Mixing, like any art, can't be reduced to a formula. All that counts is the finished product, which means there is, literally, any number of ways to proceed. Just the same, I have developed a somewhat systematic approach to mixing that usually works for me and for the types of music I tend to mix (the standard rock instrumentation of drums, bass, acoustic and electric guitars, keyboards, vocals, and so on).
In this column I'll share some of the thoughts and techniques that go into my approach, including a few gleaned from friends and colleagues. None is radical, nor is any writ in stone as the only way; indeed, experienced mixers will probably be familiar with most, if not all, of my methods. But if you're relatively new to mixing, or if you consistently struggle with particular problems in your mixes, you may find something of use here.
IT'S A STYLE THINGMix appropriately for the musical style. That may sound obvious, but what we're really talking about is musical vision-that is, having an idea, in advance, of how the final product should sound. Essentially, mixing is about getting the right balance of levels. Just as different styles of music employ different instrumentation, the relative volumes of the different instruments vary from style to style. Therefore, it's important to be familiar with the style of music that you're mixing, to know which instruments get foregrounded, which backgrounded, and so on. Of course, it's okay to BREAK the "rules"-but it sure helps to know them first.
For music with vocals, the level of the lead vocal is typically a defining element of the mix style. In heavy rock, for example, vocals tend to get mixed low, as if they were just one of the instruments-indeed, they may even be mixed lower than the main guitar tracks (check out Rage Against The Machine). Commercial country music, on the other hand, tends to feature the vocals; no instrument is allowed to distract too much from the voice and lyrics. Vocals are prominently featured in pop, too, yet slightly less so, on average, than in country.
Other defining elements of certain mix styles are drum and bass levels. In jungle, house, and other "electronic" dance styles, for example, an emphasis on the bass and rhythm elements is common. Consider, too, how drums and bass are featured in styles such as reggae (especially the modern stuff), hip-hop, and, well, drum 'n' bass. Classic rock 'n' roll, on the other hand, often showcases electric guitars or keyboards.
THE BIGGER PICTUREWhen you start work on a new mix, rather than solo the individual instruments and commence equalizing, compressing, and so on, try first to make the song sound great using only fader levels and pan positions. This approach-the art of mixing stripped to its essentials-will not only get you focused right away on the bigger picture, but it will also help you hear what's working and what isn't. For example, you may find that certain instruments are fighting for the same sonic space. In that case, try panning them far apart in the stereo field. Later, you can differentiate them further with EQ or what have you; for now, though, keep working to make the song sound as good as you can using only levels and pans.
One secret to a great mix is knowing how to select and highlight the one or two instrumental elements that make the song rock hardest (an aesthetic choice that usually relates, by the way, to the aforementioned style considerations). Once you determine the element (or elements), the rest of the mix can be built around it. In other words, great mixes are not usually "democratic," in the sense of each instrument having an equal say, but instead tend to employ a hierarchy of levels.
In vocal-based songs, the challenge often is to find the right balance between the foregrounded elements and the lead vocal. Instrumental-based songs, on the other hand, are usually a bit easier to manage, as typically there is only one instrument to spotlight.
SQUEAKY CLEANClean up individual tracks with mutes, gates, or however. To do this, you have to solo each track (in particular, those with instruments that come in and out of the mix), listen carefully for any unwanted noise, and then squelch it. If you have automation (either digital or analog), programming mutes or gates is the quickest and easiest approach. If you mix without automation, you can use outboard gates and/or expanders. As a last resort-say, if you don't have enough gates to go around-you can always erase sections between parts. (This is risky, so make a copy of the track first in case you screw up.) The goal, of course, is tracks that are free of extraneous sounds-coughing, lip smacks, chair squeaks, and so forth.
You should also check for unwanted noise on continuous-playing tracks-for example, amp hiss on an electric-guitar part. In the case of amp hiss (or other constant noise), try patching in a single-ended noise- reduction unit, such as Behringer's Multiband Denoiser SNR2000 (see Fig. 1).
THE GREAT EQUALIZERA mistake that novice mixers often make is boosting every band of EQ on every channel. Though this may initially make the signals sound "better"-at least to inexperienced ears-the seeming improvement is largely due to the signal being louder. (Remember, boosting EQ adds gain.) But there are other reasons not to go this route. Not only does it fairly defeat the idea of equalizing, but on analog mixers-at least those typically found in personal studios-it also usually adds circuit noise to the mix. Indeed, with every band boosted on every channel, the compounded noise can be considerable. (To hear for yourself, stop the tape, turn up the monitor-room output, and listen while switching the channel EQs into and out of the signal path.)
Whenever possible, then, cut rather than boost-or use a combination of cuts and boosts-to get the sound that you want. Cutting lets you eliminate undesirable aspects of the signal without adding noise; afterward, if necessary, you can add make-up gain with the fader. In general, experienced engineers end up using a combination of cuts and boosts on any given channel.
For those unfamiliar with subtractive EQ, it's instructive to spend some time experimenting with sweepable mids. First, dial in a radical cut-say, 15 dB. (If you have control of the bandwidth, set a narrow Q, as well.) Now, turn the sweep knob slowly from hard left to hard right while listening to how the cut affects the signal at different frequencies. Next, do the same thing with a 15 dB boost and compare. What you're listening for is how a cut on one side of the frequency spectrum results in a sound quite similar to that provided by a boost on the other side. For example, a low-mid cut (depending on the source, of course) can result in an apparent brightening of the signal-much like what you would get by boosting the high mids.
Here's another important point: soloing an instrument to EQ it can be helpful, but the only thing that really matters is how the instrument sounds in the mix. Therefore, be prepared to change the EQ once all the other instruments are brought in. Often, after an instrument is equalized appropriately for a mix, it sounds quite unappealing when soloed.
Still another trick is to EQ the effects rather than the instrument itself. I find this to be especially helpful when I'm happy with the quality of the signal (that is, how well the instrument was recorded) and I want to alter the tonal balance only subtly, without messing up the naturalness of the sound. You can EQ the effect either inside the effects unit (advantageous, because digital changes don't add circuit noise) or by patching the return into a channel strip and equalizing from there (which is nice, because it allows you to readily pan the signal or even compress it, if you want).
ROOM WITH A BLENDChoose effects carefully and limit the number of like effects. For example, few mixes need more than three different reverbs. Indeed, numerous hit songs have sounded great using only one. The first thing I look for-especially if the song was built overdub by overdub in a makeshift personal studio-is a global reverb that complements the whole mix. The idea here is to create the illusion of everything being in the same space. Halls and rooms are good starting points; in general, it's best to save the more unusual effects for individual instruments. Typically, I send every channel to the global reverb (except, perhaps, the bass guitar's). To hear what I'm doing, I initially crank the effects return, and I may even solo it to better hear the size, decay, and overall coloration (dark or bright) of the space. Once I've dialed in the most complementary settings, I turn down the returns to the desired level.
Just as it's helpful to spotlight a particular instrument or two in the mix, it's usually a good idea to apply supplemental effects in a similar fashion. That is, don't put a different, outrageous, and equally glaring effect on every instrument in the mix. Instead, pick one or two elements to focus on and concentrate your more creative use of effects there. You don't want to confuse the mix with too many competing elements.
When using multiple reverbs, another thing to watch out for is phase distortion and cancellations from multiple stereo returns. One of the worst offenses can come from accidentally reversing the left and right return cables from one of the processors going into the board. Talk about a comb-filtered jumble! But even with everything properly cabled, it's not uncommon to get masking and other strange interactions when combining stereo reverbs (whether they are true stereo or not).
Fortunately, it's easy to test for this problem: simply A/B the mix in stereo and mono. Going to mono almost always causes some reduction in stereo effects-often a great deal. But the thing to listen for is the phasey stuff or the disappearance of critical content. If you find a problem, try systematically returning each effect in mono (that is, by disconnecting one of the stereo return cables-starting with the right, as left is typically the preferred mono connection). This process of elimination should turn up the worst offender. Once you've identified which box is the problem child, you'll probably find that you can get by just using it in mono, especially in a dense mix. In fact, some purists make it a point to use only one stereo effect in a mix; the rest are returned in mono.
LEVEL HEADAfter dialing in all the compression, EQ, and effects, return your attention to the overall balance of levels. In the end, this is what will make or BREAK your mix. There's an art to perceiving how best to balance levels. A killer mix has an energy that seems to jump out of the speakers. It feels visceral and sounds loud-no matter what the volume. All of the instruments and elements can be heard, yet some illusion is going on in that the foregrounded elements don't sound foregrounded; they just fit into the mix perfectly, sounding neither too soft nor too loud-sort of like when an image comes perfectly into focus through the lens of a camera. Such is the power of a well-arranged hierarchy.
As you get closer to finalizing settings, note that the slightest fader move can cause a significant change in the balance of elements. For example, say you've brought up the lead and rhythm guitar tracks and found a really great balance between them and the lead vocal. But now, listening back, you think the drums sound a little low-they're just not quite as slamming as before. So you push up the kick and snare a bit, say 1 or 2 dB each. That doesn't sound like much, and yet suddenly the whole mix is thrown off because the guitars are no longer out front.
Mixing, you see, is a very delicate matter. So rather than pull more things up to compensate, figure out what you brought up too much and nudge it back down. (Note, too, that as the majority of faders cross the nominal point, it becomes increasingly important to pull some down as opposed to continuing to push them all up, or else, before you know it, you're the victim of fader creep-that insidious state in which all of the faders have crept to the top of the board!)
POINTS OF REFERENCEOne of the most helpful things you can do-especially if you're new to mixing-is compare your mix to songs from a stylistically similar commercial CD. Though this can be done at any point in the mix, I usually wait until late in the game, if only to see how far I can go on my own. On the other hand, if you don't have an immediate sense of direction for a mix, listening to a few finished mixes may give you some ideas.
Be careful when referencing, though: there's really no standard in mixing, so your mix could sound quite different from someone else's and still be very good. For me, the point of referencing is to make sure I'm in the ballpark-not to tell me how to throw the ball. I find referencing to be most useful when making final decisions about levels for bass, vocals, and effects.
QUIET ON THE SETAs we've all heard time and again, it's important to monitor at various levels, including loud, soft, and a couple of points in between. But the single most helpful monitoring trick that I've learned is to listen not only at very low levels (75 dB or less), but also in mono. Hearing the mix quietly from a single point allows you to really home in on levels and the overall balance of elements (not to mention how well you have succeeded in differentiating between those instruments that were initially competing for the same sonic space). When you can hear everything clearly this way, you know you're close to nailing the mix.
The "quiet mono" trick is also a great way to monitor if you're having to ride levels manually (and let's face it, even automated levels are done manually at some point). In fact, I often run my mixes this way while printing to tape. Of course, only the board is in mono-the signals going to the record deck are still stereo.
THE MORE THE MERRIERThe ultimate test of a mix is that it should sound great no matter what it's played back on. Therefore, the more speakers you can monitor with, the better. This can be an expensive proposition, but it doesn't have to cost an arm and a leg.
The cheapest solution I know is to use a boom box as your secondary monitor source. In my studio, I mult the mixer's stereo outputs through a half-normaled patch bay and run one stereo pair to my DAT deck and a second pair to the CD inputs on the boom box. When I want to monitor through the boom box, I simply turn down the control-room volume, flip the boom box to CD mode, and adjust the output from the board accordingly. Clients appreciate this, too; it lets them hear how their songs will sound "in the real world." (The real world is a boom box, right?)
If you can spring for an extra pair of speakers and an extra power amp (or a second set of powered monitors), a tidy and convenient solution is to install a line-level selector between the mixer's outputs and the inputs to the power amps (or powered monitors). The Coleman Audio LS3 Balanced Line Level Selector (see Fig. 2) costs only $120 and provides three sets of outputs for this type of application. And if you're looking for a pair of small, inexpensive secondary speakers, check out Radio Shack's Pro-X44AV monitors ($59.99 each), formerly known as Minimus 7 and then as Optimus 77. These are great-sounding little units, and I know several pro engineers who swear by them.
Of course, the multiple-monitoring technique needn't stop with the speakers in your studio. Once you've printed a final mix, run off a cassette copy or burn a disc, and go listen to it in your car stereo, your home stereo, your Walkman, and anywhere else you can play it. (I've even known people to take their mixes into high-end consumer-audio stores and listen to them there.) And don't forget the "around the corner" trick. That is, listen to your mix from one room while it's playing in the other-a technique that can reveal imbalances you overlooked during hours of close-field monitoring.
Time is an important buffer, as well. When I'm doing a critical mix (that is, of a song that's going to be on a CD), I always try to allow a day to "sit on the mix." I send everyone home with a copy (including myself); we all make notes about what we like and don't like; and the next day we come back, make final tweaks, and print the mix again.
INCREASE YOUR ODDSJust as photographers will shoot several rolls of film to get one usable shot, you owe it to yourself (or your client) to print several versions of the final mix. That way, you increase your odds of getting the "perfect" mix. Essentially, this is another way to increase the "time buffer," so you can wait until mastering time to choose the mix that works best for the overall project. There's nothing like some perspective to help you make the best decision.
I usually print at least four versions of a mix. The first is the one that I think is right. The second is the same as the first, but with a bit more bass. The third is the same as the first (that is, I pull the bass back down), but with a bit more vocal. And the fourth has both more bass and more vocal. In addition, if there's time, I may print four more mixes using the same approach, but each with more effects. (I tend to go light on effects; however, if you like to mix heavily with effects, you may want to print extra versions with the effects returns pulled down a bit.)
Be sure to keep clear notes (including start and stop times) about what distinguishes each mix. Professional mastering is expensive. You don't want to show up at the mastering house with twelve different mixes of each song, but no clue as to which is which!
Brian Knave is an associate editor at EM. Special thanks to Barry Cleveland and George Petersen.
Though it might seem odd to tout a piece of gear in a column about mix tips, I would be remiss not to mention a unit that graces nearly all of my mixes and truly helps them sound better. That unit is the BBE 462 Sonic Maximizer. (Note that the 462 was recently replaced by the 482, which costs $349 and is said to sound smoother and warmer and have more headroom. The higher-end, balanced version is the 882, at $599, formerly the 862.)
The Sonic Maximizer provides two different processes: linear phase shift and dynamic EQ. I often use a bit of the EQ, but for me it's the phase shifting that makes the difference. The Maximizer is said to restore harmonics to their original time arrangement-a relationship that gets "flattened out," so to speak, by speakers. The Maximizer remedies this flatness by introducing a linear phase shift across the entire frequency spectrum (20 Hz to 20 kHz), effectively slowing down the lower frequencies so that higher harmonics reach the ear before lower ones. The result is a clarification of detail and an improved sense of dimensionality; after engaging the BBE Process, as it's called, I can better hear "around the edges" of the instruments.
BBE Sonic Maximizers are extremely easy to use. But the 462, 862, and previous models employed a somewhat counterintuitive "Lo Contour" control for which unity-gain was not indicated. On those units, unity-gain for Lo Contour was around 11 o'clock-a design that allowed for both low-frequency cutting and boosting. However, the newer models have dispensed with this design and no longer provide the variable low-frequency cut (which no one seemed to use anyway).
Here's a trick I learned from George Petersen, editor of Mix magazine. I had spent many weeks recording a bunch of tracks using really fine mics and preamps. To get the best sound, I bypassed the board and recorded direct to tape (ADAT). I was meticulous, and I spent lots of time getting things just right. Everything went to tape sounding great-clean, crisp, and really tight and defined.
Once I had mixed everything together, though, I noticed a loss of clarity. To pinpoint where the loss was taking place, I subtracted things from the mix one by one-effects, EQ, compression, and so on. But none of those maneuvers remedied the situation. I finally began to suspect that the mixer was muddying my signals.
Petersen recommended that I bypass the master-output section of the board. After all, he explained, at best there are only two op amps in the master output bus (and in a budget mixer, none of the op amps is that great to begin with). So I bypassed that section of the board and noticed an improvement right away. Granted, it was a subtle improvement-but any improvement in clarity is better than none at all.
Bypassing the master-output section of your mixer is simple-that is, as long as the mixer provides stereo- master insert points. Simply remove the cables from the 1 1/4 4-inch stereo outputs (the ones that go to your 2-track record deck) and insert them halfway into the stereo-master insert points. (It's critical that they be inserted only halfway, or else no signal will get through.) Of course, this means that the master fader no longer has any effect. But as long as you're mastering (or editing) the finished mix later, you can do your fades then. (Note that it helps to have a variable input control on your 2-track record deck for this application.)