the end the amp added nothing since the lead guitarist Tim got such amazing tones just through his various guitars, set ups, and incredible use of several units. The amp in the room just made it harder for him to hear what was being tracked. We considered isolating the amp in another room but the DI worked fine alone. Tim’s been a performing pro guitarist for 40 years so he really had it nailed. Basically I would roll tape (figuratively of course) and let him do his thing and just lay in wait to capture what I could while it was happening. He would lay down three or four parts, stop suddenly, then quickly ask for another take partway through a track. He always knew what was working and what wasn’t and would change quickly so I just had to be on my game to catch. He made it plain that was my job not his and I got damned good at it. I would track everything he would do even during the first listen through while he noodled. Some of these very first takes were gold so I would always tape everything he did. Once I got good at sewing these strands together during editing that approach gave us a lot of great material.
Record: HPF at 50Hz
At the end of electric lead guitar tracking the source had its own raw track, center panned, fader set to 0.
Elec gtr DI
KEYBOARDS - organ and piano
For this album we added organ and/or piano to most of the tunes although we haven’t ever used any in performances. I played all of these using primarily Kontakt’s Vintage Organ Hammond B3 for organ using its drawbars and Spectrasonics Keyscape’s LA Custom C7 - Stage for piano. I first recorded each as an instrument track to get the MIDI file which I would then clean up with quantizing and manual shifting of note placements, volume, and articulations. I would then bounce out to audio and use them as raw tracks.
At the end of each keyboard tracking the MIDI rendered to audio had its own raw track, center panned, fader set to 0.
VOCALS:
Lead
I placed the vocalist in the middle of the room and put the stand on throw carpets to isolate any rumble. Just for S&G I set up two mic stands one on either side of the vocalist almost to each wall and configured like a very tall T. I draped moving blankets folded over them to add just a bit of a baffle on either side of him. The mic used was a Lauten Atlantis FC-387 (large condensor) that has a variety of settings including polar patterns and voicing switches. We used a cardiod pattern and toggled through the three voicing options each of which has a different frequency response. Like with the drums and guitars we recorded samples of the singer through each option and settled on one after auditioning them. I placed the mic behind a pop filter at the level of the singer’s eyes and pointed down slightly at his mouth and just the slightest bit off center rather than pointing the capsule directly at his mouth just to save the direct pressure of his voice straight on. The vocalist monitored the previously recorded tracks in headphones and would record several takes with punch-ins where needed until we had enough to choose from during editing and mixing phases.
Mic: Lauten Atlantis FC-387
Record: HPF at 50Hz
Backing vocals:
These were recorded in basically the same manner except generally the singer would step back an extra few inches just to get a little more room into the track.
Record: HPF at 50Hz
PERCUSSION:
We recorded tambourine, guero, claves, and shakers into various songs. The drummer handled all of these in separate sessions again sitting/standing in the middle of the room. I tried several arrays of mics but mostly used the Rode NT5s in stereo placement about 1’ back and pointed towards the drummer and a SM57 a little further back pointed directly at the instrument and then picked and choosed among the signals during mixdown.
Record: HPF at 50Hz
At the end of percussion tracking each source had its own raw track, center panned, fader set to 0.
Percussion L
Percussion R
Percussion C
NOTE : Although The Mercy Kit is a working band and we had been rehearsing and performing on a regular basis, it was still tricky to get everyone in place for recording sessions. 32 sessions were done over a period of seven months in total getting people in when they were available. We signed up in October 2019 to perform for a club in March 2020 with the idea that it would serve as supporting the album’s release. Even with this head start we couldn’t get everything in the can until the last week of February. As a result this gave me a three week window to compile all the recorded tracks, mix and master them. Three of the tunes had been completed before these 9 tunes were proposed and settled on. Out of the 9, one tune wasn’t battle-tested enough to be eligible for this album and a second one, although completely recorded, just didn’t quite make the cut either. So that left 7 tunes. As with the tracking, the variables of all the songs were similar enough that I decided to establish presets for each instrument that could be used as a template to be tweaked as I progressed through the subsequent tunes and this paid off greatly. Some proved not to be applicable but most were and it saved a great deal of time over the long run.
MIXING
Once I had all the raw tracks I needed for a song, I would compile them under song titles for mixing using a Cubase template that had everything but the tracks. This included premade named folders, VCAs, reverb, amp, effects tracks, busses that would most likely be used. I then imported all the raw tracks that had been created for the song being mixed in the tempo and time signature along with the Markers track created during the original scratch recording to mark the structure of the tune. Once the tracks were all imported I set up the reference track that was selected to mix the song to within the Metric A/B insert in Cubase’s Control Room to use as a guide throughout the mixdown stage. All the tracks were moved to folders on import and all the VCA links were established.
I set up the Mix Bus created to hold an insert chain bussed from the tracks through to the Stereo Out mostly with an API Console, an ATR-102 tape and a disabled Izotope Maximixer to be enabled as a placeholder limiter once the drums and bass were in place. I set faders up on everything just to hear that everything was working and that all the routing was correct.
For each of the instruments recorded with multiple mics, I consolidated them and bounced them into one track.
For the snare mic tracks, I auditioned them soloed then found a blend I liked between them, bouncing to one track. Basically the above was 100% and the below was usually about 65%.
For the overheads I panned the two single tracks hard L and R respectively then bounced it out as a single track.
For the bass I auditioned the two sources as a blend like I did the snare mics then bounced it out as a single track. Typically it would wind up as 100% GK and 40-60% DI.
For the acoustic guitar I listened for the best blend and set them in mono mostly favoring the body source then bounced it out as a single track.
For the percussion I panned the two L and R mic tracks to a fairly wide studio then blended in the dynamic mic in the middle to where it sounded best then bounced it out as a single track.
Once I had single tracks for each instrument I basically followed the Mixing Foundations workflow by setting all the faders down and then concentrating on the drums and the bass first.
For the drums mix I panned the kit L and R from the drummer’s perspective, the hihat and floor tom positions matching the overheads perspective.
I set the kick with an insert chain, added reverb, added the bass, added its reverb, got gain staging started then worked up through the drums. From there I progressed to the percussion where applicable, chordal instruments (keys, rhythm guitars), up to lead guitar and then the vocals. Although all of the tracks had been recorded with an API Console strip in the Unison preamp I used it again except for the lead vocal. On the vocal I set the Manley Voxbox as the first insert and worked up the chain from there: saturation, reductive EQ, musical compression, technical compression, de-essing, additive EQ occasionally with effects added in for width and enhancements. These went through a Send queue of reverb, plates where applicable, amps where applicable and each had treatment to get sounding right.
All the drums went through a parallel process to add intensity and power. Some of the songs had vocals that went through a similar parallel process to punch them up.
As far as keeping CPU tamed throughout the mixing stage, in Cubase I use one of two options to handle tracks as the plugins they go through start to weigh down computer performance.
- Use Direct Offline Processing to create a series of offline edit files for each plugin processed through the insert chain that is referenced in place of the original track in the project.
- Render the track as an offline file also processed through the insert chain that effectively replaces the original track with a new referencing the offline file but keeps a copy of the Sends in the new track.
Both options are very effective but personally in my workflow I currently lean towards option #2. The original track is disabled by default. It keeps the insert chain intact but as it is disabled it no longer impacts the CPU. I move these disabled files into a separate folder in the project that I label DISABLED. If I ever decide to tweak or create a new insert chain for any of these disabled files, I disabled the newer one, resurrect the original, make changes then render again.
Another habit I’ve gotten into is once I am satisfied with an insert chain is to save it off as a preset within a hierarchy of Band/Album/Song/instrument. It comes in handy as a template when starting on a new track that has the same instrument and context later on.
Once I had what seemed to be a good mix, I would bounce out the stereo out and forward to the co-producer to get his feedback. Meanwhile I would take the bounce and listen to it in several environments - mono speaker, on room monitors, through a couple of different set of headphones, streamed from my phone through my car stereo after and before listening to professional productions of tunes within the genre just to get my ears calibrated to the environment.
Each song went through several iterations until they were passed by the both of us.
MASTERING
This didn’t take nearly as much time as mixing did but did take tremendous concentration so I had to wait until all the mixes were completed and my brain and ears had decompressed to a reasonable level again. On some of the tunes I used the same references as during mixing but on others I used some of my favorite general references instead. As much as I used Metric A/B during the mixing stages, this is where it became completely indispensable. All of the components came into play during mastering especially when aiming for the PSR and loudness values as well as Dynamics. Although the songs were going to several streaming services as well as to CD, I pretty much stuck to aiming at Spotify’s -14dB mark and was able to maintain consistency throughout the entire batch of tunes. I also reopened the mastering sessions I had used for the previous 3 tunes from before keeping the same insert chains they each used but recalibrated them to meet the same mark from with Metric A/B.
Wrap Up
All said it seems the entire work was a success. But one of the most important aspects of the entire endeavor seems to have been the limited time factor and drop deadline as I had to make decisions quickly without any fooling around at all. Once I started I first charted out the timeframe I had to work within, counted backwards with what had to be done and set daily timetables based on that data. I got much nimbler and able to fly better on the seat of my pants. Did it sound good or not? Are the elements working together? Can I pick it up in an hour and still get a good feel from it? And when the feedback determined that one or another element was out of kilter, how and where can I fix it?
I hope this helps in terms of documenting an experience in recording live instruments. Any questions on any of this, post them here or hit me up online.
Ah here comes The Devils Highway, soooooo good… cheers Kevin