Equal Time

Oct 1, 1999 12:00 PM, Jeff Casey

The fact that every professional mixing console comes with an equalizer on each channel speaks volumes about the importance of EQ in the recording and mixing process. Mixing a record without EQ would be pretty difficult: we'd be forced to rely on capturing a pristine performance from a perfectly tuned instrument, in an acoustically correct room, using expensive and accurate microphones that are properly placed in relation to the sound source. How often does that happen?

So we have an arsenal of tools to help compensate, and EQ is arguably the most important one. But although every console has EQ, and most studios offer some sophisticated outboard units, not every engineer knows how to properly equalize a sound. Let's correct that problem by discussing the basics of EQ and the different types of equalizers and when they're traditionally employed. I'll also offer some specific applications that you may find useful when tracking or mixing your project.

THE GOAL OF EQ To get some outside input, I called on two professional audio engineers from New York City. Derek Martin (who hastens to point out that he is no relation to Sir George) is a freelance music engineer who floats around the Big Apple, hitting studios such as Quad Recording, Waterfront Sound, and Apex Recording. His resume is impressive, with credits ranging from Kool and the Gang to Nine Inch Nails. Greg Petricelli works at Midtown Digital and specializes in audio for video. He recently completed work on commercials for Nike and IBM. These two engineers provided valuable insights into how equalization is approached in different applications.

Before we can talk about specific applications, though, we have to ask the basic question: what's your ultimate goal when using the EQ process? Petricelli notes, "The purpose for using an equalizer is to change the frequency content of a signal. However, the reasons for changing the content vary. Usually it's to coax the very best sound out of an instrument."

In most cases, though, an instrument that sounds great by itself also needs to fit into the context of a mix. "Getting a track to blend in a mix so that its frequency response doesn't interfere with any other tracks is the very goal of the mixing process," says Petricelli. "Usually you want to eliminate frequencies that aren't needed by that particular instrument." But, Martin is quick to add, "oftentimes you'll use EQ to get a track to stand out in a mix, above the rest of the tracks-like a vocal, for example."

You can also use EQ to fix minor flaws present in a signal. For example, a 60 Hz AC hum can be notched out of an electric-guitar track, or you can use a de-esser to filter out unwanted sibilance in a vocal track. Although many sound designers use EQ creatively to sculpt sounds into new and different ones, for the purposes of this article we'll focus on how to use EQ within the context of a mix, be it for a music CD or a television commercial.

ALL SHAPES AND SIZES Equalizers are essentially filters, although they can often boost as well as attenuate signals. Both active and passive EQs are available, and some products use a combination of the two designs (for example, passive high and low shelving bands and active mid bands). Active EQs traditionally employ op amps to boost and cut frequencies. Most console EQs are active, as are most outboard EQs except for a handful of high-end processors.

Passive filters cost much more than active ones and are, therefore, not commonly found in personal studios. They traditionally use coils or capacitors to attenuate frequencies; strictly speaking, they never boost. The tone controls on most guitars, for example, are passive filters. However, a passive EQ may have an amplification stage before or after the filter (or both). That means you can effectively boost a frequency band by cutting everything on both sides of it and amplifying the overall signal before or after cutting.

The most common, and the most useful, studio equalizer is the parametric EQ. A true parametric EQ affords you control over three parameters: gain, frequency, and bandwidth or Q (see Fig. 1). Until recently, most parametric EQs were outboard units, but many digital mixers now use them, as well.

Quasi-parametric EQs have gain and frequency controls but only a limited number of preset bandwidths. Just as common is the semiparametric, or midsweep, EQ. Semiparametric EQs have gain and frequency controls but no bandwidth control.

Shelving EQs work at either end of the frequency spectrum, attenuating or boosting all frequencies above or below a specific cutoff, or knee, frequency. These equalizers are so named because the graph of frequency and amplitude looks like a shelf (see Fig. 2). Shelving EQs typically offer gain and frequency controls. There is usually a preset rolloff that gradually slopes from the knee frequency to the shelf; however, some of the more elaborate units available allow you to adjust this ratio.

The channel strips on many analog consoles have high and low shelving filters and a midrange bandpass filter with sweepable center frequency. Lower-cost mixers often have shelving filters for the highs and lows and one fixed-frequency bandpass filter for the mids. Some consoles also offer a "rumble filter," which is a low-cut shelving filter with a preset cutoff frequency.

Graphic EQs, although commonly associated with home audio equipment and P.A. systems, are also useful in the recording studio. Graphic equalizers divide the frequency spectrum into a number of evenly spaced bands with fixed bandwidths and with gain controls for each band. The result is a number of fixed-bandwidth bell curves that can be boosted or attenuated as needed (see Fig. 3). Graphic EQs are usually used to "tune" the monitoring system. There are times, however, when you might want to patch in a graphic EQ on a track so that you can play with several frequency bands simultaneously (more on this later).

EQ ETIQUETTE Although most people prefer to EQ during mixdown, some insist on equalizing during tracking. Martin, Petricelli, and I agree that dramatically changing an instrument's frequency content prior to recording is usually not the best approach. "You want the instrument to sound natural and good going to tape," Martin explains, "and this involves a whole number of things. For example, you have to make sure that the instrument is a quality instrument that's tuned properly; as they say, 'garbage in, garbage out.' You also have to make sure you've miked the instrument properly, with the right capsule and from the right angle. But aside from that, I see no benefit in making sonic changes at this point; you're not going to know what you need to tweak until you hear the track in context with the other tracks."

However, Petricelli also points out that there may be certain instances when you want to address a problem before the signal gets printed. "If I had a high-frequency buzz coming through an amp," he says, "I would certainly want to get rid of it before recording-and if that meant using a notch filter, then so be it. It's less to worry about later."

That brings up the issue of adjusting the EQ on instrument amps. Both engineers agree that this falls into the category of tuning an instrument. "You wouldn't leave an out-of-tune tom-tom to be fixed in the mix with EQ," Petricelli points out, "so I see no reason why that doesn't apply to a poorly equalized guitar amp." Martin says, "Obviously you'll also be tweaking the amp track's EQ in the mix, but you still want that amp to sound great in the room, with the mic on it."

Another debate that inevitably arises is whether additive or subtractive EQ benefits a mix more. In certain instances only one approach would be appropriate; for example, using a highpass filter to roll off everything below 40 Hz is definitely a necessary application of subtractive EQ-there's no other way to go. But some stalwarts (especially jazz and classical engineers) take a more extreme position, insisting that nothing should be added if it isn't there already. These engineers use only subtractive filters. On the other hand, some rock and pop engineers simply turn up the knobs until they like the sound.

Most engineers, however, combine both philosophies and use whatever approach works best for the situation at hand. Petricelli explains: "I usually start equalizing using a subtractive approach, turning down the gain of a parametric EQ and sweeping the frequency knob until I hear something that needs to come out. As soon as I do that, though, I may realize that I need to add a few decibels at a higher frequency to compensate. I'll then repeat the process, with the gain of another band turned all the way up, sweeping the frequency knob."

FOURTEEN EQ TIPS Now that I've discussed some basic principals of EQ and the various types of equalizers, let's look at some specific EQ applications and the ways in which you can approach each. Most of these tips require at least a 3-band parametric equalizer.

Because drums are usually the hardest instruments to equalize, I've devoted some space to them. We'll start with the percussion instruments; work our way through string, wind, and horn instruments; and finish with the human voice and special applications. Keep in mind that you can apply these tips to digital samples, as well as to acoustic instruments.

1. KICK DRUM A kick drum is one of the most important components of any mix, because it drives the beat (assuming you're not working with a chamber orchestra). But you must determine the type of character you want that beat to have.

First, you need to listen to the sound of the drum as captured by the microphone. This approach is imperative for any EQ application-you should always listen before you tweak. If the kick has a naturally hollow tone and is tuned correctly, you have a bit of room to maneuver. If the drum is padded and produces more of a "thud," your options are more limited, although you can still dial in a number of "kicking" sounds. Remember, if you're lucky enough to be involved in the entire recording and mixing process, you have the liberty to go back in the chain and make desired changes (like removing a muffler from a kick drum); if you're simply mixing a project, however, you must work with what you have.

Here we'll discuss three specific kick-drum sounds, all of which you can dial in if you have a track with an unpadded kick drum. Although these tips give you good starting points, you'll need to make tweaks because each drum is different. The first sound I call the '80s Big Hair Kick Drum. You know the sound: punchy, with lots of midrange and a bit of a thump.

Petricelli describes how he typically obtains this somewhat dated sound. "If you have a 4-band parametric EQ," he relates, "you start by rolling off everything below 60 Hz. This will eliminate rumble. Moving up one band, boost between 78 and 84 Hz by 3 to 6 dB, with about a 1.0 Q factor. Again, you may need more or less, depending on your track. This will provide the kick that hits you right in the chest. Next, dial in the punchiness between 1.5 and 2.5 kHz, boosting by about 6 dB. Here, a bandwidth of 1.5 to 2.5 should work nicely. Finally, notch out 120 Hz by about 4 dB with a 1.0 Q setting. Play with the parameters until your kick drum sounds like a White Lion record. If you have only a midsweep EQ with high and low shelving bands, try boosting the mids around 2 kHz by about 6 dB, turn the high shelf up between 4 and 6 dB, and boost the low shelf by about 2 dB."

Although this sound was popular in the 1980s, today we've returned to our roots, and engineers usually strive for a classic rock kick-drum sound, a la John Bonham. "The Bonham kick drum is the quintessential rock drum sound," Martin explains. "I usually obtain it by boosting the frequencies between 120 and 240 Hz by about 4 dB or more. You'll also need to roll off everything above 1.5 kHz. Sometimes, depending on the drum, you also might want to notch out 80 Hz a bit-not too much, just by 1 or 2 dB. Then add a little bit of 60 Hz, but again, just by about 2 or 3 dB."

A lot of alternative records have a hollower, more gritty-sounding kick drum. You can dial in this sound by rolling off everything below 100 Hz-yes, you heard me right. Boost 125 Hz by about 3 dB, and add about 4 dB between 250 and 350 Hz. Then roll off everything above 2 kHz.

2. SNARE DRUM Two of the most widely used snare-drum sounds in popular music are a tight, punchy snare and a loose, full-sounding snare (usually used on ballads). Naturally, the way the drum is tuned will influence how you decide to EQ it.

A snare never needs frequencies below 150 Hz, so roll them off. The center frequency of a snare is usually around 1 kHz, give or take a few hundred hertz, so it usually benefits any snare to boost around that point by approximately 3 to 6 dB.

For a tight-sounding drum, Martin explains, "you'll want to work with the upper mids, around 5 kHz, and some higher frequencies, like at 8 or 9 kHz. Try boosting in each area, starting with 3 dB and working your way up. Also, you'll need to sweep the frequency parameters until you nail it. Attenuating low frequencies is a good idea, so get rid of anything below 250 Hz, and if you still have an EQ band left, roll off everything above 11 kHz. A tight gate usually sounds good on this type of snare."

"A ballad snare," Petricelli notes, "needs a bit more bottom end, so try boosting around 250 Hz to give it a fuller tone. I'd start with an increase of 6 dB. You won't need to boost as much high end as you would with a tight snare-though adding a little around 7 kHz might be helpful, and you can roll off everything above that. The key element here is your center frequency: you don't want too much ring to the drum, but you want it to have some ring. Find the resonant frequency of the snare by turning the gain all the way up and sweeping between 800 Hz and 2 kHz. When you find it, narrow in on it with the Q control. Now you can adjust the gain accordingly. A loose gate and heavy reverb complement this type of snare sound."

3. DRUM KIT Drum overheads can be a bit tricky, and how you EQ them depends on how the entire drum kit was miked. "If you have individual mics on the tom-toms, snare, and kick drum," says Martin, "you only need the overheads to capture the cymbals and to blend the kit. In this case, you could probably just use a low shelving filter to reduce the amplitude of everything below 4 kHz. This way, you have individual control over the snare, toms, and kick, and the overheads take care of the rest. Depending on what mics were used, you might have to boost some high frequencies to get a decent cymbal blend-probably between 10 and 14 Hz by about 3 dB."

These days, though, a lot of engineers record drum tracks with only four mics: a kick, a snare, and two overheads. This gives the drums a more natural, almost raw sound, but it makes equalizing them much more difficult. In this case, you need to keep the overheads' low end in place to sufficiently capture the toms, but not so much as to add boominess to the kick or snare drums.

Let's look at an example from a project I recently completed. The drum overheads were recorded pretty flat-not bad, but not all that exciting. After playing around with the console EQ, it became clear that we needed some outside assistance, so I patched in a CLM Dynamics Expounder. To liven up these two tracks, I started by adding about 10 dB at 150 Hz to bring out the toms, which were overshadowed by the cymbals. I then notched out an offending ring at 250 Hz produced by the middle tom. In addition, because the snare had a prominent ring at 1 kHz and we already had a dedicated snare track, I pulled out 1 kHz by about 6 dB with a 1.0 bandwidth setting, which also helped to round out the toms. Finally, although we had plenty of cymbals in our mix, I added about 3 dB at 9.5 kHz, just to smooth things over.

A similar example can be seen in Figure 4. Here everything above 1.5 kHz sounded great, but again the toms were being washed out. In this instance, I used the console EQ, boosting the low shelf by 8 dB at 191 Hz to add some low end. This also augmented the kick drum. I boosted 3 dB at 152 Hz, and I notched out a tom ring at 666 Hz and a snare ring at 1,017 Hz.

4. HAND PERCUSSION Percussion comes in many forms, but here we'll focus on the more common instruments, such as shakers, tambourines, congas, and hand claps.

Shakers and tambourines fall into roughly the same category with regard to EQ: both need to be bright and should cut through the mix on the high-end side. Martin discusses his philosophy on hand percussion: "For shakers, I usually just roll off everything below 2 kHz and add a bit of high end, like 6 dB at 9 kHz. That usually works fine. Tambourine requires a bit more 'clanginess,' so for it I roll off below 800 Hz, boost 1.5 or 2 kHz by about 4 dB, and add a little around 7 kHz."

Petricelli often records congas for his commercial work. "If recorded properly," he notes, "congas don't really need that much tweaking with EQ. I usually find the resonant frequency of the drum by doing an EQ sweep, and either add or subtract a little depending on the situation. A boost generally helps to bring out the natural timbre of the drum. But you need to be careful not to add too much, especially if the resonant frequency is low, as it will clash with the other drums and the bass. To bring out the attack of the drum, I usually boost a bit in the midrange-say, around 5 kHz, maybe by 6 dB. You can roll off everything above and below the range of the drum."

Finally, hand claps are generally fattened up by adding some low mids, usually around 250 Hz by about 2 dB, with a 1.5 Q factor. Also, to get more of an attack, boost some of the midrange (around 1.5 kHz by about 4 dB) and add some highs (around 8 kHz by 2 or 3 dB should suffice).

5. ACOUSTIC PIANO Piano EQ can be approached in two ways, depending on how much accompaniment the piano will have. Will it be the main instrument in the mix, perhaps with a vocal and bass? Or will it be in the back of a full-band mix with seven or eight other instruments?

Martin explains, "If you have a solo piano, with little or no accompaniment, and it's been recorded properly, you really shouldn't have to do much to it. Often, I like to boost a little low end, around 140 Hz, but only if there's no other bass instrument in the mix. Also, try adding a bit of high end, around 8.5 kHz-but not too much, just about 3 dB.

"If, however, you have to fit the instrument into a tightly packed mix," he continues, "you'll need to do some subtractive EQ. A piano naturally has a 'honky' sound because much of the playing is done in the middle of the keyboard. This is what you want to watch out for. Yes, you want the midrange to be out in the mix, but you don't want it to sound obnoxious, so you may have to do some attenuating around 3 or 4 kHz. Also try rolling off the frequencies below 140 Hz, because they will definitely clash with your bass guitar and kick drum. A slight boost around 8 kHz can bring out the upper keys. All in all, how much you boost or how much you cut will depend on the nature of the mix and its components. Try to coax the midrange and upper mids from the piano, but walk a fine line so that it doesn't sound honky."

6. BASS GUITAR With electric bass guitar, you have a lot of options, especially if the track was recorded with a direct injection (DI) box. One of my favorite bass sounds is illustrated in Figure 5. This track was recorded with a DI box, and we wanted a sound that brought out each note being played, without "slappiness." We started by rolling off everything above 520 Hz and below 100 Hz. We used a boost of 6 dB at 260 Hz to fatten up the tone, and a boost of 3 dB at 730 Hz to add some finger noise. This EQ, along with some moderate compression and limiting, produces a rich, fat, full tone that works well with any type of music.

For a player who slaps and punches and needs a funkier sound, you could use the same EQ settings, but instead of rolling everything off at 520 Hz, boost some mids around 2 kHz by about 4 to 6 dB. This provides the slap. You might want to decrease the low-end rolloff to about 50 Hz to give the track more rumble.

Often a graphic EQ can be used to zero in on a bass sound. "I usually use a graphic EQ on bass tracks," says Petricelli. "If the track has been done direct, this is almost a must, because you don't have the liberty of equalizing the player's amp, moving the mic, and so on. I use a 20-band EQ that can dial in just about any bass sound I want. The nice thing is that with a graphic EQ, you can see the curve you're drawing by looking at the unit. The tips for getting a sound are the same as with a parametric EQ; you just have more frequency options."

7. ELECTRIC GUITAR As with the piano, how you treat the electric guitar depends on the other elements of the mix. For example, if you have only one electric guitar in a mix with just drums and bass, you can make the track sound large. However, if you have three other electric guitars, a piano, keyboards, percussion-the works-then you have to fit it nicely into a spot on the soundstage. Let's look at a couple of such scenarios.

Let's say that you can do whatever you want to this guitar to make it sound as big as possible, as long as it doesn't conflict with the bass guitar. Martin notes that "if I'm working with a simple rock band that has only one guitar player-you know, The Who Live at Leeds-I'll try to fatten up the low end as much as possible. Usually, a boost of 3 dB around 160 Hz is a good start, as long as it sits well with your bass. Also, try adding a little around 700 or 800 Hz. Depending on the sound of the amp, you might need to pull out some midrange; if you do, it will generally be around 3 kHz. How much high end you add depends on the sound you're going for: a boost of about 6 dB at 7 kHz will give you a crunchier sound." Again, it's always good practice to roll off anything above or below the frequencies that are not needed; but listen carefully before you cut, lest you accidentally eliminate desirable, but subtle, low-level subharmonics and high harmonics.

For our second scenario, let's assume you have to fit this guitar into a mix with a full band, including two other electric guitars. My advice is to focus on the midrange. Roll off everything below 200 Hz and above 9 kHz. Then do some sweeps in the middle to see what needs to go in and what needs to come out. I find that a boost around 4 kHz and a cut around 6 kHz generally works, although sometimes the opposite is true. You have to listen to the track and decide according to the mix at hand. If you have more than one electric guitar, make sure that they sound a little different from each other. This could mean using different guitars, amps, mics, miking techniques, or processing. There's nothing more grating than three guitars playing the same thing, all with the same sound.

8. ACOUSTIC GUITAR To illustrate how dramatically the EQ of an acoustic guitar can change from one mix to another, let's look at Figures 6 and 7. Both EQs were applied to the same track, but in different mixes.

Figure 6 shows an acoustic guitar track that has been worked into a mix with drums, bass, piano, electric guitars, percussion, and lots of vocals. This track was recorded with a very nice Martin acoustic guitar miked with an Audio-Technica AT 4033. Here, I rolled off everything below 90 Hz and added a bit of body around 360 Hz. I really boosted the middle and upper frequencies, though: 10 dB at 2 kHz and 9 dB at 7.1 kHz. With everything else happening in this mix, the acoustic guitar had to sound bright-and it did.

Now look at Figure 7, which shows the same guitar track, only within an alternate mix of acoustic guitar, vocals, and percussion. This time, I decreased the low-end rolloff to 43 Hz to give the guitar a fuller sound, and I made a 3.4 dB boost at 166 Hz to achieve the same goal. The midrange boost has been moved to 3.1 kHz and is at only 7 dB, and instead of increasing 7.1 kHz with a 9 dB bell curve, I've used a smoother shelf to gradually raise the high end.

Martin notes that strings or string patches follow a similar approach. "A string section has many of the same tonal characteristics as an acoustic guitar," he explains. "I usually roll off most of the low-end content, assuming that the strings will be placed in a full mix. I then boost the upper mids a bit around 4 dB at 7 kHz, and maybe give the sample some 'airiness' by boosting with a shelving EQ around 10 kHz."

9. BRASS AND WINDS Some horns are naturally strong in the midrange, so bringing out those particular frequencies is important. For instruments such as the trumpet, roll off the low end completely, around 200 Hz or even higher. Depending on the recording or sample, you might have to tame some midrange to avoid a honky sound.

Conversely, with brass instruments that rely on the low end, you might want to roll off the upper frequencies-say, above 9 kHz. Usually a boost in the low mids helps bring out these instruments in a mix, probably around 1.5 kHz. When working with tuba or with tuba samples, be sure to use a filter to roll off frequencies below 40 Hz; this way you ensure that no speakers get blown, especially during low notes.

Most wind instruments need to be airy, which commonly involves frequencies above 9 kHz. Usually a shelving EQ works well to augment this characteristic. A bassoon can produce extremely low notes, so don't filter out its low frequencies; but for many wind instruments you should use a low-frequency filter.

10. LEAD VOCALS A vocal track must have body, presence, and air, but not so much as to interfere with the rest of the mix. Vocals also need to be out front in the mix, and this is usually accomplished by boosting the midrange. Petricelli elaborates: "Vocals are what sell the product, whether it's a song or a sung commercial, so they need to be audible but not annoying. It's a fine line to walk." Understand that vocal tracks are the most sensitive you will have to EQ, and because so many variables are involved, making categorical statements about vocal EQ is difficult. You'll have to listen and decide for yourself.

It's important to note, however, that there is a big difference between recording male and female voices. "I just finished a session in which we had two singers, a man and a woman, who were doing a duet," Martin explains. "They were recorded in the same room, with the same mic, roughly ten minutes apart. When we went to mix, the tonal differences were amazing."

Based on what Martin told me, I re-created the two EQ settings shown in Figure 8. The tracks were recorded with a Neumann U67 mic and run through a UREI LA-2A limiter. This illustration should give you an idea of how vocals are typically equalized. The male singer's track appears on the left EQ, and the female's track is on the right.

Immediately, you'll notice that the male vocal gets an upper-frequency boost of 1 dB with a shelving EQ, while the female vocal requires a 3 dB shelving cut at 8.8 kHz. The male vocal also needs a 2 dB boost at 7.5 kHz and a 5 dB cut at 5.1 kHz. Martin told me that this was because the singer had a cold and sounded a bit nasal. Conversely, the female singer's track was treated with two 4 dB low-end boosts-one at 733 Hz and another at 283 Hz-to make its low-end content complement that of the male singer's track. Interestingly, both vocal tracks got a similar boost around 2.5 kHz, illustrating how important it is to bring out the midrange in a vocal track.

11. BACKGROUND VOCALS "There are two kinds of background vocals," Martin points out. "The first is your standard harmony that sits an octave above the lead and comes in on the chorus. This is almost like another lead vocal, and I generally EQ it using the same approach. The other type of background vocal is the choral effect, in which you may have three or four backgrounds, at all varying octaves, panned across the soundstage. For this I use a different approach: these vocals typically sound better when they have a bit of 'air' to them and sound almost unearthly.

"To get this effect, I usually roll off low frequencies on high harmonies, usually around 400 Hz, and attenuate extremely low frequencies on low harmonies, around 100 Hz. Then I take out as much midrange as I can without losing the clarity of the voice-try filtering out anywhere between 1 and 4 kHz. Finally, I add a lot of upper mids and high frequencies. I usually put a shelving EQ on around 8 or 9 kHz, sometimes higher, and boost it until I have that angelic effect."

12. NARRATION Obviously, if you're working with a narrative that doesn't need to blend into a music mix, you can fatten up the sound of the announcer's voice. Petricelli commonly does this.

"In this case," he explains, "you can leave a lot more low end in the voice than you could if you were doing a music mix. Announcers are supposed to have big, godlike voices, so you need to make them sound that way. Start by boosting a bit between 60 and 120 Hz, depending on the tonality of the voice you're working with. Then go all the way up top and add some high end, probably around 7 kHz. You might need to pull the mids back a little, but make sure you don't lose any clarity. It's somewhat similar to the way you'd set your home stereo."

13. REMOVING ARTIFACTS As mentioned earlier, filters and EQs are great tools for removing sonic garbage. The most obvious example of this is using a filter to notch out a 60 Hz AC hum. However, you can use a notch filter to get rid of almost anything, and you can use a parametric EQ to create a notch filter. Martin relates a story.

"We were doing a session in which we had a percussionist playing the congas" he says. "When we listened to the playback-months after he had originally cut the track-we noticed an odd clicking sound that came up every so often. We couldn't figure out what it was. It turned out that the guy was wearing a ring on his right index finger that occasionally hit the lug on the side of the drum. We couldn't redo the track, so I thought I'd try to EQ it out with a parametric. By turning the gain all the way down and gradually sweeping the frequency knob and zeroing in with the Q control, I was able to isolate not one, but three different frequencies that were causing this click. We eliminated the noise without affecting the sonic characteristic of the drum."

14. ONE EQ FOR TWO Often, especially in the world of digital audio workstations, sound designers wind up with more than one instrument on a particular track. If your automation package allows you to automate EQ changes dynamically, there's no problem: you simply switch from one EQ setting to another at the appropriate time in the mix. If your software doesn't offer this feature, however, you'll have to get creative, or at least reach a compromise.

Let's look at another example from a project I recently did (see Fig. 9). One track contains a lead guitar, background vocal, and hand claps. The vocal and the hand claps were recorded using the same mic (an AT 4033) and had a similar sonic content-but, as you can imagine, the guitar sounded quite different. So we needed to make some compromises. The first thing I realized was that none of the instruments needed anything above 15 kHz, so I gradually rolled it off at 11 kHz. I then boosted the midrange a little at 4 kHz to bring out the vocal; that, unfortunately, also made the guitar sound clangy. So to compensate, I added some low end for the guitar, around 300 Hz, which didn't really affect the vocal in a bad way. I then boosted about 2 dB at 8 kHz for some clarity to the guitar and vocal. Lo and behold, with all these changes, the hand claps sounded fine!

The moral of the story is that if you have to comp tracks together, make sure the instruments on them have similar frequency responses.

TWEAK IT, BABY! An EQ can either ruin a recording or make it shine; it's a very powerful tool. That power is harnessed by the ear, experience, and knowledge of the person operating it. The first two things can't be taught in an article or a classroom, but the tips offered here should be helpful to anyone, whether you're a novice or a pro. These solid approaches to EQ have been used for many years by seasoned engineers. But that's not to say they're set in stone; like anything else in this business, creativity is the key.

EM associate editor Jeff Casey has a 10-band fully parametric EQ on his car stereo.