You’ve done your part. You researched the best products and successfully put together a working studio. You’ve studied hard to master your studio tools—each menu, knob, and switch. You’ve laid all the tracks for your first masterpiece. You’re ready to begin mixing. Suddenly you come to the realization that all those manuals haven’t even begun to teach you how.
In a way, mixing starts much earlier—during every phase of the recording process, when each step you take affects the manner in which your sound puzzle fits together. But in mixing, you have to deal with all of the decisions you’ve made, and with all the problems you’ve swept under the rug since laying the first track.
Art and science
Mixing is an amazing amalgam of art and science. Fortunately for most musicians, the science is increasingly deeply imbedded in the tools—these days we deal with fewer hands-on technical chores than in the days of alignment tapes, machine calibrations, and oscilloscopes.
The art of the mix is up for grabs, of course, as always a matter of taste and personal preference. A small (but noisy) segment of the population abides by no rules. Their delight is to make every LED light and every needle pin, regardless of the sonic mayhem that ensues. The point is that there really aren’t any rules, just guidelines for creating sounds that suit your purpose, and commonly accepted practices.
If your ambition is to create high-fidelity mixes, there are some things that all of you will have to watch, things like not lighting that last peak LED, and making sure that your stereo tracks are in phase (otherwise you’ll have trained engineers craning their necks when they listen to your work, and sounds disappearing when your tracks are played back in mono). Also it’s good to remember that the more sounds you dump into the soundfield, the fuller it becomes, reducing the listener’s ability to clearly distinguish individual elements and so forth.
Establishing the goal
As a mixer, my goals differ depending upon whom I’m mixing for. If I’m mixing my own music, my goal is to please myself and I adjust the sound accordingly. If I’m mixing for a client, unless otherwise instructed, I believe it’s my responsibility to deliver a mix that satisfies their artistic intent. For me to learn their taste quickly, I’ll often ask them to cite commercially available CDs as an example of a sound they like. If they’re local, and the condition of the track suggests a variety of possibilities, I’ll often mix with the client present.
In any case, once I’m clear on where I’m headed, the task at hand is to take the multi-track recording and mix it into a two-track format that can be delivered to a mastering engineer. Choosing a two-track format is no mean feat. So many choices abound. Just a few of the pathways I follow in this regard include:
• Bouncing at the session’s resolution within Pro Tools, then burning that as a data file onto a CD I can hand to a mastering engineer.
• Bouncing within Pro Tools and choosing sample rate/bit depth conversion to conform to the standard CD audio format.
• Using the Pro Tools stereo analog output and routing directly to an analog or digital stand-alone recording device.
• Fanning out individual channels into a console or analog summing device (such as the Dangerous 2-BUS), then routing that to a two-track recorder.
And so on…consider your own options when making these decisions.
Once I understand the artistic direction and the technical specifications of the delivery medium, it’s time to get to the real task at hand—polishing the collection of related tracks into a musical experience. People create music for a variety of reasons. This pursuit can be commercial (as in commerce—not a musical style!), intellectual, emotional, or a combination of these and other intentions. In my opinion, a truly great recording embodies the human spirit in a rich tangle of emotion and intellect. In a truly free market, it just so happens that this can sometimes lead to great commercial success when a recording resonates with the human spirit in a universal way.
Mixing for someone else
It’s one thing to record and mix your own tracks. After you have painstakingly constructed the set of tracks, one can only hope that you already have a strong idea of where the mix is headed. But mixing someone else’s project is another thing altogether, so for the purpose of this article, I’ll use the scenario of freelance mixing as an example.
When a client asks me to mix, we discuss their artistic intent and delivery medium. That done, I’ll pull up the project where it was last left and listen. In Pro Tools this generally means that I’ll be listening to at least some signal processing and automation that has been crafted by the tracking engineer. I like to begin where the last person to work on the tracks left off, even though most of the time I’ll soon dump much of their treatment and start over on the processing. That first audition gives me a good overview and insight into what the artist/producer has been experiencing as “the state of the track.”
Thoughts on gear
I own and operate a relatively large Pro Tools TDM system. When I mix for clients these days, most of the time they will have constructed their tracks on a Pro Tools LE system. Their system limits them to fewer simultaneous high-quality signal processing options compared to mine—in general they will have used fewer plug-ins.
I’ve carefully collected a sizable set of plug-ins, most of which sound more robust to me than those typically found in LE systems. Generally, I find that the tracks of projects that were created on Pro Tools LE systems don’t sound nearly as robust as I think they should. There are so many reasons why this can be the case, but let’s just say that expensive gear is usually expensive for a reason, just as great engineers and studios generally earn and deserve their reputation.
So there are a few main tasks awaiting me. One is to make each of the sounds more robust; another is to make the sounds fit together in a complementary way; yet another is to manipulate the musical dynamism in a manner that elicits more emotional/ dramatic impact from the track.
There are several ways to get to know what’s on each channel and track. In a live setting where line/ sound checks are often the only preparation an engineer is allowed, it is prudent to quickly go down the list of channels, optimizing the sound of each one before the band hits the stage. If you’ve ever been to a live sound check, after the interminable “testing 1, 2” you can be sure that the first word out of the FOH sound mixer will be “kick,” followed by “snare” etc.
As far as I’m concerned, this is the wrong model for beginning a mix in a studio environment. If a project comes to me with no previous mix intact, I’ll throw all of the faders up to the same level, and listen to the mix briefly to ascertain how things fit together in this raw state. To avoid master bus distortion, I place all track faders somewhere in the vicinity of –10 to –6 dB, with the master fader set at 0.
Once you begin to understand the relationships between the tracks, soloing is very useful for pinpointing signal processing. One good example of this is the relationship between kick and bass. Since they both produce sounds within a similar frequency range, while adjusting one, the engineer has to be considerate of the other so that the two sounds complement rather than fight with one another.
Depending on the musical style and the recordings of the bass and the kick, if I’ve recorded them I’ll generally eq one of the two to be more authoritative in the really low frequency range (20–50 Hz), and the other to be more authoritative in the 60–90 Hz range. I also tend to give each of the two a unique frequency boost in both the upper and lower mids, so that the two sounds work around each other.
All of this needs to be done by ear, according to the musical style and the recordings of both instruments. I often give the bass a little boost around 800–900 Hz, and then again at 2.0–4.5 kHz. I’ll usually give the kick a bit of a boost around 1.0–1.3 kHz, then again around 1.6–2.5 kHz.
I’ve been a musician most of my life, so the musician in me evaluates the arrangement when I first listen to a mix project. Is it dynamic enough to hold my interest? Sometimes I’ll go so far as to suggest structural changes like cutting out a bar or two here and there. Most of the time it is possible to do a little bit of judicious subtractive editing, so that when all the pieces are in, there’s more of a dynamic lift to the song. It is also possible to dramatically increase the intimacy of the performance by paring down the number of tracks.
The trick in doing subtractive editing is to make the most of the pieces you leave in. This can mean turning up the solo instrument when you break down to a solo instrument and a voice, so that you compensate for the suddenly missing roar of instruments you just muted. When you simplify a mix you’ll often have to rework reverbs and delays to increase the spatial dimensions.
Another issue with subtractive editing: you have to make the edited exits and entrances believable, unless you want to jar the listener. This often means finding another spot in the song (often the very end) to paste in for the last attack before an edited pause.
Pitch accuracy is another area of musical interpretation. Tuning problems can turn off the listeners. Pitch correction software (both automatic, e.g. Antares Auto-Tune, and manual, e.g. Serato Pitch ‘N Time) makes this task a little less tedious than it was back in the days of re-recording through an Eventide Harmonizer.
Panning is hugely important. I’ve heard one famous Hollywood engineer quoted as saying that there are only three proper pans—left, center and right. I disagree with this assertion but take part of the message to heart. It’s hard to go wrong being bold with a mix. Being bold generally means stretching mix concepts as far as you can without disrupting the overall presentation of the piece of music.
Mixing in stereo, there are a few different types of source tracks—mono, stereo and multiple. Generally, modern pre-packaged sound sources such as electronic keyboards, synths, samplers, etc., have stereo outputs. Generally these devices sound best (alone) when hard panned (sending the left output all the way to the left and the right output all the way to the right). If there are a number of such sources in your track, then hard panning all of them will most likely reduce the complexity, ergo the efficacy, of each individual sound.
Good panning helps to present a musical arrangement dynamically. In the case of multiple stereo keyboards, try to figure out which sounds suffer the least from being collapsed a little—meaning moving the pan of one or the other or both sides away from hard left or right. Sounds that have a lot of useful and interesting stereo information built into them can remain hard panned. Sometimes, turning off one side or the other is a good way to de-emphasize the robustness of a given stereo track. Sometimes, to create a mono track from a stereo track, it sounds better to pan both sides exactly the same, thereby summing the sounds.
Listen through the whole song with just the pre-packaged stereo sounds soloed. Adjust their pans so that the various musical parts support and balance each other, while—at the same time—you try to create interesting movement between the left and right. Depending on the rest of the arrangement, you may not be able to permanently leave the pans where you set them on this solo run, but at least you’ll have a strong idea of their interplay.
Panning bass and drums
Bass sounds are by nature less directional than higher-frequency sounds. It also just so happens that bass sounds require more energy to be reproduced. For both of these reasons, most engineers usually pan bass sounds to the center. In a typical rock band setting, this would mean that both kick drum and bass guitar would be panned center.
A fairly typical drumset panning would include kick and snare panned center. Beyond that, there are two camps, one advocating the drummer’s perspective and the other favoring the audience perspective. Of course, whether your drummer is right-handed or left-handed factors into this equation. Assuming a right-handed drummer recorded from audience perspective, my choice is generally:
• cymbal mics: hard panned or close to it,
• floor tom: at about 35% to 55% left,
• middle rack tom: at about 15% to 25% left,
• high rack tom: at about 5 to 15% right,
• hi-hat: at about 35 to 70% right.
The hi-hat treatment varies a lot based on the song, the player, the part played, etc. The way in which the drums were originally recorded has a great deal of effect on how they’ll ultimately need to be panned in the mix.
One good place to start with panning is to consider how a live performance would be set up in reality, then try to duplicate that in your mix.
Panning the band
Finishing out the rock band scenario, if it’s a four-piece band, I’ll generally pan each of the other players either entirely or mostly hard left and right. If overdubs exist, then the puzzle becomes a little more complicated. I try to create interesting movement within the stereo field, by balancing parts in such a way that layered parts are spread apart from each other, and consecutive parts (orchestration that carries a thematic thread) weaves from one side to the other and back, and so on.
And there’s more
There are many other aspects to creating a powerful, interesting, dimensional musical mix, too many to discuss here today. The processing for any fairly complicated mix will be likely to include eq, compression, limiting, time domain effects such as delays and reverbs, distortion, and modeling of increasing varieties.
I’ve mixed whole albums in a half day. I’ve spent literally in excess of 24 hours mixing one track. Usually, a good mix can be made within five to eight hours. When I used to mix in analog, even with automation, the paradigm was to start and finish a mix in a single session. Even with total recall, resurrecting a mix that includes a lot of outboard gear is a time consuming, inexact process. That has changed in a useful way.
Anyone who has mixed in analog knows how frustrating and disappointing it can be when—days after you finished a mix—you come to the realization that you didn’t get it exactly right. Going back to tweak the mix is expensive and not precise, so generally, unless you were on a big budget, you’d just live with it. With Pro Tools almost every setting can be automated and brought back repeatedly any time a session is opened.
With Pro Tools, I now actually suggest to my clients that we work in stages, crafting the mix toward its final destination in a number of sessions. This gives the client a chance to listen on different systems, with time to compare the mix to commercially available CDs, and allows time to live with the mix before signing off on it.
How do you know when you’re done? When—after a few weeks of careful listening and tweaking—there isn’t anything more to do.
Bruce Kaphan is a pedal-steel guitarist, recording artist, producer, and engineer in California. His latest album is Slider, about which you can learn more at www.brucekaphan.com.