Sign up for our newsletter and get tutorials and tips delivered to your inbox.
Photo by Oskar Wimmerman
The mixing of a live record is no easy task, regardless of the genre or the instrumentation. You have to reproduce both a reality and the reality’s quintessence—something that goes beyond the reality.
It’s not unlike trying to catch lightning in a bottle. To take the analogy further, it’s more like crafting a mechanism that recreates the appearance of lightning, and then stuffing that in a bottle; others might see the lightning, but you know the hard work that went into creating its verisimilitude.
What follows are some quick tips to get you started—a breakdown of the best possible processes and practices for pushing a live mix to its pinnacle. These aren’t necessarily specific tricks for esoteric situations. Instead, what you’ll find here is my general approach, whether I’m mixing rock, jazz, classical, or podcasts; as long as it’s taped in front of a live audience, these six tips have worked for me. The seventh tip is a bonus.
A guitar player who supported Ornette Coleman once told me the following: Every morning he'd wake up, roll out of bed, walk over to his upright piano, lay his hands over some random keys, and, as the unholy noise planked out, he’d bless that chord. This was his morning ear-training ritual, and it always stuck with me—never more so than when mixing live tracks. See, with rare exceptions (overdubs provided in studio), what you have is what you have. You must make peace with the audio.
How does this translate into an actual tip? The very first thing you do is listen to the tracks and come to terms with their limitations. You bless the mess, embracing these tracks exactly as they are. It clears the mental cobwebs, allowing you to move forward.
Next, going through each track in solo—and constantly comparing to them the overall jumble—find the channel (or channel pair, if it's a stereo overhead/ambience capture) that best encapsulates the performance. This is now your North Star. Your job is to make sure every other track follows its light, and enhances its every aspect.
I don’t approach mixing a studio recording this way. Indeed, there’s a lot about mixing a live record that’s quite different and counterintuitive.
And that brings us to tip number two.
To a certain extent, a live band mixes itself on stage. Various factors can surely undermine this principle: The band could be terrible; the recording could have problems; the guitar player could’ve had an off-night. If you run into signal problems, you have plenty of tools to correct hum, noise, clicks, pops, and other issues (iZotope RX 6 comes to mind), though not a lot beyond global editing (always risky) to save a lousy performance.
One assumes the band wouldn't be terrible, or else they wouldn't want you to mix their show for public consumption. So, assuming you’ve got a decent band on your hands, you're going to get the best results with our two simplest tools: level and panning.
EQ and compression are obviously mainstays of studio mixing. But these processes, though useful in spots—don’t get me wrong—can often do more harm than good on a live record, and here’s why:
Because of the bleed inherent in live recordings, everything you're putting up contributes to the sonic picture in an interdependent, relational way. Thus, if you EQ the drum overheads, you’ll probably have a palpable effect on whatever instrument is nearby, which could be the piano, the bass, or any number of elements.
Of course, songs are recorded with bleed in studios all the time. But in the studio, instruments are usually mic’d with mixing in mind. It tends to be fifty/fifty with a show: Unless a distributed recording is intended from the outset, the mic setup usually facilitates live performance—and that’s a different discipline from studio recording. Less thought might be given to how the spill distributes, because the engineer is counting on the old principle of “if it sounds good, it is good.” The live engineer’s “good” exists in the room. Yours, unfortunately, must exist in every playback system under the sun.
Onto compression, which also might also make things worse. Consider the vocal: that SM58 is cardioid, but it'll still pick up all the sound coming directly into it. So when you're compressing the channel, you're not only bringing the vocalist down—you’re also bringing up the crash right behind the singer’s shoulder.
Patti Smith performing in Finland, using an SM58 | Photo via Wikipedia
More often than not, both upwards and downwards expansion are more suitable in these contexts. The former process allows the instrument to pop out of the mix when it needs to, while the other brings ambience down to a more suitable, blendable level (more on that later).
In a way, it's almost more beneficial to think of mixing a concert like mastering, in that targeting an instrument with EQ will have invariable effects on all the instruments taking up that frequency space. The analogy falls apart when you get granular (a stereo track versus a multitrack, for instance), but the idea of preserving the cohesion is very much at the forefront, as it is with mastering.
Because compression can emphasize unwanted blemishes in the live sonic picture, you're going to want to use other means to make sure every note is heard (and every word is understood). The best method I know of is automation—painstaking, seemingly endless automation. You bring up the sax for the solo, then bring it back down when it's silent, never killing its ambience outright, as that’s part of the picture.
Consider that the spatial relationships of everyone on stage could necessitate panning automation as well: remember, we're using every mic we have to communicate the quintessence of the night. And that sax? Its ambience might have a lot of ride cymbal in it, so off toward the ride cymbal it goes.
Yet when the sax solo plays, that far-right pan position is invariably distracting; a solo often craves to be put down the middle. Since there’s now sound blasting into the mic at close proximity, you’ll find the offending ride cymbal eclipsed, so here we have a good case for automating your panning. Of course, when the sax player stops, you’ll have to move the track right back to where it was. I know I’ve had to automate panning between phrases.
Don’t be afraid of automating down to a tee in live mixing: this is where the nitty gets gritty!
Sometimes, to cut down on automation, a certain amount of clip gain and gating can go along way in speeding up the process. Rather than draw in every move, clip-gaining the ambient parts down to a suitable level might be the quickest solution.
I'd pick clip-gaining over gating because when that instrument comes in, so will a certain amount of the overall spill. Yes, we just mentioned that sax solo as an example for automating your panning, but every instance is different; sometimes, the spill around a new element can shock the ear unpleasantly, especially if it only exists in a smaller part of the frequency spectrum. Clip-gaining the signal down, rather than gating or muting it, preserves some of that ambience. Thus, when it’s time to raise the level, it’s not too much of a shock, if you do it gradually and musically.
A case could be made for using a gate with a higher close-threshold, or fine-tuning the hysteresis controls to control the gate’s behavior. Still, I prefer clip-gaining regions to a suitable level and using automation to bring levels of music up and down. Call me nit-picky, but I'd rather retain manual control over something as variable as a live performance than farm the process out to a piece of processing.
Plenty of engineers do get good mileage out of gates, particularly in ways described above. On percussive instruments, whose flubbiness might need to be curtailed in a live record (a kick drum, for instance), I find gates to be quite helpful.
Believe it or not, good applause is not as easy to come by as you'd think—it's inevitably ruined by something the band does (such as tuning or engaging in terrible banter), or by a particularly loud/annoying member of the audience. Furthermore, real-world applause and on-record clapping are two entirely different things: as Bob Katz astutely mentioned in his book, Mastering Audio: The Art and the Science (a must-read), real-world applause tends to last much longer in the venue than you'd ever hear on the record.
His discourse on the subject informs my approach: He recommends prepping a few well-sourced and varied applause breaks to intersperse between tracks as need be. In my career, I've done this often, but I’ve also had to sweeten applause that wasn't loud enough because it wasn't captured well, or because the audience simply wasn't there.
You will need to manufacture techniques for manufacturing applause, whether it be editing a bunch of sparse clapping from different songs together to create a larger audience (don’t forget to pan them in the interest of fullness!) or bringing in a sound library to do the job. In that case, I often take an impulse response from the gig (a drummer clicking the sticks together works in a pinch—it's like the clap of a slate), and feed the fake applause through some ambience.
Whatever bedrock principle you employ in lying to the listener, you need to analyze how applause works, its inherent flow, the way it tapers, and what sounds signal the fake when you’re looping a sound bed.
Yes it's laborious, but the editing skills you'll pick up in learning how to manipulate ambience and applause will help you in other gigs (post production comes to mind).
One of the trickiest aspects in mixing a live record is improving ambience without calling attention to unnatural reverbs, delays, and modulation effects. While there are plenty of pointers I could throw at you on the subject, the most effective one is this:
Give yourself plenty of time. If you’re going to make a decision about sweetening the reverb of a snare, or adding delay to a guitar, make your decision and then walk away. The longer break you take, the better your ears will be at immediately determining whether or not you made a good decision upon resuming the mix.
When I’m mixing a live record, often the last thing I’ll do in a night’s work is work on ambience, because I know that in the morning, with a cup of coffee and a good night’s sleep (okay, that should be in the other order), I’ll have a better perspective, not only on whether the ambience I’ve sweetened is too loud or soft, but whether it’s even appropriate in the first place.
I’m not saying don’t take risks. In creating the feel of the concert, you can go to town in this area, pulling out crazy EQ curves and kooky modulation moves to enhance a reverb for the right reason (the right reason being to fit with the vibe of the venue). But always step away after doing so and come back later with fresh ears; you wouldn’t pile on more decisions about level and balance after making a faulty ambience-based judgment call. It will wreck your perspective.
If you really want to up your skill set in mixing live recordings, then I recommend asking your friends in the live-sound business to lend you the multitrack board-recordings of their worst gigs.
Assuming you can get them to agree (you might need to bribe them), why would you ever put yourself through this? The answer is simple: You need to develop a set of tricks for dealing with horrible situations, and these tricks must fit within your workflow.
I could give you tips on how to use Vocalign to time-align overdubs into live recordings, but if you don’t have the software, it won’t do much good. I could tell you how to turn every spare mic you can grab into ambience, but my solution may call for calling up more tracks than you’re comfortable with, or might cause a heavier CPU strain than you’re willing to impose.
The point is, you’ll need to figure out your own tricks for working effectively and efficiently through problematic material. There’s no shortcut around this; my tips will only get you started. There’s no substitute for experience, so in the absence of experience, you must manufacture your own.
The biggest thing to remember about mixing a live record is that you are not there to put your own stamp on the proceedings. The best live records aren’t showcases for the flashiest mixers to demonstrate their abilities. They are opportunities to showcase a band, both as they are, and in their best possible light. A studio record has this objective too, but the arrangement can be shaped and aided by the mixer. Live recordings are different: You are beholden to the sound of the musicians in a truly accountable way. If you keep that in mind, and honor it at every turn, you’ll be on your way to getting great results.