PLEASE NOTE: This article has been archived. It first appeared on ProRec.com in September 2000, 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!
The “Point to Point” article was released in December of 1999. Since then, I have received hundreds of e-mails from readers thanking me for the level of detail provided in that article. And hundreds more with follow-up questions. And even more asking for another Point to Point.
Since December I’ve worked with a number of artists worthy of mention. But I really wanted to write an article focusing on a killer modern rock band, since the majority of questions I am asked revolve around recording modern rock.
So when I got a chance to record the new Happiness Factor CD, I knew I had the perfect band for the followup article.
Happiness Factor is the newest band to emerge loudly on the Dallas music scene. Formed as a kind of local supergroup, Happiness Factor features Salim Nourallah (The Moon Festival, The Nourallah Brothers, Novachrome), Paul Averitt (The Daze, DoublePlusPop), Steve Duncan (Grand Street Cryers) and the incomparable John Jay Myers (Crash Vinyl). The CD also features guest appearances from members of other major Dallas bands. At the time of this writing, Happiness Factor has played only three shows – all sold-out shows in the three most popular venues in Dallas.
They’re clearly a band on a steep trajectory. Moreover, they’re a group of artists who all have significant experience making records. This is a band composed of four people each of whom could easily produce an album. So I was pretty excited – and a little nervous – about producing their debut CD.
Happiness Factor covers solid modern rock territory from loop-based rock to ballads to punk-pop to ultramod. Their sound is all original, but influences include the Attractions, the Clash, Radiohead, and Imperial Drag. Production on the CD ranges from the stripped-down sound of two guitars, bass, drums and vocals to layered indulgent sonic overkill.
Since their CD includes perfect examples of different approaches to producing and recording modern rock, I intend to cover the making of three different songs. Between these three songs we will explore several alternative paths to creating a modern rock sound. Hopefully, by the time we’ve covered them all, you will have an inventory of new ideas that will get you excited about trying new approaches in the studio.
The song we’re going to cover first is “Everything’s a Lie.” To best understand the context of the article, I strongly suggest you begin by checking out the song, and listening to it several times to get a feel for the parts and sections. If you haven’t heard the song you’ll have no context for the rest of the article. We will also augment the article by providing short MP3 clips of tracks and submixes as we build the song from the ground up.
You can checkout the song here on Youtube
And for you licensed cynics, the usual disclaimers apply. This ain’t my band, I didn’t write these songs, and I’m not making any money off their record. Cool?
Setting the Stage (and some boundaries)
It can be really challenging to record a project with an artist who has no clear idea where they want to go.
It can be even more challenging to record a project with four artists each of whom has a perfectly clear idea of where to go.
Happiness Factor is a group of seasoned pros. No, they ain’t Steely Dan, but between the four members of this group is a history of about a dozen or so CDs. Any member of this band could have produced the CD.
Fortunately, the members of this band are also mature enough to recognize the value of a dedicated producer on the project. Especially when there are four people in the band with convincing, valid, and mutually exclusive ideas about the direction of each song. I knew from Day One that I needed to establish clear ground rules about the production of this CD.
I take different approaches with different artists. For example, when working with a solo artist who has a clear direction, I believe my job is to fully understand where the artist is coming from, and to essentially take direction except when my opinion is requested or when I have a very strong opinion. Conversely, when working with a solo artist who has no clear idea where to go, I see my role as collaborator – almost co-writing the songs, helping with arrangements, performing, and engineering.
With Happiness Factor it was clear that my role would be equal parts director, collaborator, and referee. On some songs the live arrangement was ideal – the goal would be to capture a straight-ahead four-part rock band sound. On some songs the goal would be to create an “augmented” four-part rock band sound by adding tasteful overdubs or mixing nuances. And on other songs the band had no clear idea how to achieve the sonic vision present in the music and I would be called upon to take pieces and parts and assemble music based on my conceptualization of the song.
Getting Ready to Rock
One of the things we needed to do before we got started was to make sure that we had the right tools for the job. The first thing you have to realize when recording anything is this: garbage in = garbage out. You have to have a sound source that is capable of producing the sound you want to record before you can set about recording it.
It would seem elementary, yet many newcomers (and some old-timers) in the audio scene are only too willing to cut corners when it comes to getting the right equipment (read: instruments) for the job. Drums have to be in tune and well maintained. If you want a “pro” sound and you have a squeaky kick pedal, you need to get another one. Amps need to be checked for rattles, hum, and microphonic tubes. Guitars need to be in good shape. You need a master tuner that everybody tunes to.
The simple fact is that a little time invested up front in getting the right instruments will result in any easy, brief mix. Good sounds just sound good. I know, it sounds silly. But have you ever sat up late at night trying to get a killer “shattered glass” guitar sound with a Fender Twin and some distortion pedals? Get the right tools. If you want a Marshall sound, get a well-maintained plexi and get it over with. Borrow it if you have to.
For this record we set about making sure we had access to all the gear we needed. Guitars included a 59 Gretch hollow body electric, a Les Paul, an SG, a Magnatone strat, a Martin acoustic, and an old nylon-strung acoustic. Bass guitar was all Fender – J-bass and P-bass. Bass amps included a killer vintage Ampeg fliptop and an Ampeg SVT complete with 8×10 cab. Guitar amps included a Top Hat Club Royale, Mesa dual rectifier, Marshall plexi, Vox AC30, Fender Pro Junior and Fender Twin, Rocket A30R, and, of course, the trusty battery-powered Vox amp with the dual 3″ speakers. The drum kit would be my studio kit – a 4-piece Tama Rockstar kit with Fiberskyn heads, diligently maintained, with Zildjian cymbals and a choice of Yamaha birch snare and a Pearl floating-rim brass piccolo snare.
Setting a Direction
When the band brought me “Everything’s a Lie” we all knew the song had serious hit potential. But it was clear from the nature of the song – intense, repetitive, powerful, modern, somewhat mechanical – that a straight-ahead production approach wouldn’t fly. I had the band play the song for me a few times and knew that I was dealing with a piece of loop rock.
Selling the band on a loop-based production approach was challenging. In particular, drummer John Jay Myers was somewhat put out by the notion. To understand this you have to understand that John Jay is a ROCKER. He’s tall, lanky, powerful, somewhat technical, and plays a drum kit with deadly authority. In suggesting a loop-based production approach, I was essentially ruling out any chance for John Jay to drive the song.
Even though I envisioned the song being created with loops, I wanted to keep true to the sound of the overall project. So, the approach we took was to have the band perform the song live with guitar, bass, and drums. I would then create loops of each, and arrange the song with loops. Then we would go back and re-track extra guitars, effects, and finally the vocals.
In the end, the entire song was constructed using only loops of sounds performed by the band with the exception of two electronica loops used at the end of the song. All of the other loops and grooves were harvested from the tracking sessions with the band.
Tools of the Trade
We tracked the CD at my project studio using the equipment available to me at the time. My computer is, essentially, a Roll Your Own machine with a 533 MHz Celeron overclocked to 800 MHz running on an ABIT BH6 motherboard. It’s a two-year old rig with an upgraded processor. I have an 8 GB OS disk, a CD-R and a pair of removable 20 GB audio disks.
It has served me well.
A number of people asked about the removable drive bays. These are simply THE way to go when working with audio on a computer. The drive bays I have are from inclose (http://www.inclose.com). I think they are a little wimpy, but they work. Kensington has a model that seems sturdier, but they cost a little more. The Inclose models are really cheap. The starter pack costs $20 and includes the drive bay and one cartridge. The drive bay installs into a 5 1/4″ drive space in your computer. You then take any 3 1/2″ IDE drive and mount the drive in the cartridge. The cart pops into the drive bay. Additional carts are $10 each.
So instead of thinking of the hard disk as an expensive, fixed solution, I now consider them cheap, disposable media. For under $200 I have all the media I need to do a full length CD project: all my multitrack audio, multiple versions of mixdowns, and all the work files can be kept on a single removable disk. When the project is done, I pop the cart out and pop in a new one. If you think about it, it’s cheaper than analog tape.
I said it in the last article, and I’ll say it again. A lot of folks are a lot more interested in winning the Big Dick Contest than they are in making music. C’mon people, if you spend all your time and money on the gear, there’s no time and money to make music. I don’t care if I have the fastest computer, the most tracks, the most expensive mics, the largest tracking room, or the most expensive preamp. I just like recording music.
I got a lot of questions about “investing” in gear, so I’ll reiterate the point about “investing” in computer hardware. Computers have the worst depreciation of anything you can buy. A top of the line workstation will set you back $5000. In two years that computer will be yesterday’s news, and will have a street value of well under $1000. But a decent workstation can be had for about $1000. It, too, will be pretty valueless after two years – but instead of losing $4000, you’re only out less than $1000. Here’s my advice: take the extra $4000 and buy yourself a nice mic. It won’t lose its value.
I’ll also take a moment to clarify a part in the last article that left some of you a little confused. If you don’t use a computer to mix, but are still working with tape, listen up! Software effects (plug-ins) are an incredible value. Consider that you can drop $1000 on a nice outboard compressor unit. When it comes time to mix, that compressor can be used on one track (unless you send a submix to it, but it’s still only compressing once). A software plug-in can be used on multiple tracks at the same time. Each “instance” of the plug-in functions as a seperate unit with its own controls and settings. It isn’t unusual for me to use a software compressor on ten tracks at the same time. This would require ten rackmount units!
We used some nice mics on this project. However, I have to confess that I don’t own a Neumann. Mainly because I’m cheap – and, also, because I tend to have decent mics available for review purposes. For this project, our high-end mics were a Rode Classic Valve, an Alesis AM62 tube, an Alesis AM40 tube, a Groove Tubes MD1a, a Groove Tubes MD2a, and pair of AT 4050s. Preamplification was provided by a dbx 576, an ART Pro Channel, a MindPrint Envoice, a Joemeek VC6Q, an Aphex 107, and a Mackie 1402. We used an M-Audio Delta 1010 for tracking, and a Waves L2 as an extra set of converters to create a 10-input system. We used Cakewalk Pro Audio 9 as our recording software.
Some of this gear is pretty good, the rest is clearly “prosumer.” One of the points I want to make in this article is: it’s the witch, not the wand. True, to make an audiophile CD, you need audiophile equipment. Of all the gear used, only the MD2a, dbx 576, and L2 could fall into the “audiophile” category. Some might suggest that even these pieces aren’t “audiophile.”
However, we’re not making an audiophile CD. We’re making rock and roll, and frankly, good enough is good enough. The real question is: does it rock?
Remember: it’s the ear, not the gear.
When recording rock rhythm tracks, I prefer to have the band all play together. That way the drummer and bass player are hearing all the cues they need to make performance decisions. For example, when a drummer doesn’t hear the loud guitars he’s used to hearing, he’ll back off and won’t play hard enough. Then you add in the guitars and wonder what happened to the energy in the drum track. I also like to attempt to get a keeper rhythm guitar track whenever possible. The guitar that’s cut together with the bass and drums will have the raw energy of live performance.
For this record we wanted to get a natural, versatile drum sound. The goal was to capture a “studio” drum sound – tough and slightly larger than life, but not overly gated or processed, and with a nice dose of real room sound.
The drum kit was set up in my main tracking room, which is L-shaped and connected to a hallway. The L-shape and the attached hallway provide some variance in the reflections – some short, some long, some simple, some complex. The room has a wood floor with rugs. Equipment cases and shelves provide some diffusion.
This is a room with character. It seems to be a goal in some schools of recording that tracking should be done in a room that is somehow sonically neutral (whatever that means). Not as far as I’m concerned. To me the worst problem is recordings without character. In particular, when recording rock drums, I really prefer a live-room sound. I like to be able to hear the whole kit breathing and speaking. That’s why I don’t overdamp my tracking room. I’d rather hear too much room sound than not enough.
For this recording session we used my session kit, a 4-piece Tama Rockstar kit with Zildjian cymbals and a Yamaha Custom Birch snare. The toms are outfitted with Remo FibreSkyn III heads which, to my ears, have a far more natural sound than the typical plastic heads. I also use large toms tuned low when recording rock. Small toms sound great live, and work fine on some kinds of music, but for rock, I prefer a big tom sound.
As for brands of drums, as far as I’m concerned, any decent tom or kick drum, properly maintained, can get a good sound on tape. The real exception is the snare. Snares are magical instruments that need lots of loving care and attention… and even then, some simply never make the cut no matter what you do to them. We had five snares on hand to audition, including Yamaha, Pearl, Tama, and DW models, and decided that the Yamaha was our favorite. It had less character than some of the other snares, but it sounded great. Don’t get me wrong. I like snare drums with character. But a quirky snare drum can get really out of hand in a mix. Take a look at the EQ work we had to do on the Four Mile Mule project and you’ll see what I mean.
When you compress and limit a snare drum, you lose attack in order to beef up the body. If the snare has ring, then the more you compress and limit, the more amplified the the ring will become. Since we were doing some serious rock and roll I knew I’d be using a lot of compression and limiting – maybe distortion – on this drum mix, and I wanted something that would sit nicely in the mix no matter how badly I tortured it. And thus the Yamaha snare.
I used a fairly typical mic arrangement for rock drums: Crown CM700MPs overhead in an X-Y configuration, Shure SM57 on snare and rack tom, AKG D12E (the Brick) on floor tom, Shure SM81 on hi hats, Audio-Technica AT25 on kick. Sometimes I will use a mic under the snare, to get a more “snarey” sound if the snare is too dark. But the Yamaha snare had plenty of snare sound and a strong attack, so we didn’t use a bottom snare mic.
In setting up mics for a rock tracking session, I usually take the same approach: a pair of overheads as well as a mic on each drum. If I know I am going for a particular funky sound, then I might take a different approach. On this project we were going to cut a dozen songs at a sitting, so I wanted a versatile configuration. So I used the tried-and-true approach.
The overheads are either a spaced pair or X-Y. Spaced pair mics offer more separation in the mix, whereas an X-Y configuration will offer a more focused and tighter sound. When using either, I make sure that the snare will be “centered” between the mics. This keeps the kit sounding balanced. On this project I knew we’d want to err on the side of natural and focused, so I chose the X-Y configuration.
The 57 on the snare usually goes under the hi hats pointing at the center of the snare and aimed directly away from the bottom hi hat. This reduces hi hat bleed, which can quickly be a problem if I need to boost the treble on the snare. The tom mics are aimed toward the center of the drum and mounted approximately 2-3 inches away from the head. The hi hat mic will go over the hats aimed toward the far side of the hats. This minimizes bleed from the snare.
Preamplification for the kit used the onboard preamps of my Mackie mixer for the overheads, hi hats and tom mics. The kick was preamplified with a dbx 576, with just a little limiting and compression. The snare was miked with an ART Pro Channel and amplified until I got a nice tube-limited sound. The slight distortion fattens and thickens the sound. Remember, in this approach the snare mic is only used to beef up the overheads, so if the sound of the snare mic is somewhat exaggerated, that’s OK. I do not usually make EQ decisions when miking the drum kit and all mics were recorded flat.
When dialing in the drum mix, I start with the overhead mics full-up and usually completely dry – no EQ, reverb, gates, nothing. When I start like this, I’m pretty much guaranteed a fairly natural drum sound. The typical sound of a dry, flat overhead pair on drums is “natural to a fault.” So I then start adding in the close mics. But I EQ, gate, and compress the close mics so that they fill the in the spaces. So if the kick is a little wimpy, rather than jacking up the bass on the overheads (which would tend to muddy up the sound of the whole kit) I’ll create a fat kick mic sound and dial a little of that in.
I usually start with the just the overheads. I then EQ (and usually compress and gate) the kick and add enough kick mic to beef up the sound. Now I listen to this three-mic drum mix for the snare. Is it ringing? Is it too dark? Not meaty enough? After I think I’ve identified the direction in which the snare needs to change I’ll EQ and compress the snare mic and add a little in. Then I’ll usually EQ the hi-hats and add them in for a little extra sizzle, and gate, compress and EQ the toms if they need a little extra meat.
To hear how this all builds together, take a listen to the overheads by themselves. You’ll notice that while they seem boring, they sure sound “real”. You can hear the whole kit. Now add in the kick. I already had a good sounding kick. Miked with the AT Pro 25, I usually will roll off some bass below 40 Hz until the kick sounds “tight”, and I may also pull out some low-mids so that the kick isn’t too “thick”.
For this song in particular I wanted to get extra slap – I really wanted to hear the attack of the kick drum – so I provided a pretty serious presence boost around 4KHz. Remember that this track isn’t supposed to stand on its own – it’s supposed to augment the overhead mics. Listen to the kick track and get an idea of how it sounds soloed.
Now I add in the snare. The Yamaha snare was nice and punchy, with a little ring in it. A couple of notch cuts with the Waves EQ pulled the ring out, and a smooth presence boost pulled the snare forward in the mix. Note that for this snare, the EQ did not need to be overly radical.
After the EQ comes the compression. For a snare like this I prefer to use the Renaissance compressor. I typically get about 6 dBs of compression at an approximately 4:1 ratio, then drive the limiter hard until just before the signal starts to break up. Driving the limiter hard makes the snare meatier, but you have to make sure you don’t remove the attack completely. Finally I added a little Arboretum Hyperverb to provide some extra “space” on the snare. You can listen to the snare completely dry and after EQ and compression to get an idea of the effect I’m going for.
Putting the Drums Together
When mixing drums in Cakewalk, I dial in the sound I want, then submix the drums to a stereo track. This frees up CPU resources for further processing. Also, since Cakewalk doesn’t have a proper submix bus layout, it allows me to use a Renaissance compressor on the entire drum submix, something that I almost always do. After dialing in the drum sound, I’ll drop a Renaissance compressor on the master outputs and get a few dbs of compression and a few more dbs of limiting. This results in even more weight and volume on the kick and snare.
Remember this is a drum mixing approach I take for in-your-face rock drums. This is not a drum prescription. There are many applications where you’d want much less compression and limiting, and fewer open mics. When you compress and limit the drums as much as we’re doing here, they fill up a lot of the available space.
For this arrangement, we wanted a dry, tight, almost overly focused drum sound on the verses, and a larger, thicker sound on the choruses and the bridge. The purpose of this arrangement was to keep the verses sharply focused in order to emphasize the grandness of the chorus. To achieve this effect we used very little reverb on the drum track, and created a dry drum track for the entire song. Then I cut out the bridge and chorus to another track where I ran them through a Cakewalk FX3 room simulator and another Renaissance compressor. I added enough room effect and compression to start to approach a Bonham-like size, but I didn’t want the transition to be distracting. I think the result works: the verse drum track is tight and focused, the chorus and bridge have more meat, and yet the transition doesn’t call attention to itself.
Almost all of the drum tracks for this record were recorded in a single long session. John Jay is a beast who does not tire, and he delivered the goods on 12 songs in essentially one sitting. Between John Jay’s excellent performances and the drum kit we used, we got an exceptional drum sound on this record. Since it is loop-based, “Everything’s a Lie” is probably not the best example of the drums that we recorded – for a more compelling example download “Facelift” from MP3.com and check out that sound.
The bass was recorded with a Fender Jazz Bass over an Ampeg SVT amp and Ampeg 8×10 cab. We miked the cab about three feet back with the AKG D12E. When working with a bass amp – especially one as massive and loud as an SVT full-stack – I will ALWAYS get a direct sound as well. A loud amp loads a room, making the entire room resonate. Strange and unfriendly sounds can result unless your room is designed to handle extreme bass SPLs. In this case we were driving the SVT well into distortion, so it was LOUD.
The bass direct sound was preamplified with a Joemeek VC6Q “British Channel”. This is a great preamp for direct guitar and bass applications. For recording I used just a little compression, to add sustain to the notes, and rolled off a little high end to keep the bass from sounding too metallic. On mixdown, I added a Waves Renaissance compressor to get just a little compression and limiting.
When mixing the direct and amplified bass tracks, there is always delay between the direct sound and the amplified sound. If you’re not careful, the two sounds will be out of phase with each other, which really messes up the tone. Even a little phase shift can affect the tone. Therefore I will advance the amplified track and line it up to be coincident with the direct track. The result is a larger, more focused bass tone that stays in control better.
Another factor to consider when recording a bass amp is settle time. Cabinets resonate. Speakers continue to vibrate for a moment after the signal stops. Amps also continue to put out signal after the input signal stops, especially when overdriven. And loud amps cause the room to resonate. The combined effect is a lack of definition – it takes a moment for the entire system to fire up, blurring the attack – and even longer for the entire system to get quiet, slowing the release. Notes lose definition and the important “black spaces” between the notes become filled with nonmusical garbage.
For this application I really only wanted the distortion and attack characteristics of the amp. I ended up using EQ to cross the two tracks over at around 200 Hz so that the direct track provided the bass below 200 Hz, and the amp, lower in the mix, provided the “edge” above 200 Hz. You can really hear the amp on the bass track to “Everything’s a Lie”, but the bottom end is tight, not boomy.
Cuttin’ Guitar and Other Odd Noises
More than any other song on this project, “Everything’s a Lie” makes use of layered guitars and oddball noises. Most of the other songs on the CD, by comparison, are straightforward two-guitars-bass-and-drums arrangements.
The drums on this track were cut live, then cut into loops and reassembled into the final arrangement. The bass was cut live to the looped drum parts. Then the guitars were cut live, and either looped or used as-is.
Paul played a Gibson Les Paul over a Vox amp with a variety of distortion effects. Faris played a late-50s vintage hollow-body Gibson over his Top Hat Club Royale amp, using nothing but natural amp distortion. The warmth of Faris’ parts helps to offset the harsh, gritchy sound of the ultradistorted Vox.
We recorded the Vox amp with an AKG D1000E mic. This is a thin-sounding dynamic mic that is great for controlling overly bright or muddy guitar amps. I rarely have to use any EQ or compression on a guitar cab recorded with a D1000E, and it was perfect for knocking some of the Vox’s brightness down a notch or two. We miked the Vox very close to the speaker – maybe six to twelve inches – to capture all the attack of his part. The Top Hat amp was miked with an SM57, in a room with a tile floor and wood walls. We backed the mic off the amp a couple of feet to minimize the proximity effect and to capture some of the reflections the room was putting off.
I had some questions about the last article regarding my preference for the D1000E. Most of the comments were to the effect that the D1000E is “thin” or “brittle”. Other comments suggested that “everyone knows that a 57 is THE mic to use on guitars.” To my ears, the D1000E complements the SM 57 perfectly – where the 57 tends to be a little bright and has a strong proximity effect, the D1000E tends to be midrangey and rolled off on top. So I usually will use both if I can, so that one guitar’s color complements the other. But any time both mics are used, the D1000E tends to be a little more out front, since it’s thinner-sounding and tends to cut through better.
The first guitars you hear in the song come in at the chorus. There are three guitars that make up the sound of the chorus.
First off, we had Paul play around with some heavily distorted parts until we found a part that worked, then we simply created a loop of that part and used it where necessary. The effect is a chopping, stabbing part that sounds mechanical. Paul also played a double of that driving loop with less distortion and with an orange MXL phaser. You can hear the overdistorted and the phaser guitars here.
On the first two choruses, Faris is playing “punctuation” parts that emphasize the chord changes. For the final choruses, we have the same arrangement as the previous choruses, except that now Faris is playing stabbing eight notes along with Paul to really kick the song into high gear.
In the second verse we have some seriously out-of-control crap. I had Paul plug his already-overdistorted guitar into a Vox wah-wah, and he just made strange noises and modulated the wah. This resulted in some great screams and squeals. The sound is teetering out of control, coming apart at the seams. These out-of-control guitars also make a reappearance after the bridge, before the final choruses.
Also in the second verse is a part that sounds like a sequencer, but is really a guitar run through a square wave tremelo. This kind of tremelo is either “on” or “off” and it sounds very staccato. I ran this part through a Waves Mondo Mod to make it bounce around between and outside the speakers. If you listen to this tremelo guitar track soloed you can hear the attack as Paul hits the notes on the guitar.
For the bridge, a number of guitars work together to produce the sound. Faris and Paul are both playing arpeggios through the chords, each on their own guitar rigs providing both warm and thin sounds. After the song had been recorded, we decided to bolster the guitars on the bridge by adding a pair of guitars that play the chunky chords at the end of each phrase. The result is big but doesn’t sound overly layered.
Now we come to the end of the song. Here we have the same arrangement from the last chorus, plus we’ve added in another guitar. As before, we ran this guitar through a bunch of distortion boxes and finally a wah pedal. All it was capable of doing is feeding back. Paul just slowly rolled the wah from bass to treble, and the guitar just screamed higher and higher. We used a long delay and hall reverb to make it wash out over the mix and end the song. This wailing guitar track made the song’s ending much more dramatic.
We used absolutely no EQ or compression on any of the guitar parts. This is amazing, especially since my typical approach would be to start by rolling off at least SOME bass, and then seeing how much compression I can get away with. For this song, though, I just didn’t think EQ or compression was necessary. I guess if you live long enough, you really do get to see everything once.
Maybe some explanation is in order. Let’s start by tearing down the first assumption: to get a heavy, distorted guitar sound, you need a lot of distortion. FALSE! Fact is, a heavily distorted guitar amp can often LOSE aggression when recorded. It turns to mush. Even the ultradistorted sound of much of today’s modern rock is created with amps that are probably playing less loud (or with less pedal / preamp distortion) than you’d guess. When you turn an amp up past its sweet spot, you lose attack, definition, and warmth. You start just getting white noise.
My preference is usually to find the LEAST amount of distortion required to make the part speak. Then I use compression to thicken the part up and make it denser. The result is warmer, thicker, and heavier than you get by turning the amp up. It has more tone and less garbage. Believe me, I love distortion. I love the high-end sparkle of a Marshall “shattering glass”. But that high-end noise will absolutely step all over everything in the mix – and end up making the guitar thinner and wimpier. So, with distortion, less is more.
On this song, though, we broke ranks and got really distorted. In hindsight, too distorted, maybe. For example the looped chorus guitar doesn’t have as much attack as I’d like. At any rate, after all the tracks were cut, I didn’t hear anything about the guitars that suggested they needed EQ or compression. They mixed themselves.
For this project we had access to five vocal-quality mics, and I wanted to make sure that we chose the best tool for the job. Of course, all mics perform differently depending on their placement. Placement is not really a question with Salim. Salim’s voice is very rich, and is utterly stuffed with all kinds of strange harmonics and overtones. These overtones contribute to Salim’s vocal sound, and must be properly captured if the vocal is to sound right.
Part of the magic of recording Salim’s vocals is to get him right on top of the mic. The further he is away from the mic, the less apparent those overtones become. So we tested all of the mics available using a close-miking approach. Salim was less than six inches away from the mic at all times, and got as close as two inches at times.
The easiest mics to eliminate were the Audio-Technica 4050, the Alesis AM62, and the Groove Tubes MD1a.
The 4050 is a great mic on some vocals, but my experience with using it on male rock vocals is poor. It is a little too pristine to make the vocal have appropriate size and weight. It’s response to sibilants can be a little metallic as well.
The AM62 suffers from the same trend we’ve seen in a lot of new “home studio” mics. Assuming that the buyer is going to make a purchase decision with a pair of headphones in a booth at the local GearSlut store, designers of many new prosumer mics are trying to get the mics to sound “more” — more bass, more sparkle, more oomph, just more. The AM 62 suffers most of these traits – it has a weighty proximity effect, scooped-out mids, and a harsh top that makes sibilants really stick out. We used the AM 62 on “Loveable” and while that track had some nice traits, you can hear the sibilant problem.
With the MD1a the problem was reversed. The MD1a is really warm and compressed, and didn’t quite have enough oomph for this vocal application, although I’ve used it on other vocals with great results. In this case the song simply had so many distorted guitars and crashing cymbals that the slightly dark MD1a would need too much EQ to sit up front, and with all that EQ, the track might not sound very poilte.
That left us with the Groove Tubes MD2a and the Rode Classic Valve. The Classic Valve is a real piece of work. It has some of the most unusual high frequency peaks and notches you might ever find. It is anything but flat. But it does sound good, and that strange high frequency response can make vocals sound really wet. The MD2a, by comparison, is strangely pristine and reasonably flat, but is a little larger than life and adds a nice controlled sparkle. We cut most of the vocals with these two mics. In the end, when all the songs were mixed and it was Miller Time, I decided that I liked the tracks cut with the MD2a better. To me they sounded more natural and up-front. Your mileage may vary.
On this song we used the MD2a. Preamplification was provided by a dbx 576. I set up the 576 to provide a little compression with a 2:1 ratio, and set its limiters to engage only if something got out of hand. As it turns out, something did get out of hand. On the bridge, Salim got unbelievably loud and slammed the limiters. Lit up both limiters and also overloaded the preamp. I was concerned that we might have blown an otherwise great take, because the 576 does not overdrive gracefully. Fortunately, the touch of distortion actually accentuated the roughness of the vocal, and it all went together well in the mix.
For the verses, we applied an effect that sounds distorted, though we did not use an amp or amp-sim. To achieve the effect of the verses, we applied a “telephone style” filter with a Waves Renaissance EQ, cutting the low end at about 700 Hz and cutting the high end at around 3000 Hz. We then applied a Waves MetaFlanger to gently flange the vocal, making it move around in the mix. Finally, the distortion was created by a Renaissance compressor. We ran the compressor hard, getting about 12 dbs of compression with a 6:1 ratio, and overdrove the output until the limiter kicked in hard.
The vocals were tracked close to the mic and we ended up with more proximity effect than I had intended. So, for the choruses, we used a litte low-shelf EQ to reduce the bass about 8 dBs at 180 Hz. Compression was provided by a Renaissance compressor set up like the compressor on the verses, only the output was lowered so that the sound wasn’t so overdriven. We also wanted a big reverb effect on this vocal, so we dialed up an Arboretum Hyperverb with a long pre-delay. This makes the reverb sound delay-like, and also increases the apparent size of the vocal.
This song was a mixing effort from the countoff to the mixdown. Most loop-oriented projects are. The process of recording first, mixing last really breaks down when you start constructing a song with loops.
To begin with, we started by recording and mixing the drums. Once mixed, the drums were hacked into loops and rearranged. A bass part was also constructed with looped bass. At this stage in the recording, the arrangement of the song was still in question. I took some of the scratch guitar tracks, cut them up into loops, and build the arrangement. All of this mixing and arranging was used just to arrive at an arrangement.
Next Paul came back into the studio to cut his guitars. Paul’s guitar worked well with Faris’ scratch guitar, so we decided to keep Faris’ guitar. After Paul’s work the guitar mix was pretty close to complete.
Then Salim came back in to recut the bass. In order to keep the song from sounding too loop-ish, the bass was not looped. After bass we cut the vocals on the song. At this point we had finally arrived at a mix that sounded close to being finished.
The band had urged me to really take some mixing chances, and to add loops, effects, or whatever I thought worked. In some ways that’s like giving me a license to screw up their song!
Truthfully, I felt like we had already taken some cool chances. Moreover I really wanted to avoid overly “Rip-izing” the song. Knowing that the song already featured a lot of production work, I wanted to err on the side of underproducing. Remember that when you’re making an album, you have to think about each song’s relation to all the other songs. Most of the songs on this CD are straight-ahead rockers, and we already had thrown two or three cool curveballs. I really didn’t want to throw a wild curveball with this song.
So, once we had a good “guitars – bass – drum” mix, I brought the mix into Acid to audition some loops. In the end I found a lot of loops that sounded really cool, but I didn’t find any loops that were particularly meaningful to the song. In other words, there were great loops that were different than what the band had already recorded, but there were none that were particularly better.
In the end I used two electronica loops – old workhorse loops from the Acid installation disk – to spice up the ending of the song. They added a nice color and movement to the song without taking away anything that was already present. The average listener might listen to the whole song and never really notice the loops. That’s the way I wanted it.
After getting the song close to a final mix stage, we decided that the bridge didn’t have enough drive. At this point the “accent” guitars hadn’t been cut. Paul suggested re-cutting the guitars, but I didn’t think there was anything about the original guitars that needed changing. I thought we just needed a little extra punctuation when the vocals stopped. So we added the extra guitar parts on the bridge.
It was only at this point that I decided to make the bridge vocals stronger. First we cut a backing harmony vocal. Next I added reverb and delay to the vocal. With reverb and delay on the vocal, the song started to take on a somewhat cheesy quality, like so many 80’s hair-band songs. Rather than going backward and removing the effect, I plowed onward and used a reverse reverb effect.
Building a reverse reverb effect in your DAW is easy, if your DAW supports destructive editing or if you have a good editor. In this case I had already brought the vocal into Cool Edit because I didn’t know what effect I was going for and just wanted to play with the vocal until something hit me.
First, I converted the vocal from mono to stereo, so that the reverb would be in stereo. Next I reversed the vocal so that it played backward. I then added a generous reverb, 100% wet, with a long 2-second reverb time. Next I compressed the reverb so that the tails were hotter and the loud parts weren’t so loud. Finally I reversed the audio again, ending up with a nice reverse reverb effect. This effect track had to be reimported into Cakewalk and properly aligned so that it was correctly timed with the lead vocals. The final effect is strong and over-the-top, and we think it worked well.
The only part of the song that I still wasn’t happy with was the ending. This song was already tagged as the album closer, so I wanted the ending of this song to be really dramatic, something that would retain the over-the-top quality of the bridge and the screaming guitar that takes the song out at the end. In this case the answer was to take some of the guitar-noise tracks we had cut earlier into Acid and mess ’em up.
Like other time-stretching tools, Acid works by looping tiny portions of audio. When used reasonably, it does a great job of stretching (or compressing) time up to 25% or even 50% of the original. In this case I took an 8-beat loop and told Acid that it was 64 beats long – 800% of the original. This forces Acid to decimate the loop and reassemble it into completely new and original sounds. The ending of the song is some wierd guitar noise that has been tortured into a completely new sound by Acid.
Because this song was constructed one part at a time, it was essentially mixed by the time we quit adding tracks to it. Essentially the construction process is the same track-at-a-time process I usually use when mixing: dial in the drums and bass, add guitars, add vocals, polish to taste.
These songs are all being sent for mastering for their appearance on the final CD. For the band’s MP3.com EP, we decided to forgo mastering and simply get the songs “close enough.” The question is: what process should one use if mastering is going to be skipped? To me the answer is obvious: compression and volume leveling. Any other process – EQ (especially EQ), reverb, multiband compression, etc. should be avoided like the plague. If you hear an EQ problem, go back to the offending track and change the EQ there. So for the songs on Happiness Factor’s EP, we just patched a Renaissance compressor across the output bus of Cakewalk and applied enough compression and limiting to make the song a little hotter. For mastering, we will remix without the compressor on the output bus.
I fully intend to produce more articles about the making of the Happiness Factor CD. I would certainly appreciate it if you, Valued Reader, would take a listen to more of their songs, and let me know if there’s a song that you would like to know more about. I hope that these articles are constructive.
As before, I don’t want to appear egotistical with any of my work. It’s not as if I think that I hold the keys to the secrets of recording. It’s just that I know that whenever I watch someone else work, or read about how some other project was produced, I always come away with my imagination fired, ready to try some new things. In many cases, I am not learning any new techniques but rather being reminded of techniques I once used and forgot about. In other cases I do learn some new approaches, and sometimes I get a completely new idea to try out.
So use this article as food for thought, take some new chances, and get out there and make some art!