The process of mixing can be broadly divided into two primary challenges: whipping individual tracks into shape and achieving a proper overall balance when blending those tracks together.
Regarding that last bit, no processor will automatically put the bottom end in proper perspective to the rest of the spectrum or sit the lead vocal perfectly in the mix. Those tweaks are balancing acts that require cultivated ears, experience and an accurate mix room and monitoring chain.
Polishing individual tracks is a lot more straightforward, though some dogs take a lot more grooming than others. Thankfully, modern plug-ins can perform miraculous makeovers on even the ugliest of tracks. Here are a few ready solutions to five particularly intractable problems you’ll inevitably face at the mix position, whether working on a major-label release, racing to meet a broadcast deadline, or polishing a song-demo project on a shoestring budget.
The Missing Snare Mic
Problem: When bringing up the fader for the snare drum track, it’s immediately obvious the drum was captured only with a top mic. You can barely hear the snare’s rattle, and the drum sounds like a bongo. How can you add the missing sizzle?
Solution: One way is to route the output of a white-noise generator to a new track, slap a gate on the track’s insert and bus the snare track via a send to the gate’s external sidechain input. Every time a snare drum hit exceeds the gate’s threshold, a burst of white noise will be triggered—a good proxy for rattling snares. Problem is, quieter rolls and grace notes won’t trigger the gate and will still sound like they were played on a bongo. It’s a moot point, though, as you don’t have a white-noise generator on hand.
Another solution is to “re-snare” the snare drum track. Send the track to an active monitor placed in your tracking room. Aim the monitor skyward, directly underneath a snare drum. Mike up the drum’s snares, and route the preamplified mic signal to a new, record-ready track. Hit Record and play the pre-recorded drum track from start to finish. Every time the snare drum is played, the speaker will excite the live drum’s snares, adding back the missing bottom-mic signal. What if you don’t have a snare drum on hand? You could always layer or replace the snare drum track with samples. But it would take some time to get grace notes and rolls to trigger properly and eliminate false triggering from mic bleed.
Fig. 1: Lexicon Resonant Chords and SPL Transient Designer plug-ins together process a top-miked snare drum track to simulate the sound of an additional bottom mic.
The fastest and easiest solution is to synthesize the missing snare rattles by instantiating on a track insert the Lexicon Resonant Chords plug-in, followed by the SPL Transient Designer plug-in in the next insert (see Fig. 1). Once you’ve programmed and stored the custom presets I’ll describe for the two plug-ins—do this before you need them—it will take only seconds to recall their composite effect and convincingly simulate the sound of a bottom-miked snare.
The Resonant Chords algorithm makes up to eight delay voices resonate on specific musical notes when excited by a percussive input signal—in this case, your snare track. The upshot is to make Resonant Chords generate, using very short delay times, a discordant cluster of eight detuned notes spanning as a group around two semitones. The attack portion of the densely packed tonal cluster will imitate the musically unrelated harmonics snares produce when excited. But because the generated tones will ring so long as to create a bizarre-sounding reverb, you’ll use Transient Designer after Resonant Chords to severely cut the tones’ sustain and match the tighter envelope snares typically produce.
Lexicon provides an outstanding factory preset to give you a head start in creating the effect: Kettle Room 5 (stored in the Odd Spaces category). Edit the preset thus: Increase the resonance value for L Voice 4 to 19%, and bump up the resonance values for the other seven voices to 13%. In the Resonance control section, hike the Resonance Master value to at least 19% and lower the Blend control to -40.0 to give the resonating tones more prominence. In the plug-in’s Master control section, lower the Mix control to 59% to include a healthy amount of dry signal at the output.
Open Transient Designer in the immediately following insert, and plunge the plug-in’s sustain control to its lowest possible setting to dramatically shorten Resonant Chord’s effect envelope. If the effect still sounds too loose or ambient, lower the Delay Time Master in Resonant Chords’ Master control section. Store the settings for both plug-ins, and you’ll be able to recall them whenever you need to add a virtual bottom snare mic—in seconds!
The Icy Amp Sim
Problem: The electric guitar track was recorded through a guitar-amp simulator, and the pick strikes sound like glass breaking. The DI’d source track was erased, so you can’t re-amp the track—you’re stuck with the sim. You try EQ’ing the track using a lowpass filter to melt the icy attack, but that just dulls the overall sound without fixing the problem. How can you warm up this brittle-sounding track?
Solution: When amp sims bite, the problem is usually due to the guitar’s attack sounding glassy. The rest of the envelope may sound fine. The solution in this case is to de-ess the track. A de-esser works more transparently than static EQ because it acts dynamically, or only when high-frequency transients sound too prominent.
Fig. 2: Waves DeEsser warms up an icy-sounding electric guitar track.
The Waves DeEsser plug-in and iZotope Alloy 2’s De-Esser module both do an excellent job warming up nasty-sounding amp-sim tracks. In Waves DeEsser, select a highpass filter for the plug-in’s internal sidechain and set the HPF’s corner frequency to about 5500 Hz (see Fig. 2). Toggle the Audio button to the Split setting so that DeEsser will process in splitband mode. With this setup, DeEsser will compress only the high frequencies above the inherent cutoff for a typical analog guitar cabinet; the amp sim’s spectra below 5500 Hz will not be processed. The final touch is to lower DeEsser’s threshold slider to the point where sharp string attacks trigger gain reduction but the guitar’s airy sustain does not. Around 4 to 6 dB of gain reduction on peaks should give your guitar track a smooth tone with a rounded attack and fully preserved presence.
The Paper-Thin Floor
Problem: The floor tom’s track has no bottom end to speak of. The tracking engineer, in his infinite wisdom, tried to eliminate kick-drum bleed in the tom’s mic by rolling off low frequencies below 120 Hz with an HPF. Your challenge as mix engineer is to rescue the thin-sounding floor tom track and make it sound huge again. But you can’t boost a bottom end that no longer exists. What to do?
Fig. 3: Missing bottom end on a floor tom track is restored using Waves LoAir.
Solution: Obviously, you could replace or layer the tom track with samples that have more heft. But a quicker solution is to synthesize the lost bottom end using the Waves LoAir plug-in (see Fig. 3).
LoAir filters audio with a lowpass filter before sending the signal to a subharmonics generator, adding spectra an octave lower. The plug-in splits the dry signal into two copies and sends one copy to its LPF, the corner frequency for which you set using the plug-in’s Range control. The level of the unprocessed dry signal is adjusted using the plug-in’s Direct control, whereas the filtered (post-LPF) signal’s level is set using the Lo control. A mult of the LPF’s output is routed, in a parallel path, to a subharmonics generator, the output level for which you adjust using the LoAir control. The takeaway is that the Range control delineates the upper-frequency limit for audio routed to both the Lo and LoAir controls. Although there’s some spectral overlap in the Lo and LoAir signal paths, the LoAir control outputs audio with far greater bass extension.
To rescue the emaciated floor tom track, instantiate LoAir on one of its inserts. Set the plug-in’s Range control to 120 Hz to demarcate the upper limit of the bottom end you wish to restore (spectra the tracking engineer threw away). Adjust the Lo and LoAir controls to set the respective levels of bass and sub-bass reinforcement, sculpting the bottom end to taste.
For technical reasons I don’t have space to explain, you may find you need to raise the Lo control higher than the LoAir knob to prevent a hole in the mid-bass response. Keep in mind that raising either control above its 0dB mark will progressively add saturation and produce a more aggressive sound; to bring the bottom end into good balance with the unprocessed mids and highs—without unduly increasing saturation—you may need to lower the Direct control a tad. Once you’ve mastered using LoAir, it should only take a minute or two to restore missing bottom end on any drum track.
The Erratic Demo Singer
Problem: A music publisher has handed you a basket of song demos to mix. The publisher’s budget is razor-thin, so you have to work quickly. Unfortunately, on one song the lead singer’s timbre and dynamics are all over the map. The tone is way too bass-y on verses and strident on choruses, and the levels swing wildly throughout the entire track. How can you rush through this mix when the money track needs so many fixes?
Fig. 4: Pro Audio DSP’s Dynamic Spectrum Mapper V2 plug-in simultaneously reins in a singer’s erratic tone and levels.
Solution: Luckily, there’s a quick fix. Use the Pro Audio DSP Dynamic Spectrum Mapper V2 plug-in (DSM) to rein in the singer’s timbre and dynamics simultaneously (see Fig. 4). DSM automatically splits its FFT-based processing across at least 16 frequency bands, so you won’t have to spend precious time adjusting crossovers. Processing in all the bands is governed by one set of controls for ratio, knee, attack time, decay (release) time and makeup gain, speeding your workflow further. And because DSM’s processing is more transparent than a true compressor’s would yield, you don’t have to worry about using a sledgehammer on your out-of-control track.
Instantiate DSM on an insert for the lead vocal track. Listen to the song and, when the singer’s timbre sounds the most balanced, click and release the plug-in’s Capture button to create a so-called threshold spectrum. Represented by a static yellow curved line in the GUI’s X-Y display, the threshold spectrum is the instantaneous frequency response you captured—and the timbre you’d like the vocal track to more or less have throughout the song. As you play back the singer’s track, you’ll also see a blue curved line moving around in the X-Y display. The blue line represents the vocalist’s changing frequency response from moment to moment. Any time a portion of the blue line rises above the yellow one, DSM will compress the vocal in the corresponding frequency bands to move the track’s timbre closer to that represented by the yellow line.
The threshold spectrum might not be the absolutely ideal vocal timbre you want for the song, but you can edit the yellow line’s shape to make it so, using the parametric threshold controls at the bottom of DSM’s GUI. Lower the line below 1 kHz to put a firmer lid on the boomy and muddy verses. Also lower it around 5 kHz to more heavily compress highs during the louder and brighter choruses. Raising the line at 12 kHz will preserve detail and air throughout the track. Set DSM’s ratio to 100:1. Don’t worry—DSM stays transparent no matter how hard you hit it. Jack up the makeup gain to compensate for the ultra-heavy compression. Activate DSM’s mastering-grade limiter to bridle any peaks that might get past the compressor. The vocal’s tone and dynamics will now sound incredibly balanced and controlled throughout the entire song—and it will have only taken you five minutes and one plug-in to accomplish.
The Heavy Breather
Problem: You’re engineering the final mix for a video project. The mix has to be delivered by mid-afternoon, in time for broadcast on PBS that evening. On one of the dialog tracks, the speaker’s breathing is annoyingly loud in the gaps between each and every one of their sentences. Some breath noises are much louder than others—you can tell that a noise gate on the track would be prone to chatter. You don’t have time to manually erase each breath noise, but they must be subdued, if not completely eliminated, and the clock is ticking.
Fig. 5: iZotope Breath Control tames breath noises on a dialog track.
Solution: Either of two plug-ins will tame the wheezing in a heartbeat. Waves DeBreath and iZotope Breath Control can automatically attenuate the breath noises to a consistent level you specify (or mute silence). Unlike a gate, Breath Control doesn’t use threshold-based processing. Instead, the plug-in analyzes the harmonic structure of inputted audio to determine which portions are breath noises and need to be processed—no matter their level.
Slap Breath Control on an insert for the dialog track and activate the plug-in’s Target-mode button for the most natural-sounding results (see Fig. 5). Raise the Sensitivity slider until the orange-colored meter to its left begins to show signal. A waveform display at the top of the GUI gives you visual confirmation that attenuation is occurring in the track’s timeline where it should: An orange gain-reduction trace at the top of the display will dip lower in the gaps between the waveforms representing dialog. Drag the Target slider to the level you want all detected breaths to be reduced to.
Activate the Breaths Only button to hear in isolation the audio being attenuated; if you hear any hint of dialog in Breaths Only mode, lower the Sensitivity slider until you hear only breathing. Once your settings are optimized, deactivate Breaths Only mode. Then bounce the dialog through Breath Control to a new track in real time while listening, to confirm the processing is transparent from track start to end. Relax, because Breath Control is remarkably discriminating (and way more reliable than a noise gate in this application). Set it and forget it!
Mix contributing editor Michael Cooper is a recording, mix, mastering and post-production engineer and the owner of Michael Cooper Recording in Sisters, Oregon.