PLEASE NOTE: This article has been archived. It first appeared on ProRec.com in December 1999, contributed by then Editor-in-Chief Rip Rowan. We will not be making any updates to the article. Please visit the home page for our latest content. Thank you!
As ProRec’s Editor-in-Chief I get a lot of email on a variety of subjects. Questions range from “what kind of mic should I use” to “how do I control room noise” to “why is my hard drive so slow.”
Hey, I’m just a guy. I do not proclaim to be the most knowledgeable person about anything. In fact I’m the classic Jack-of-all-Trades-Master-of-None. I’m a half-decent drummer with half-decent engineering skills running a half-decent studio and using my half-decent programming and writing skills to create this half-decent website you’re reading.
One thing that I’ve noticed over the years is a lack of hard factual information about recording techniques. I read a lot of interviews, most of which offer anecdotal stories but very little “how-to” information about the nuts and bolts of creating a recording.
There’s some good reasons for this.
Recording is NEVER prescriptive. What works great in one situation will not work at all in another. The fuzziness of the process is why a great engineer gets paid the big bucks. He or she must be able to respond interactively to any given situation, drawing on a wealth of experience and talent to optimize whatever situation arises. So it can be quite misleading to offer “recommendations” in recording techniques when the problems can be so unquantifiable.
And there are really a lot of issues to cover. We’re talking everything from electrical engineering to marketing, from acoustics to psychology and from computer science to ergonomics. It’s a lifelong learning process for anyone who is committed to the work.
Armed with this information, I’m going to dissect a recent demo recording made by a client of mine, and talk about getting their mix from the ground up. This article is targeted directly at the person who is in the learning process. Experienced recordists might learn something along the way, too. My hope is that, although you may have come to scoff, you’ll stay to learn.
The band I’m going to cover is Dallas-based Four Mile Mule, an alt-country group that has recently completed their first demo CD, “Black and White Movie.”
Several key points need to be made about this article. Clearly the recording process depends highly on the material being recorded. As an alt-country band, Four Mile Mule has a very different sonic goal from Limp Bizkit, Frank Sinatra, Dave Brubeck, DMX, or Supertramp. Their project was a demo project, focused on documenting the band’s entire song catalog, not on creating the perfect radio-ready sound. We spent as little time as possible tweaking and as much time as possible recording.
So I’m not holding this work up as a hallmark of sonic perfection, rather, it’s a useful learning tool. Moreover, it’s the most recent thing I’ve done, so it’s fresh on my mind. I will cover the making of a song from beginning to the end in as much detail as I can muster. So pop open a cold one, and let’s dive into the details and look at how this project came together.
Defining the Sonic Goals
There’s an old saying that goes, “If you don’t know where you’re going, any road will take you there.”
There’s a corollary to that saying. “If you don’t know where you’re going, how will you know when you get there?”
Followed closely by, “If you don’t know where you’re going, who knows where you’ll end up.”
There’s a point to be made here.
Before I ever hit the record button, the first thing that must be determined is the artist’s sonic goals. Sonic goals refer to the kind of outcome the artist wants to achieve. Should the overall sound be majestic and lush, or tight and defined? Will we be developing a Wall-of-Sound mix, or a sparse, “exposed” mix? Aggressive or approachable? Angry or melancholy? Lo-fi or hi-fi? Dramatic, or understated?
If you’re the producer as well as the engineer, then these are decisions you’ll make with the band or artist. If you’re just the engineer, then you’ll work with the producer to understand the sonic goals. If the band is a developing band, you may want to set these goals on a song-by-song basis. As the project progresses, a sound should emerge that will define the rest of the project. On the other hand, if the artist already has an established “sound” then you’ll probably want to set these goals for the project up-front, then evaluate each song’s place in the whole.
Bob Lichty wrote an excellent article on this topic a few months ago. In it he covers the importance of clearly defining the roles and goals of the project. If you’re interested in production and project management, be sure to read it.
The simple fact is: you may not know where you’re going when you get started. And that’s OK. As long as you all agree that somewhere along the way, a decision will be made and a direction will be taken.
Decision. There’s that word. From the Latin: de – off, from; caedere – to cut. To cut off from, to separate, in making a choice. That’s right. A decision to do something almost ALWAYS involves NOT doing something else. Think about it.
In the case of Four Mile Mule, the band already had a decent song catalog of about two dozen songs. All of these songs were solid, well-written songs, but none of the songs had “radio-ready hit” qualities. So there was no clear direction where the project would end up. As it turned out, the lead-off song emerged during the final stages of the recording sessions, and was the last song we recorded. This song helped define which songs would be included on the demo CD and which ones would be cut.
Four Mile Mule has a sound that is similar to adult alternative rock sung with a country twang. It’s guitar-based powerpop music with strong riffs, but the lyrics are earthy and “down-home” and delivered with a decidedly country flair. This is not “Black Hat Country” – it’s country music designed to appeal to a rock audience.
Using the song’s stylings as a guide, we decided early on that the overall sound needed to be simple, reasonably hi-fi, and just a little edgy. The goal was to create a sound that would be a little prettier than your typical rock radio hit, but tougher than your typical country hit. The guitars needed to be edgy, the drums needed to be fat, and the vocal needed to be full-range and upfront. We felt this would be a sound that would best show off the group’s rock / country crossover potential.
Let’s take a close look at a song from the album, and build it up track-by-track so you can see – and hear – how the pieces fit together. The song is the band’s lead-off single, “Black and White Movie.” You can buy the CD and listen to it first, before reading further. You’ll need to hear the song in order to get an idea of how the tracks fit together in the mix. Click here to get the CD. “Black and White Movie” is the first song on the list.
Please note that I neither wrote the song nor performed on it, and the song is used here with the artist’s full permission. In other words, this ain’t my personal plug.
“Black and White Movie” was cut in my studio, using Cakewalk Pro Audio v.9 and a MOTU 1224 audio interface. My primary DAW is a “Roll Your Own” computer (Celeron 300A @ 450 MHz, 128 MB RAM, 3x 20 GB EIDE hard disks, yada yada yada). Other than having a lot of online disk storage (60 GB) this is a pretty average computer by today’s standards. It is capable of performing most of my recording needs, though I can always use more CPU bandwidth. I guess the point here is that I am NOT using the latest, greatest computer hardware.
See, I do NOT consider the cost of a computer to be an “investment” in my studio. Repeat: the computer is not an investment. It is an expense, not an asset. The reason for this is that computers simply depreciate far too quickly to consider them an investment. Also, to purchase state-of-the-art computer technology you usually have to spend five to ten times what you would spend on “decent” computer technology.
For example, a top-of-the-line Pentium will cost you up to $1000. A very good but not state of the art Celeron will cost you about $100. The very best and fastest SCSI drives will cost you about $750. But a very good EIDE drive only costs about $250. And here’s the kicker: in two years all of these products will have a street value very close to $0. If it’s going to be virtually worthless in a couple of years, I’d rather not drop my money into it, thank you very much.
I tend to stay practical in my selection of technology. In my view, whatever will get you there the cheapest is often the right choice. This may sound like heresy our world of More Bigger Better Faster NOW! I just feel that for many people, having the biggest, fastest, most powerful computer is seldom related to Getting The Work Done and usually related to some kind of Big Dick Contest.
You heard me right. The recording market – a male-dominated market for better or worse (for worse, if you ask me) preys and profits on male ego insecurity. This same biological imperative that tells me that I need a new BMW M3 and a new wardrobe and a new, extremely sleazy girlfriend also tells me that whatever new, cool hardware I don’t own makes me inferior to the guy who owns it. I try hard not to listen to that imperative. It ALWAYS gets me into trouble.
On the other hand, I do view my software as an investment. In fact, software is the primary reason I like computer-based recording. Say, for example, that you spend $500 on a compressor plugin that has virtually identical features and sound as a $500 hardware unit. If your computer is fast enough, you might be able to run 10 compressor plugins simultaneously. The hardware unit is, however, always just one piece of gear. That means that your $500 software investment purchased the equivalent of $5000 in hardware.
Now, let’s say it’s two years later, and you upgrade to a new CPU that’s twice as fast as the one you used to have. Now you can run 20 compressor plugins. Congratulations, you have just received $5000 in free gear! And by this time the software manufacturer has enhanced the plugin, and for only $50 you can have warmer, richer sound than before. I’ve never had a hardware company sell me an upgrade for a tenth of the sales price of the unit – no, to get the new version of the hardware, you’ll have to buy a new unit.
The only hardware gear I will happily spend money on is mics, preamps, and maybe speakers. A great mic or preamp will hold its value very well. Some even appreciate in value. But, when it comes to buying a computer, I want to buy in the “sweet spot” of the price / performance curve. By not purchasing at the leading edge I spend 50% of the money and still get 90% of the performance.
One last note on the computer. The advent of low-cost, high performance EIDE disks has coincided with the development of cheap removable drive bays. For $10 you can purchase a drivebay that installs in a 5¼” drive space, and for $10 per disk you can purchase the cartridge for the hard disk. That’s right: $10 for the bay, $10 for the cartridge. A good 20 GB drive costs about $250. So, for well under $300 you can buy all the media that an artist will need to record a 32-track 24-bit double-album-length project. I can charge this back to the artist as a media cost.
It’s a beautiful thing.
I know I’m going to get e-mail from the SCSI folks out there. “Pro’s use SCSI.” “SCSI is better.” “My friend bought an EIDE drive and his house caught on fire.”
Really, you ought to read my e-mail.
I’m not talking about BETTER or WORSE. I’m talking about What’s Practical and Gets the Job Done. For whatever reasons, UDMA/33 EIDE drives are plenty fast (my UDMA/33 drive can sustain 48 tracks of 24 bit audio, and I really, really don’t need any more than that), plenty reliable (even the worst hard disk has a higher MTBF than an ADAT), and awfully cheap ($250 for 20 GB). If you have SCSI, cool! I’m not saying that SCSI is bad! I’m just saying that the price/performance ratio of cheap disks has changed the way I think about hard disks.
See, I now no longer need to think of the disk as a fixed part of the system. I treat it like disposable media. That also means no backups. I can hear you now: “No backups! You’re crazy! What if your disk crashes?”
Used to be that a DAW would have one or two really expensive drives that were in constant use. My old DAW was built on this philosophy, complete with $1000 9GB Barracuda UW SCSI drives. I backed those drives up weekly! When you’ve got just one disk for all your sessions, and everything depends on the one disk, and you’re hitting this one disk all day long, for days on end, month in and month out, you bet: back that puppy up. But that’s the old way of working. When you only use a drive for a couple of weeks or months, it’s hardly broken in by the time you’ve pulled it out of the computer and put it on the shelf.
I look at it this way. I have never seen an engineer back up 2” analog tape, even though the recording business is full of horror stories from well-known bands like Boston and Queen who came really close to just wearing out their masters during the sessions. And I know only a very few people who regularly back up their ADAT tapes, even though ADATs can get pretty damned glitchy. Hard disks are probably the single most reliable audio recording media ever made.
Anyway, enough ranting. Let’s make some music.
Four Mile Mule had already recorded rhythm tracks for about twenty songs when “Black and White Movie” was written. We all felt strongly that it was the catchiest song in their catalog and decided to record it and package ten other songs with it as a demo CD.
With only 8 inputs into my recorder, we would cut drums and bass at the same time, then go back and add acoustic and electric guitars and vocals. We made the artistic decision to NOT use a metronome or click track for the rhythm tracks. We wanted to hear the natural ebb and flow of the tempo that you get when a band will “push” and “pull” the energy of the song. Therefore, even though we were only recording the bass and drums, the entire band would set up so that the rhythm tracks would be cut with the live feel.
The band wanted the CD to have a nice, polished sound, so I wanted to get a pure drum sound on tape. Setting the entire band up in the tracking room would create too much bleed on the drum mics, so the drummer set up in my tracking room, while the rest of the band set up in the control room. In this way, everybody could hear each other really well, but the drum mics would be getting only the drum sound.
We took the bass direct for this album. The bass player plays a 5-string Fender over a GK head and a Hartke 4×10 cab. I am NOT a fan of transistor guitar amps or Hartke cabs, at least not for this kind of music. We needed to get a warm, round, fat bass sound with a little “hair” on it. The GK / Hartke combo – a solid live rig – proved to be too bright and pointy in the studio. After capturing the direct sound of the bass, we would run the recorded signal back through a low-wattage tube amp to add a little “cardboard” to the sound.
For the song “Black and White Movie” we actually wanted a distorted bass sound. To get the sound we were going for we ran the recorded bass tracks through a Fender Pro Junior amp cranked up nice and loud in my large tracking room. The mix of the compressed direct bass sound and the Fender amp turned out to be the fuzzy edge that the song needed. Take a listen to a brief clip of the bass guitar to get the idea.
You may be surprised at how distorted the bass actually is. It’s not the bass sound I’d usually record… but it’s the sound we decided to get for this song. See, once you place a distorted bass track into a mix with even one distorted guitar, the ear begins to hear all the distortion as coming from the guitar. In listening to the complete mix, you probably can’t tell that the bass is distorted. The trick in this case is that the bass was played with slightly more movement than the guitar. The distortion added an extra layer of thickness to the song without actually adding another part. The result is that the mix is heavier and thicker, but still retains the simplicity of a four-piece arrangement.
The drummer plays a typical 4 piece kit: kick, snare, two toms, hats, ride, and two crash cymbals. I like to work with drummers who play relatively simple kits. There are precious few drummers in the world who can truly make use of more than two or three toms, and very little music that requires a double-bass. Great drum parts are usually the result of expressiveness and groove, not any specific number of drums. Less is more.
On snare, of course we used an SM57. When I use a 57 on snare I usually line it up pointing directly away from the hi hats. This cuts down on the bleed from the hats, and is usually out of the drummer’s way. I point the mic at the snare at about a 20-degree angle down, about ½” away from the head, and 1-2” inside the rim. The mic should be pointing roughly at the center of the drum.
The key to getting a good snare sound is: have a good sounding drum. This should be a no-brainer, but you’d be surprised how many snare drums just don’t stand up when recorded. And price is not the issue. I’ve seen some really crappy little chrome snares produce a great sound and some high-dollar snares that were really terrible. The key is the selection and tuning of the heads and snares.
Depending on the kind of music and drummer I will sometimes use a “ring-off” on the snare. You’ve seen these in the music stores hanging by the drum heads. It’s a ring that sits on the top of the head right inside the rim. I’ve also known drummers who will damp the snare by placing their wallet on the snare.
Why are these good ways to mute the snare’s ring? Why not use a bunch of duct tape? Here’s the deal. Tape, or anything that adheres to the drum head, will change the tuning and harmonics of the head. Usually this screws up the attack, making the snare sound really dead. But a damping ring – or even a wallet – does not adhere to the head. When the drummer hits the snare, for a split-second the ring is lifted up, giving you the entire attack and briefly allowing the full sound of the snare to develop. Then, the damping ring makes contact with the head, cutting the ring.
In the case of Four Mile Mule we wanted a natural, open snare sound, so we did not use a damping ring of any kind. Instead we used gating and EQ to control the snare’s ring.
A snare’s ring is composed of a fundamental and several harmonics. The specific frequencies depend entirely on the drum, but are typically in the range of 200-800 Hz with the fundamental in the 200-300 Hz range and the harmonics at multiples of the fundamental. Be aware that sometimes the first harmonic can seem louder than the fundamental.
In adding EQ and gate to a snare drum most engineers seem to use gating before EQ. I think this is a result of the fact that on most consoles the insert is placed before the EQ. I usually place the EQ before the gate. This way the gate is not “hearing” so much overtone and seems to react more evenly. Either way will work. For best results use a fully parametric EQ with adjustable frequency and bandwidth (Q). If these terms are meaningless to you, read Lionel’s articles on EQ for more help with EQ.
Crank up the gain on a given frequency and set the Q (width) as high (narrow) as possible. Now slowly sweep the EQ around until you hear the snare ring loud and clear. It should ping like a submarine. Now cut that frequency. You’ll want to only cut as much as needed to get rid of that part of the ring. You shouldn’t need to do a full-cut. Repeat for each offending frequency.
I used the Waves Q6 EQ for the snare. Below you can see the EQ settings I used. We wanted a bright, crisp sound, thus the serious high-shelf EQ used. I usually apply a low-cut to the snare to keep it from being too dark. This totally depends on the application and the snare. Remember, you want to leave some ring. Otherwise the snare will sound dead.
Next comes the gate. The gate is a downward expander which will attenuate the signal as it falls below the threshold. If these terms are meaningless to you, read my article on Dynamics Processing for more help with gates and compressors. For a snare I usually use an expander with a very fast attack, a medium fast release, and a 2:1 or 3:1 ratio. The threshold needs to be set according to the volume of the snare track. Below you can check out the Waves C1 Gate that I used on this snare track.
Finally comes compression. I like to really compress the snare hard and usually limit it as well. Serious compression adds to the “crack” as well as the “meat” of the snare. For snare compression I use a fast attack, medium release 4:1 compressor. I then crank up the output until I’m hitting the limiter section hard enough to just start to bite off the tops of the attack. This makes the drum fuller and thicker sounding. Below is a shot of the Waves Renaissance Compressor used on this track
When recording snare I like to use a tube mic preamp. The snare drum is probably the single most dynamic instrument on God’s Green Earth, and there’s no way to record it at a decent level and not clip the transient. A little tube preamplification can add some depth and bite to the snare’s attack and help get more tone to tape. Sometimes, for a tougher sound, it’s good to even drive the preamp into moderate distortion.
The trick to getting a good snare sound is to work interactively with the instrument and to not be afraid to really jack around with the sound. It’s not reasonable to think that by placing a mic ½” from the head of a drum that you will capture the full sound of the instrument. Instead you need to work with the sound as it is coming out of the speakers to get it sounding good. If you will closely listen to most snare sounds on the music you like, you’ll probably realize that they’re all a little “ugly”. Each snare has it’s own character. Don’t kill the character with too much gating, EQ, and damping. In fact one of my biggest problems with Black Hat Country Music is the sterile, overprocessed, “perfect” snare drum sounds that are prevalent in Nashville these days. Boooorrrrrr-ing.
You can listen to the dry snare sound and the effected snare sound to get an idea of what we’re doing here with EQ, gating and compression. You’ll notice that the processed snare drum is much brighter (maybe too bright), with lots of snare sound and a crisp attack.
EQed, Gated, and Compressed Snare Clip (MP3)
Next comes the kick drum. Like the snare, the trick to a good kick sound is a good sounding kick. Fortunately, kick drums are pretty easy to get right. In rock, country, and jazz, the kick is usually damped with a pillow or some kind of material to cut down on the boominess. The right amount of damping depends on the drum and the drummer, but it ought to be somewhere between “whap” (no tonality) and “boooooom” (all tonality). If it’s at either extreme, adjust the damping material to get a balance between the attack (“whap”) and the decay (“boooooom”).
Now, I’m a drummer, and sometimes it’s cool to play with an undamped kick drum with a full front head on the drum. As with most instruments in rock music, though, what sounds good in rehearsal or on stage may not be the sound that best translates to tape. Undamped kick drums usually fall in this category. Unless you’re looking for an unusual kick sound, don’t record an undamped kick drum.
We used an AT-Pro25 kick drum mic. I am not a fan of the ubiquitous AKG D112, which to me is too heavy sounding. The D112 is good if you want a really big, thuddy kick sound… a sound that personally bores me. I like mics like the AT Pro-25, EV RE20, Sennheiser MD412, or even a Shure SM57 for kick. These mics offer more attack and personality, and create a tighter kick sound than a D112.
With kick, I like to use a little downward expansion to reduce the amount of noise bleed from other drums. I use pretty much the same setup that I use for snare: fast attack, medium release, and about a 2:1 or 3:1 ratio. To me it’s important to set the threshold and release so that I’m not cutting off the decay of the drum. If you cut off the decay with a gate you run the risk of taking the “butt” off the kick sound. It’ll get small-sounding. Just be sure to leave enough decay in the kick that it has its full size in the mix. If you think that the kick sounds too small, yet there’s clearly a lot of bass content (demonstrated by your popping woofers) chances are that there’s not enough decay. Adjust the gate or the damping of the drum.
Next comes EQ. Each kick needs its own EQ depending on the sound of the kick. An important thing to note is that while a kick drum does produce a lot of really deep bass (down to 20 Hz and below), the part of the kick that thumps you in the chest is the octave from 100 to 200 Hz. These are the same frequencies, however, that are usually over-present in a muddy mix. The right balance of deep bass and midbass is paramount when EQing a kick drum.
Some engineers like to pull out all the mids and boost the treble on a kick. This creates a kick sound reminiscent of metal bands with a bid “thud” combined with a treble “tick”. I do not care for this sound, nor is it appropriate for Four Mile Mule. For this kick we used a little low-cut @ 35 Hz (to control the deepest bass), a nice, round bass boost @ 90 Hz, midbass cut @ about 160 Hz, and a high-shelf cut of a few dBs.
Finally comes compression. I like to use heavy compression on the kick drum. Pretty much the same settings as I use for the snare: fast attack, medium release, and about a 4:1 ratio. For both kick and snare I like to drive the compressor into about 6 or so dBs of gain reduction, and then use the Renaissance Compressor’s output limiter to take off a little of the attack and add a little meat.
EQed, Gated, and Compressed Kick Drum Clip (MP3)
The Rest of the Kit
For hi hats we used a Shure SM81. I like to mic the hats on the far side of the cymbal from where the drummer hits it with the stick. This cuts down on the attack and gets more of the cymbal sound. It also lets me point the mic slightly away from the rest of the kit, resulting in more hats and less kit. Here the signal chain is easy: I use a serious low-cut EQ @ about 100 Hz combined with a hi-shelf boost @ about 8 KHz. For the Mules I used a lot of overheads, so there’s plenty of hi-hats in the overhead mics. The hi-hat mic is just there to add a little sizzle and sparkle.
For overheads we used a pair of Crown CM700s. For this application I placed the mics up and over the drummer’s shoulders spaced a few feet apart. Spaced pairs and X-Y configurations work well for drum overheads. If you intend on creating a large, more spread-out image, use a spaced pair. For a more focused, mono sound, use an X-Y configuration. The main thing to be concerned about is to make sure that the mics will not line up on edge with any of the cymbals, otherwise the cymbals will take on a strange, underwater sound. It’s good to keep the mics sufficiently far enough away from the cymbals that they will hear the whole sound the cymbals are producing. The signal chain here was also easy: we used a low-cut at about 300 Hz to reduce the drone of the entire kit and to promote the cymbals.
Finally, the toms were miked with a pair of AT4050s. Shure SM57s are also great mics to use on toms, as are Sennheiser MD421s. Unless the drummer is playing a lot of fills, I prefer to manually edit the silence out of the tom tracks rather than using a gate to isolate the toms. It improves the DAWs efficiency, and I can usually make the edits in about as much time as it would take to patch in and set up the gates.
As with the other drums, to get a good tom sound, start with a good sounding tom. If your drummer can’t tune the damn drums, fire him and get a drummer with a clue.
Once the toms are in tune, all that needs to be done is to set up the EQ. I usually don’t compress the toms unless they’re integral to the groove or not sitting in the mix well. To make the tom sound richer, just start boosting the bass frequencies around the fundamental, and cutting the mids about 2-4 octaves above the fundamental. For example we used the following EQ on the floor tom in this track: +6 dBs @ 90 Hz, -4 dBs @ 450 Hz. Instant roundness.
Finally, unless you’re recording in a large live room, you’re going to want a little reverb on the drums. Tastes vary widely here and depend largely on the format of the music. In our case, we wanted a fairly dry drum sound, but a little extra “space” on the snare and kick, so only the kick and snare were fed into the reverb. The rest of the kit was dry.
For reverb we used the Cakewalk FX3 Sound Stage Designer. This is a good reverb unit for lifelike room and hall sounds. It has a nice combination of “slap” (natural-sounding early reflections) and “space” (decay) that works well on snare drums.
Unless you’re running a DSP-loaded DAW, by the time you’ve gated, EQed, compressed, and added reverb to the drums, you’re probably approaching the limits of the real-time processing power of your system. Therefore, after dialing in the drum mix, I like to print a stereo track of submixed drums and “off-line” the drum tracks. With older 16-bit DAWs you’re losing a lot of signal by submixing, but not with 24 bit tracks. They retain pretty much all the relevant sonic information in the mix, so I submix 24 bit tracks all I need to keep the computer running nice and fast.
When submixing drums I like to add compression to the drum submix. I can always re-print the submixed track later if I added too much or not enough compression. For drum submixing I use the Waves Renaissance compressor with a gentle 1.5:1 ratio, medium attack and medium release. I set up the compressor to just ride the track with about 1 dB of gain reduction, then start turning up the output and driving the peak limiter until the drums get to the right fatness. Here is a sample of the submixed drums.
Submixed Drum Track Clip (MP3)
A little bit goes a long way. As you get more and more peak limiting, you’ll start to lose the attack of the drums and increase the reverb and ring. A little bit of compression and limiting gives the drums are thicker, denser, fuller sound. Listen to the submixed drum clip. I think you’ll agree that it has good energy.
So now we have four tracks worth of audio: a direct bass, a distorted bass, and a stereo drum track. Time to get busy with guitars!
First up is acoustic guitars. This is easy work with Four Mile Mule. We were recording with a nice-sounding Seagull acoustic with a piezo pickup. For this track I recorded the guitar direct as well as miked with an SM81. We used a rather typical miking position at about 2 feet away from the guitar pointing roughly at the 12th fret. You really need to be careful in setting up a mic on an acoustic guitar as tiny changes in the mic’s position can dramatically change the sound. The direct track and the miked track were panned left and right to create a large-sounding acoustic sound that is more focused than a doubled track. The idea for this song was crisp, jangly energy.
For rock rhythm acoustic, I like to compress the acoustic hard. Heavily compressed acoustic rhythm guitars are something I learned to appreciate listening to a lot of Boston albums. For these tracks I set up a compressor with fast attack, medium release, and a high ratio (7:1). The compressor was set up for about 6-12 dBs of gain reduction.
For rhythm acoustic I also like to cut most of the low-end out of the tracks. A warm-sounding acoustic will have a lot of questionable program material in the 80-240 Hz range, and if this is rock music with bass and electric guitar, you probably don’t need any of it. Getting a well-defined mix involves eliminating what you don’t need, and a bassy acoustic is a good place to start.
You may be surprised by the sound of the acoustic by itself. On other material I might not make it so thin. But on this song the purpose of the acoustic is to add energy with its rhythm, not to fill up the sound with thick-sounding chords. On this song, that’s the role of the electric guitars.
There are three tracks of electrics here: a doubled rhythm track (this is what is playing the chords in the song’s intro) and a third track that plays the hook. The guitarist played a hollow-body Heritage guitar (similar to a Gibson) over an ADA Electronics Rocket amp. The Rocket has been reviewed here on ProRec. If you can get your hands on one, do so immediately. I’ve yet to hear a bad tone come out of one.
In all cases I prefer to record with a low-wattage tube amp. 100-watt Marshall stacks might sound great live, but when recording in a studio you’re going to be dealing with an absurd amount of volume by the time you get the amp loud enough to do its thing. After all you’re dealing with an amp loud enough to fill up a large club, and you’re stuffing that amp into a small tracking room. I find that I get better sounds with smaller amps.
You need to be careful when recording guitar with one amp. In particular, using the same guitar, amp, speaker, and mic over and over can print the same sonic fingerprint into all of your guitar tracks. The result can be that individual parts blur together, and certain frequencies stick out, making the sound “honky” or “pointy” or just plain dead. By using different guitars, and by making subtle changes in the mics, mic positions, and speaker selection, you can get a bigger sound and help the listener differentiate each guitar part, even if the overall tone is more or less the same.
For this song we wanted to get bright, edgy guitar sounds. Four Mile Mule keeps it simple and usually does not double their guitar parts, preferring a more natural guitar sound. But we wanted a powerpop production style on this song, which required a big guitar sound. Therefore we chose to double the main rhythm part using different speakers on each part, a 12” Celestion Greenback and an EMI Eminence 12”. We also slightly changed the amp’s tone and gain to create slight differences in the quality of the distortion. When it came time to record the solo, we kicked the gain up a notch and rolled off a little bass to get a more “pointy” sound that would cut through the mix.
We miked the cabs with single mics, using an SM57 on the rhythm parts and an AT4050 on the solo. The SM57 is a guitar cab standard, and is always an easy choice when miking up a cab. Like the 57, the 4050 also has a presence boost, but it has an extended treble range that the 57 does not, and is less “covered up” and boxy than the 57. I think the 4050 is the perfect compliment on guitar cabs to the 57 and I like using them in conjunction this way.
I like to close-mic powerpop guitars to get them “up-close” sounding. Room or distance miking can really take the edge and attack off the sound which can move the guitars back in the mix. This is not what you’d usually want in a powerpop production. The problem with close miking is that the mic’s proximity effect creates a bassy sound.
Guitarists, listen up. The thick, chunky sound of a close-miked distorted guitar may sound fat and heavy when the track is soloed, but if you throw that track into the mix, it’s probably going to turn into mush. A tough guitar sound will sound surprisingly heavy in a full mix even with a lot of bass rolled off. For these guitars we applied a low-shelf to cut about 4-6 dBs of midbass at about 400 Hz. This dramatically lowered the mud factor and helped the guitars to cut through nicely.
And here again, compression is the key to getting good sound. When you drive an amp into power amp distortion, you’re going to lose a lot of attack. By setting a medium fast attack and a medium release, you can add back a little extra “bite” into the attack and help the rhythm guitars to have more punch. This is the key to the classic British guitar sound, and I compress every guitar track. I have also been known to jack up the compressor’s output until I’m slamming it’s peak limiter to get a more “distressed” sound (for an example of the “distressed” sound, check out the guitar solo on the song “Two Silver Dollars”).
Take a listen to a bit of the guitar mix on Black and White Movie to get a better idea of the tone that’s making it into the mix.
Finally we recorded the vocals. Here I feel that I really fell down on the job. The band’s singer has a fairly sibilant style, really hissing out his S’s. Normally I would take the mic a little further back and slightly off-axis to help moderate the S’s. In this case the mic was already ready to go, so I forgot about the sibilance problem and just told him to get up there and sing his part. What’s worse is that I didn’t catch the sibilance problem until after the track was cut and we were ready to mix.
For the vocals we used the Crown CM700 through the preamps of my Mackie mixer. On this record we also used the AT4050 and a Groove Tubes MD1, but for this song we knew we wanted a bright, sparkly, clear vocal sound, and the CM700 will do just that. The CM700 worked great on some of the other songs on the record, but this time he was singing “right down the tube” of the mic, and was too close, and we got some nasty sibilants happening. If I had caught the problem, the right thing to do with that mic is to back it up a foot or so and/or to angle it about 20 degrees off-center. But, as I said, I fell down on the job.
So when it came time to mix, I was in trouble. With sibilants you often don’t realize the problem exists until after you start compressing and EQing the vocal. Compression quiets the loudest parts, so that the quiet parts can get louder. Sibilance is always quieter than the voice’s tone, so you always wind up boosting the relative volume of the sibilant consonants.
I dialed in a compressor for the vocals, medium fast attack and medium release, with about a 3:1 compression ratio. We used a lot of compression on this vocal, from 6-12 dBs of gain reduction during the loudest parts. For the vocal’s EQ we used a little midbass cut to deal with the proximity effect from having the singer too close to the mic, and added a high boost of a few dBs at 9KHz to bring out some extra sparkle.
By the time I got the vocal compressed to where it was sitting nicely in the track, and EQed to bring out the sparkle and shine in the voice, I knew I was in trouble. The S’s were enough to make me wince. Really the only two things you can do in this situation is (1) re-cut the vocal or (2) patch in a de-esser and do some damage control. We didn’t have the luxury of time to re-cut the vocal, so I used a Waves De-esser plug-in to help tame the most annoying sibilants. Now the harshest sibilants don’t just take your ears out, but their tone is pretty poor.
Ultimately, though, I believe that most listeners do not tune into these brief instances in the song, but rather the vocal’s overall texture during the melody. In other words, I compress and EQ as if the sibilants aren’t a problem, rather than EQ’ing and compressing “around” the sibilants. This is for rock music where the overall sound is dostorted, compressed, and limited. This ain’t “pretty music.” If the take is good and the energy is strong, who cares if the consonants are a little harsh? If I was doing an adult contemporary ballad where the vocal was more exposed, then I might have to re-track the vocal.
There are only a few moments in the song where I notice the sibilants, so I presume that most listeners would never hear the problem. I think that the sound of the voice’s tone is far more important that sibilant control in these tiny moments. So the vocal could have been recorded better, but I doubt that the result we got would actually hold the song back from any kind of artistic or commercial success. Overall, the vocal sound was precisely what the song needed.
When putting together a mix I usually start with drums, add bass, then vocals, then guitars. Sometimes I start with vocals, add drums, bass, and guitars. And sometimes I just start with everything up and start pushing things around.
With me there’s no secret formula. Often the song itself dictates the mixing approach. For example, a ballad usually revolves around its vocal. For a ballad I might start with the vocal track and add the rest of the instrumentation around it. For a guitar-rock song, the bass / drum groove and the electric guitars usually create the song’s energy. In that case I might mix the instruments into a fairly cohesive mix, then set the vocal into the mix.
The key, for me, is to be able to determine what “drives” the song. Is it a killer drum and bass groove? Is it a strong lead guitar hook? Is it a catchy vocal melody? I think a good producer / engineer needs to be able to comprehend the song’s key hooks and build the mix around the hooks, supporting and strengthening them. In the case of Black and White Movie, I felt that there were two key elements in the song’s hook: (1) a powerful, driving guitar riff, and (2) a smooth chorus vocal that is capable of selling the lyric, “Just like a black and white movie.”
This song pretty much mixed itself. The EQed electric and acoustic guitars sat nicely in the mix, with only a little tweaking needed to get them to not be too thin. Remember that when you add the midbass back into the guitar track, the sound actually moves further back into the mix, and you wind up having to turn up the guitars to make them cut through! As intended, the distortion on the bass became almost invisible in the mix and just added the desired movement when the bass player played 8th notes during the chorus. And the drums were dead-on.
We brought the bass distortion out of the mix during the quiet verse and chorus of the song, and dropped the guitars and vocal reverb substantially to create a more intimate sound. When the acoustic guitars come in toward the end of the quiet chorus, we pushed them up a little to add some urgency to the song. But this is all really basic mixing. I think with simple guitar-pop music like this, once you get the instruments balanced, there shouldn’t be a lot of mixing required. And there wasn’t.
Two reverbs were used during the mix. The same Cakewalk FX3 used on the drums was used for the electric guitars. FX3 has a very natural room sound which I think works great on instruments. For the vocals, we used the denser-sounding Waves Trueverb which I prefer on vocals.
Like I said at the beginning, I do not hold this song up as the paragon of sonic excellence against which all other songs must be measured. I think it’s a strong song that’s competently recorded. You can check out the other songs the band recorded for their CD, and get a more complete idea of the other kinds of sounds we got for the recording. In particular I would point out “Gasoline and Wine”, a Stonesy rockabilly song that could be subtitled “Adventures in Compression and Distortion”; “Killer in the Sun”, a very warm folk-rock tune that really shows off the warm sounds we got for the CD, and “New York City” which has a very balanced clean sound with enough edge to really rock out at the end. Take a listen to the rest of the band’s songs and you’ll get a better idea of the sounds that we were able to achieve with essentially the same recording process outlined here.
Hopefully this article helps demonstrate some of the processes and tools we discuss in the articles on ProRec. Pretty much every piece of gear used in the creation of this record has been reviewed at some time or another on ProRec. None of it is particularly esoteric, and all of it is good enough to get some great recordings.
In conclusion, I want to state: it’s the WITCH, not the WAND. The gear manufacturers and retailers want to sell you a better wand. They’ll challenge your self-esteem along the way, daring you to buy their products and insinuating that you’re a loser if you don’t. Don’t fall prey to the marketroids. Great music always begins with a great song and a great performance. Nobody can sell you that. A half-decent recording of a great song will take you much further than an excellent recording of a half-decent song.
Focus on the music, and the rest will take care of itself.