In their conquest of the pop charts, Pentatonix’s only weapons were the human voice — and the skills of mix engineer Ed Boyer.
A cappella acts have been around for many years, but have only occasionally broken through to a wider international audience. The Flying Pickets had a UK number one in 1983 with ‘Only You’, a feat Bobby McFerrin emulated five years later with ‘Don’t Worry, Be Happy’, and since then, there’s been the emergence of beatboxing — a hip–hop–influenced form of vocal percussion — and the editing and treatment options provided by DAWs. Both have made it possible to create a cappella recordings with punch, precision, width and depth to match those made with musical instruments.
It’s nonetheless taken nearly three decades for another a cappella act to hit the big time. Enter Pentatonix, a quintet that formed in 2011, and in that same year won the third season of the American a cappella TV talent show The Sing–Off. Skilful use of social media furthered the group’s popularity, and as of 2012 a string of EPs and singles were released, with the group’s take on anything from Psy’s ‘Gangnam Style’ to Imagine Dragons’ ‘Radioactive’ to Gotye’s ‘Somebody That I Used To Know’. Their homage to Daft Punk, ‘Daft Punk’, was awarded a Grammy in the category Best Arrangement, Instrumental or A Cappella.
After signing to RCA in 2014, the label repackaged Pentatonix’s extant releases with some new material to form two compilation albums. The same year, the group’s Christmas album, That’s Christmas To Me, sold more than a million copies and reached number two in the USA. October 2015 saw the release of a new, self–titled album containing almost only original material. Pentatonix enjoyed impressive success, reaching number one in the US, and achieving high chart placings in many other countries around the world.
Given Pentatonix’s ambitions for the latest album, it’s not surprising that they hired collaborators with proven track records in the pop world, in this case Thaddis ‘Kuk’ Harrell, a songwriter, vocal producer and arranger who has worked with the likes of Rihanna, Celine Dion, Mariah Carey and Beyoncé, and star pop mixer Manny Marroquin, whose credit list could easily fill the rest of this page. However, Pentatonix also remained loyal to the people who helped them get to the top, because several earlier collaborators also feature on Pentatonix, notably vocal producer and arranger Ben Bram, and mixer Ed Boyer.
Ed Boyer is a producer, arranger, engineer, mixer and singer who specialises in a cappella music. He’s worked on an impressive amount of projects, including TV shows such as Glee, The Sing–Off, Sing It On, Pitch Slapped, Sesame Street, Ellen and many more, plus with groups such as Straight No Chaser, Home Free and the Swingle Singers. His home studio, which he prefers to call his ‘office’, is in Connecticut, a 75–minute drive north of New York City.
“I worked with Ben [Bram] on The Sing–Off show, and when Pentatonix came on in 2011 he was helping them with their arrangements and brought me on board to mix them. We continued with this role division since then. I wasn’t there for the recordings of the new album, but I know from Ben that the band wrote a ton of songs, 40 or 50 I think, alone and with many different songwriters, just trying to figure out what direction they wanted to go in and who they clicked with. One of the notions people have about a cappella is that it is just a medium for covering other people’s material. So this was a new experiment.
“Pentatonix have a very grassroots fans base, which has been following them from the beginning, and which is comfortable with the identity the group have established. But with the new album the group wanted to achieve a sound and production that appeals to people beyond their fanbase. The contemporary a cappella subculture is quite small, but within that there are production approaches that range from the totally organic — ie. just set up a couple of room mics to record a group live — to very intricate productions, with people using all sorts of amps and effects and triggers and production techniques. They are technically a cappella, but basically their voices don’t sound like voices.
“In this spectrum Pentatonix are kind of in the middle. They’re at a balancing point where they don’t sound traditional, but also don’t do productions that sound so out–there that people don’t know what they’re listening to. A very typical reaction when people first hear Pentatonix is that they think that the band is using some degree of synthesis, especially subharmonics, and drum samples. And then when they listen more closely and pick the sounds apart, they realise that it’s all just done with voices. They’re not so heavily produced that you can’t work out whether loads of elements of their productions are fake or not.”
Boyer conducted all his Pentatonix mixes at his studio, despite the fact that it is, says Boyer, “pretty minimal. I run everything out of a Mac laptop. As Pro Tools native has become better and better, I can run sessions on my laptop and pile things on for days and never get past 10 percent CPU usage, whereas some of the Pentatonix sessions we later loaded onto Germano studio’s HD system slowed down. I am now pretty committed to do everything in a laptop. It has the added advantage that if I am stuck at an airport for a two–hour layover, I can get out the session and do some non–critical work.
“Other than my laptop, my rack is tiny. I have a few preamps, like a couple of Avedis MA5s, which is a Neve–style 500–series, which are great to mellow out a very bright mic, and a Five Fish Audio SC501 500–series, which is very clean, and which I use when I don’t want to hear the pre. My interface is the RME Fireface UFX, which I have because of the TotalMix FX software, mixer and router, which allows me to go in and out of Pro Tools and my hardware and gives me latency–free cue monitoring. My main mics are the Flea 47, an Advanced Audio CM12, and a Brauner VM1. The Brauner is really beautiful for the right voice and has a lot of detail, though there are voices where you don’t necessarily want to hear all the detail, as it can be too revealing. In addition I have the Shure SM7B, which is great for recording beatboxing, and I’ve held on to my Rode NT1A, which I got when I began more than 10 years ago! I have GIK Acoustics panels and diffusers around the room, plus some diffusers made by my dad.
“My AKG K240 headphones also are important. So much of tracking (and sometimes mixing) happens in non–ideal places. Remote tracking is popular because it’s much cheaper for one producer or engineer to travel than it is for an entire group, and this means you’ll often work in spaces that are not meant for recording, and often in the same room as the performers. Familiarising yourself with a pair of headphones is a good way to achieve some consistency when you’re away from your main monitors. I like the AKG K240s because the work I do on them translates fairly well. Unlike a lot of models, they don’t flatter the source too much. Also, when working in the studio, headphones are great for checking for mouth ticks and pops. When you have 100 tracks of vocals, noise–hunting can be quite an endeavour!”
As anyone who works from home can testify, the comfort and convenience of having a studio in the house can quickly become a challenge when there are family demands, an issue Boyer was confronted with during his mixing for Pentatonix, which took place over four weeks during August and early September 2015. His wife had a back problem and was advised not to lift anything for a few weeks, so she and their newborn stayed in the room next door, and Boyer rushed out every time the baby needed moving or its nappy changed. “I’d run in, do what I needed to do, and run back into the studio to carry on with mixing. It was an insane month!”
Boyer’s mixing stage came at the end of a lengthy recording process which saw Pentatonix recording “all over the world,” says Boyer, “because the band was on tour and would record wherever and whenever they could get into a studio, all over the US, Europe and other places. If you look at the liner notes, every song has several engineers! Pentatonix usually work through overdubbing, although they will occasionally record everyone live in the same room, with just some baffles, if they have a simpler song or a song that needs rubato, like ‘Take Me Home’ on the new album.
“Many of the arrangements happen when they are on tour, backstage, waiting to go on, or when they have a night off and spend time in somebody’s hotel room. Sometimes it’s just one or two of them. Someone gets an idea, and they take it and run with it, and there may be one or two collaborators, and as they get further down the line they’ll hash out a full arrangement and they’ll then demo it with a keyboard or some backing from GarageBand or something like that. Their demo usually is laid out to a grid, and for the final recordings they will, one at a time, sing to the demo, which supplies the pitch centre and the tempo. They are never going to try to nail their parts in a vacuum. Ben [Bram], who was with them most of the time, and he and people from the label took notes of the microphones and signal chains that were used, so there was some degree of consistency between sessions.”
Consistency between the different sessions was not made easier by the fact that several producers worked on the album. When the sessions came to Boyer for the final mix, there were, he says, “definitely some logistical hurdles. I recall one session where Avi Kapler wanted his vocals treated differently than how they had come in, and we did not have the source tracks, and the song had been recorded in 10 different studios all over the world, so quite a bit of detective work was involved in tracking down his original vocals. But most of the sessions came to me after they had been treated by the producers, meaning that I received three kinds of session layouts: Kuk [Harrell]’s, Ben [Bram]’s and Martin Johnson’s, who was the third main producer. Kuk did all the editing and tuning before he sent his sessions to me, but Ben and I have worked together for a long time, and he simply gets the performance the way he wants and then sends it to me, and I do the editing and tuning.”
Not so long ago, ‘tuning vocals’ was a bit of a dirty phrase, implying a lack of skill on the part of the singer. But it now has become widely accepted that pop vocals have to be perfectly in tune, and post–production tuning is the only way to achieve this. Boyer explains that his tool of choice is the stand–alone version of Celemony’s Melodyne.
“I use Melodyne for both pitch and rhythm. I do that before I start the mix, even though I may finesse things later on during the actual mix. I can bring the entire session into Melodyne, so I can see and hear everything at the same time. I don’t like other pitch–correction plug–ins because you just hear the track that you are correcting. The important thing is not whether one track is pitch perfect, but how it sounds in the context of the entire arrangement. So I treat all vocals at the same time, and Melodyne then renders the changes for me and prints out the files, which I then import in a ‘Save As’ copy of the original Pro Tools session.
“The degree to which I tune depends on the a cappella genre. If you are dealing with a more classically orientated group, you may be more inclined to leave the tuning looser. But with pop music, even if people don’t know the difference between tempered and true tuning, they subconsciously expect tempered tuning. Some people have a purist attitude and don’t agree with tuning, but Pentatonix is a pop group, and not only do you want it perfectly in tune, you also want to hear some of the timbre of pitch correction, because it is part of the sound of modern pop music.
“Pentatonix are not sloppy or far off, so it’s not like I have to clean up a mess. They occasionally release videos of them singing together in a room, recorded without multitracking and hence pitch correction, and they do that to show that any production choices that they make are not used as a crutch, but because they are choosing to make pop music. Melodyne is used to get their vocals to sound a little tighter and right in the pocket, so it feels and sounds a bit more like EDM and pop music.
“I don’t always import the vocal percussion into Melodyne, but when I do, I turn the ‘click–to–grid’ function off, so nothing is automatically moved in any direction. I just go through by hand and place things where I want them. Sometimes it’s right on the beat, sometimes a bit after, or before. With beatboxing I often move things slightly forward, because no matter how good the person is, you never quite have the velocity or the crack of somebody swinging a stick at an actual snare drum. So many beatboxing sounds have sort of a ramp up that you would not necessarily get with drums, so sometimes I centre the peak on the grid and get the ramp before the grid.”
Boyer’s work in Melodyne is part of the general preparation work he does when he begins work on a mix. “The first thing I do after opening a session is put the song on loop and listen to it a number of times, to get a basic idea of what’s there and what vibe the group is going for. I’ll poke around and have a look at the individual tracks, and I’ll move things around and will label them, and do more of that kind of housework. By the end of that process I’ll have a feeling what the band is actually going for. Pentatonix don’t like giving many instructions at this stage, they just like to see where I take it, and then they give feedback. So I open the session, listen to it, and run with it, and only very occasionally do I take it in a direction where they say, ‘Oh, actually, we were thinking of something completely different.’
“Once I know what direction I want to take the mix in, I’ll do the tuning and timing, and then the mix itself is pretty much similar to mixing a band, where the first thing I do is listen to the kick, and try to get that to where I want it to be, and then the snare, and so on. I then add in Avi’s bass vocals, work on them, make them sit with the beatboxing, and then I add the lead vocals; and once I have these three elements sitting right, I fit in the other background harmonies.
“I did almost all the mixing at my studio, in several rounds, each time getting the group’s input, until they were happy with it. For the songs that were produced by Kuk [who was involved in nine of the album’s 16 songs, seven of which were mixed by Boyer] I went to New York City after completing the mixes at my place. We rented a room at Germano Studios for four days to finalise the mixes with Kuk. It was like the group had put its stamp on the mixes, and then he wanted to put his stamp on them, and because we were in New York people from the label could stop by to have their input, so we could get everyone on the same page. It was a matter of fine–tuning, with us opening up a session and perhaps spending a couple of hours on it, though with a few songs we took half a day. The work at Germano also was all ‘in the box’, with us just going through the first two channels of the desk there.”
- Beatboxing: Wave Arts TrackPlug & MultiDynamics, SoundToys Decapitator, Waves Trans–X & Renaissance EQ, FabFilter Pro–Q 2, XFer LFO Tool, Avid Trim.
“Kevin [Olusola]’s main beatbox track is right at the top, and below that is a ‘BBox kik’ track, which came with the session and is a kick sample taken from his beatbox track. I sent both tracks to the track below that, ‘BBox and kik’.
“Next are two more sample tracks, one with a kick and another with a snare sample. I added both from a library I have built up over the years of Kevin’s sounds. I will find a kick, or snare, in his beatboxing that just hits the right spot. With beatboxing the kick will move around, and sometimes it’ll be 25Hz higher than the one before or doesn’t sit right for other reasons. So if you blend a beatbox track with one or more samples you get a lot more consistency. I added the snare sample because sometimes his snare comes out with a little bit of phlegm on it, which isn’t necessarily intentional, but I like it because it beefs up the snare sound a bit.
“I treated all these tracks with similar sets of plug-ins. I added the Wave Arts TrackPlug to the kick sample that came with the session, which is a typical channel strip, with gate, compressor, EQ, limiter and so on. It’s straightforward, no–frills and I can just open it up and dial things in very quickly and easily. It doesn’t have a character of its own per se, but I can get a sound pretty close to where I want it with a few moves. It’s high–passing on the kick track, to get out some low rumble, and then I added some compression, letting the transients through and then it clamps down. It just cleans up the sound a bit before it hits the ‘BBox and kik’ track, on which I have the Wave Arts MultiDynamics compressor, which is gating out certain frequencies, so that, for example, when he’s doing a hi–hat, and there’s an inadvertent low noise, it takes that out, but without affecting the kick.
“Then there’s the SoundToys Decapitator, which acts a bit like parallel distortion. I drive it pretty hard, but then turn the Mix down, so it just thickens the sound and adds some character. Then there’s the Waves Trans–X Multi, which is controlling the spit, and adding some transients to the low and mids, trying to get some more punch. Next is the FabFilter Pro–Q 2 EQ, which is bumping around 55Hz, where the kick is, and then the Waves Trans–X Wide, with which I try to add some more transient to the sound as a whole. The kick and snare samples below both have the Trans–X Wide, which I always use to get a little more pop out of things. As I mentioned before, the thing about beatboxing is that no matter how good the person is, just in the first milliseconds, they don’t get the same transient as a real drum, and so the Trans–X Wide adds a couple of decibels right at the beginning of the sound to make it more crisp and punchy.
“Below these tracks are some beatbox overdubs. Pentatonix love re–recording. If they don’t like something, they choose to re–record it. Sometimes this can be frustrating because you just want to move ahead with what you have, but it’s also nice because it gives you more options. In this case we felt that the chorus needed a bit more to move it along, so Kevin went in and overdubbed a pattern with more subdivisions. That overdub might have been done during our last four days at Germano. Next is the ‘BBox hat’, which is a sample that repeats, and yet more beatbox overdubs.
“The ‘BBox shaker’ again has the TrackPlug to clean up the sound, and also the Waves REQ6 boosting some high end, and the XFer LFO Tool is a DJ tool that I use to draw an envelope under the shaker, essentially manually gating out some of the in–between sounds that don’t add to the musicality of the part. There’s also a Trim at the end of the path, and this is because often someone will ask, ‘Can you bring down that shaker 3dB?’ I’ll do that and then next time they’ll say, ‘You know what? It was better where you had it.’ If I had automated the fader I wouldn’t remember where I had it, so part of my process these days is to add a Trim, bring the gain down with 3dB, if requested, and if necessary I can then later simply mute the plug–in and the track will go back exactly to where I had it.”
- Body percussion: Wave Arts TrackPlug & FinalPlug, Waves Trans–X, Avid Time Adjuster, Fabfilter Pro–Q 2.
“The snaps, foot stomps and handclaps are all done by the band. The first snaps track is the one I was sent, and I liked it, but wasn’t so keen on the same snap sample being used all the time. So I took some band snap samples that I had from previous sessions, maybe seven or so, and pasted those in so that they were never in the same sequence. I really like the attack of one of the snap samples, and pasted that in separately on the ‘SNAPS_god’ track. By blending all three snap tracks together I felt I got the right balance between consistency and variation. The snap tracks again have the TrackPlug and Trans–X Wide.
“The first clap track came with the session, and I added a sample of a group of people clapping in a church, purely to get some reverberant space in there. I have the TimeAdjust plug–in on both, because I’m delaying one of the clap tracks by a few hundred samples to create more of a stereo effect. I also have the Wave Arts FinalPlug on it, which basically is a limiter, for level and punch. There were four foot stomps tracks in the session, which are bused to the ‘stomps’ group track, on which I have the TrackPlug, FinalPlug and Pro–Q. Below that are two stomp samples I added, taken from the previous album, because they had some natural room sound on them. Once again I time–adjusted the second sample for some extra width.”
- Bass vocals: iZotope Ozone, Kush UBK1.
“There was one bass track of Avi [Kaplan] singing, but I pulled one phrase out to a second track because I wanted to treat it with some slightly different EQ. Both tracks have the iZotope Ozone, which I use as a channel strip, applying EQ, multi–band compression and a bit of parallel excitement, and the Kush Audio UBK1 compressor, which adds colour. I tried to create some additional harmonics in the bass vocal with the UBK1, so it also translates to smaller sound systems. What I like about that plug–in is that every parameter has a dry/wet switch, so you can drive it to get some fuzz, but then set it to the exact degree of wetness that you like.”
- Lead vocals: FabFilter Pro–Q 2 & Pro–DS, Wave Arts MultiDynamics, Kush UBK1, Waves Renaissance Compressor & EQ, SoundToys Echo Boy, Avid D–Verb & Mod Delay III, Klanghelm MJUC.
“The first three tracks, ‘LV Sco 1–3’, are where Scott was really trying to get his lead vocals just right, and he ended up adding some stuff later in the game for that. So some of that came from different sessions using different mics. The same plug–ins are used on all three vocal tracks, but with slightly different EQ settings to match them. The chain starts with the Pro–Q, which is just a high–pass, the FabFilter Pro–DS, for some de–essing, the MultiDynamics compressor, the Kush UBK1, for some more compression, adding colour, another instance of the Pro–DS, another Pro–Q for the real EQ shaping, and the R–Compressor. I use two instances of the Pro–DS because if I had tried to take all the sibilance out with one de–esser, it would have affected the sound too much, so it was better to have two working less hard. Kuk likes the RCompressor and he added this during our four days at Germano, for a little bit more pop. There also are four sends, one of them going to my Echo Boy quarter–note delay, one going to my Echoboy slap, and one to Kuk’s D–Verb reverb and one to his Mod Delay III. My delays are more symmetrical, whereas Kuk likes delays that ping–pong between left and right.
“The chorus has lead vocals from both Scott and Mitch [Grassi]. Really it’s Mitch’s lead, but it’s a duet. Coming from the verse Scott takes one step back, and Mitch one step forward, if you want to put it like that. Mitch’s lead vocal treatments are mainly done by Kuk, with the RCompressor, REQ6, and his reverb and delay, and I added a MultiDynamics compressor. Scott’s vocal has some of the same treatments as his other vocal tracks, plus the Klanghelm MJUC compressor, which is really great for colour. I used it to squash and darken the vocals a little bit, just to keep them underneath Mitch’s.”
- Backing vocals: FabFilter Pro–Q 2, Pro–DS & Pro–MB, Wave Arts MultiDynamics & S1 Imager, IK T–Racks White 2A, Avid Trim & TL Space, McDSP Futzbox.
“Both Mitch and Kirsten [Maldonado] sang two background vocal parts, which they doubled, for imaging purposes, so they can be panned left and right and get out of the way of the lead vocals. This meant that they each had four tracks of BVs, and each set of four was sent to its own bus. The treatments on the two buses are similar as on the lead vocals, with the Pro–Q2 EQ, Pro–DS and MultiDynamic plug-ins, and also the T–Racks White 2A Levelling Amplifier, which is an LA2 emulation. There are Trim plug–ins for the same reason as with the shaker. There’s also a stereo Choir track, which are the teachers and singers from the a cappella camps that Avi started [see www.acappellacademy.org]. The choir was recorded at Westlake in LA, and is quite roomy, in contrast to the band’s vocal tracks, which they like to be quite dry.
“The choir has the TrackPlug, MultiDynamics, REQ6 and the Waves S1 Imager, the latter because I wanted to spread it out and take it out of the middle to leave space for the lead vocals. There’s also a ‘Yup’ backing vocal track, which has the McDSP Futzbox, on a megaphone setting, because I wanted to give that some character and make it sound as if it was sampled from somewhere. The bridge vocals were a last–minute addition, because there were a couple of bars where nothing was happening, and so Kevin, who also sings, went in and laid those vocals down. I have the Pro–Q 2 EQ, Pro–MB multi–band compressor, S1 Imager and a Trim on these vocals. Almost all the backing vocals had a send to my Verb aux track, which had TL Space on it, on a church setting. Finally, there’s a master track at the bottom which has the Ozone, for listening purposes, but I stripped that out for the final mix bounce down, which was to disk.”
As one of the US’s leading a cappella engineers and mixers, Ed Boyer has built up a lot of know–how when it comes to recording this highly individual musical subgenre.
“The recording of most a cappella vocals is pretty straightforward, if you record them individually, as was done with Pentatonix. It’s just like you would record any other vocals. With bass vocals it’s a matter of using the proximity effect to your advantage, ideally using a hefty, large–diaphragm condenser mic, up close, like the [Neumann] FET 47, which I have been using a lot recently. The Shure SM7B dynamic also works well. They’re both mics that Pentatonix like to use. Of course you have to make sure that you don’t get too much wind in the diaphragm, so you have to have screens. The alternative is to use a mic that can take some beating, like the SM7B.
“You can record beatbox — or ‘vocal percussion’, as we say — on any good mic. There are, however, different challenges with different types of mics. Dynamics are good because they can handle the plosives and because they de–emphasise unpleasant spitty sounds, inadvertent lip smacks, and so on. [Electro–Voice] RE20s are good mics, and SM7Bs are great because they give a particularly good thump on vocal kick sounds when recorded closely.
“Sometimes a beatboxer will give a better performance on a dynamic handheld because that’s what he or she is used to holding in a live performance. The down side, of course, is that you don’t get the same detail as you get with a large–diaphragm condenser. With those, you get more of the total sound spectrum, even though there is the issue of the distortion from plosives and wind. But if you have a beatboxer with good control who specialises in detailed hi–hat patterns and so on, you’d probably want to use a large–diaphragm condenser.
“When I was producing sessions for the movie Pitch Perfect 2 with veteran film engineer Joseph Magee, we took a best–of–both–worlds approach to record beatboxer 80 Fitz. We gave him a Shure SM58 to hold (which he cups to sculpt his sounds) then set up an Advanced Audio CM47 a few inches away to get a more natural and detailed sound. This can get a bit tricky with phase, since he moves around a bit, but with a little policing, it worked out.
“I’d say the biggest trick for recording contemporary a cappella is maintaining performance energy. For a number of reasons — there’s more isolation, it’s more logistically feasible — singers usually record one at a time or in subsets of their group. But not all singers are comfortable with that process, as they’re used to being on stage with their entire ensemble, especially the groups that don’t use live effects or any sort of sound reinforcement/monitoring. So getting them comfortable and making sure the vibe is right is huge. That means getting a good cue mix, but also pacing the session in a way that’s comfortable for the singer, making sure the energy in the room stays up, or relaxed, depending on what’s needed, and, most importantly, not settling for takes just because they’re technically correct. With digital recording, the tuning and timing of recording can be 100 percent, so the focus should be on what kind of vibe and character the recording has.”
Ed Boyer, who is originally from Ohio, studied Classical Voice at the New England Conservatory of Music and Liberal Arts at Tufts University, both in the Boston area, Massachusetts. He graduated from both in 2004. He recalls, “At Tufts I joined an a cappella group, the Tuft Beelzebubs, for no other reason than to hang out, have some beers and sing some pop songs. Ironically, that is what led to my career. Being in that group taught me about arranging and how to run and conduct a group. Also, in 2000, when we were recording in a major studio, we realised that if we did it ourselves, we could save tens of thousands of dollars. So we spent about $8000 on a Pro Tools LE rig and some cheap mics, and started recording. It was fun, it was productive, and that aspect of my career snowballed from there, and bloomed into something that I did to help others.
“I’m largely self–taught on Pro Tools, though a very good a cappella mixer named Bill Hare was a mentor for me. Bill mixed the first Beelzebub album that I tracked and produced DIY–style, and since then we’ve remained in touch. He now masters many of the things I mix, including past Pentatonix albums. But in those early years I dived into projects and made tons of mistakes, and every year I’d look back and think: ‘What was I thinking doing that?’ I got a lot of my knowledge from reading. I’d just go to Barnes & Noble and park out in the magazines aisle and read every music tech magazine they had — including Sound On Sound, of course. Every time you find an article covering something you don’t know, you read it and try to learn from it.”