Search Gear

Control Freaks! Try Micro-Editing Vocal Tracks

June 30, 2014

EVER WONDER why most famous singers sound so much better on their records compared to performing live? It’s not just pitch correction responsible for their transformation. Deep composite editing and dozens of edits at the syllabic and sample levels have been implemented in the recording. Standout vocal excerpts have been cherry-picked and combined and blemishes removed, creating a track that consistently presents the singer’s very best capabilities from moment to moment. You can do the same. Use these micro-editing techniques to make your vocal track the crème de la crème.

Mix and Match Record several takes of the lead vocal track—more than a dozen, if the singer can handle it without flaming out. Listen to every take in turn as you play back the first vocal phrase. Copy the best parts—even if they’re just fleeting syllables— to a new, blank track reserved for your final composite vocal. In choosing the gems, you should generally be listening for five things: intelligible lyrics, great intonation, strong projection, excellent tone, and a strong vibe that conveys the emotional essence of the song. But don’t overlook momentary dross. It’s not uncommon for me to snap up a breathy syllable, a momentary crack in the singer’s voice, or other ostensible imperfections that ooze emotion for a nanosecond in an otherwise unfeeling phrase.

Fig. 1. The waveform to either side of the edit point is trending downward, toward negative phase, avoiding producing a click.
Fig. 2. This splice, made at the zero crossover point for both the leading and following waveform, nevertheless produces a click because of a phase-trend reversal and high-amplitude spikes with steep slopes on either side of the edit seam.
Fig. 3. This splice produces no click despite the reversal of phase trend (from positive-going to negative-going) at the zero crossover because the amplitudes and slopes are mild to either side of the edit seam.
Once you have the first vocal phrase cobbled together, move on to the next and repeat the process. But don’t worry about poorly assembled snippets with slapdash transitions that cause clicks and pops in the composite track. We’ll address all those pedestrian issues later. Your sole focus should be on creative cooking until you’ve worked your way through the entire song and have the best vocal bits assembled in one track.

Slip and Slide After you’ve built the composite vocal track, you’re ready to start micro-editing it in your DAW. Pay attention to how crisp the phrasing is; it should propel a rhythmic, uptempo song along or lay back a bit for a ballad. Within each phrase, cut up and nudge syllables milliseconds earlier or later along the timeline where necessary, to make them lock to the beat. Don’t go overboard, though, or the vocal will sound robotic.

Copy and Paste In your composite track, listen for consonants that aren’t clearly voiced, making the lyric unintelligible. Replace any swallowed consonants with the same ones articulated more clearly elsewhere in the song by copying the champ and pasting it in place of the lemon. If the consonant in the composite track should occur at the start or end of a word or phrase, be sure to paste it so that it voices squarely on a beat or strong subdivision of the song. While nudging one syllable or consonant may not make a noticeable difference in the overall performance, slide a couple dozen bits in a four-minute song and the result will be a more authoritative vocal that’s firmly in the pocket.

Bump It Up or Down Your composite track is bound to comprise bits and pieces sung at different volumes. If any snippet dips or leaps out even slightly in an unnatural way, bump the region’s volume up or down to even it out with surrounding ones. Doing this will preclude the need to use heavy vocal compression during mixdown, preserving depth and nuance. If your DAW doesn’t allow non-destructive volume adjustments for regions, draw fader-automation curves to smooth out the ride. The goal here isn’t to beat the vocal dynamics into submission but to mitigate any dips that make the lyrics unintelligible or peaks that sound jerky.

Smooth the Seams Don’t rely on crossfades to hide poor-sounding edits; subdued clicks and pops have a nasty habit of resurfacing when dynamics processing is subsequently applied during mixing and mastering. Instead, optimize each edit seam by adjusting region boundaries so that each transition occurs at a zero crossover for both the leading and following waveform. If possible, place the edit seam so that the waveform on each side of the splice is trending toward the same polarity (that is, consistently rising or falling; see Figure 1); an instantaneous reversal in phase will often cause a pop even when it occurs at a zero crossover (see Figure 2). If the edit still causes a click after finessing its placement, implement a crossfade on the seam.

Transitions between audio regions often sound the smoothest where waveforms have mild slopes on either side of the seam. (You can often get away with instantaneous polarity reversals if the amplitude is low and the slope very mild to either side of the edit seam; see Figure 3.) The exception is with hard consonants: Placing the seam right before the amplitude spike for a t or k usually sounds best. And an edit that doesn’t occur at the zero-crossover point, if artfully placed, can sometimes enhance a following hard consonant that wasn’t inherently loud enough.

Let It Breathe Take care not to place edit seams in the middle of breath noises, as that will make them sound artificial. Don’t completely eliminate breath sounds either, unless they’re distracting in some way. Extinguishing all breathing will make the vocal sound unnatural—who doesn’t breathe for four minutes?—and, for uptempo songs, bereft of urgency.

Enjoy the Fruits of Your Labor If skillfully and artistically rendered, you can make a hundred or more edits to a vocal track without anyone realizing it’s been touched. The total effect of all your nips and tucks will be a vocal performance that far surpasses the quality and impact of any one of the original takes. It typically takes hours of work to build such a virtuosic track, but the huge payoff is totally worth the sweat.

Michael Cooper is a contributing editor for Mix magazine.

Show Comments

These are my comments.

Reader Poll

Do you make more money songwriting, recording, or performing?

See results without voting »