Nowadays the traditional tools—eq, compression, echo, reverb—are well-represented by a bewildering array of top-notch plugins. The most recent additions to this set of vocal-processing options allow us unprecedented control over aspects that for years had been out of the engineer’s domain: namely the actual elements of the performance itself. And along with this new power have come questions about how (and how often) to use it
Engineers have always had the ability to manipulate the pitch of a vocal recording to some degree. Even in the days of magnetic tape, slowing down and speeding up the recorder was a trick that could be employed to help a struggling vocalist reach that elusive high note. And in the digital age, pitch shifters (both hardware and software) have always been able to shift the tuning of a performance, although not without noticeable artifacts. But the art of manipulation reached new levels of sophistication with the advent of plug-ins that could effectively and smoothly adjust the intonation of vocal recordings, note-by-note, on the fly.
The first of these to achieve wide recognition outside engineers’ circles was, of course, the plug-in Auto-Tune, from Antares. For anyone who hasn’t been in a recording studio recently, Auto-Tune automatically corrects the pitch (intonation) of a monophonic audio signal passing through it, based on a user-defined scale (most often this is a chromatic scale, but it can be restricted to a diatonic or modal one as well). The plug-in also offers a manual mode for more problematic tracks, where the amount of pitch correction can be graphically entered for a specific section of the track.
If applied gently, the pitch correction can be virtually seamless—slightly sour notes are miraculously put right, and no one need be the wiser. On the other hand, if applied too aggressively, notes’ pitches sound “quantized” to discrete values, for a somewhat mechanical effect (this is usually referred to as the Cher effect, from a well-known song by that artist, “Believe”, that deliberately featured the effect prominently—whether or not Auto-Tune was actually responsible has been debated, but the effect is easily reproduced in the plug-in).
For some years Auto-Tune ruled the roost when it came to pitch correction—eventually it was joined by other companies’ efforts, such as Yamaha’s Pitch Fix, and Logic’s (built-in) Pitch Corrector plugins. These only offer the automatic mode, lacking any king of graphic way to manually zone in on problem areas, but as auto mode was by far the most common usage anyway, they pretty much provide the same basic service.
While many musicians were at first wary of having their performances toyed with this way, eventually many came to appreciate the upside of these plug-ins—being able to tighten up slight “pitchiness” digitally, or salvage an otherwise good take marred only by one or two off-key notes. With DAWs, it was always possible to fix up a performance with editing, overdubs, and punch-ins, but auto-correction was a new level of efficiency and unobtrusiveness—things could be cleaned up without a significant impact on the flow of a session.
The good, the bad, and the ugly
At their best, pitch-correction plug-ins help the singer to relax, and go for a performance with great feel and energy, without worrying about the stray “squeaker”, knowing the occasional flub can be brought under control efficiently, without breaking their musical momentum (and we all know how important that is in a session). Of course, as with any new technology, there can be a dark side as well....
Many people feel that pitch-correction tools, even good ones subtly applied, have fostered a certain amount of laziness on the part of some performers, and unrealistic expectations on the part of others. While I do embrace the technology, I must admit that this can be true. In sessions, I’ve sometimes seen singers gloss over the (real) need for a retake, or come in unprepared or not fully warmed-up, assuming that “that thing can fix up any roughness, right?”
And, unfortunately, from what I hear, many engineers are also too quick to turn to the automatic fix, and too heavy-handed in their application! I more than occasionally hear some warbling that sounds like an artifact of pitch quantization on commercially-released tracks that clearly has nothing to do with any kind of deliberate creative use of the effect, and everything to do with a pro-forma application of the processing without carefully monitoring the track for artifacts—or worse, no one in the session could even pick up on the degradation to the performance that had resulted from their quest for “perfection”.
However, despite these issues, pitch-correction technology is here to stay, and overall that is a good thing. In fact, the latest developments have upped the ante even further, bringing another degree of sophistication to the basic corrective process, and providing more opportunity for creative use as well.
Where is the harmony, sweet harmony...
Not too long ago, a company called Celemony brought out a piece of software called Melodyne, which many engineers and musicians credit with bringing the art of pitch processing to another level. Unlike the products mentioned above, Melodyne is not strictly a plug-in, but a standalone application, requiring a bit more processing power to perform its magic. (Editor's Note: At the time of writing in 2006, Melodyne plugin did not yet exist. There is a plugin version now that inserts on an individual track.) It can be called up from a plug-in within a DAW, or via Propellerhead’s ReWire protocol, for access that, while not as simple as regular plug-ins, is nonetheless relatively convenient in a typical session situation.
Instead of merely correcting off-pitch notes automatically, or even allowing for a graphic automation line to be drawn, Melodyne applies the “piano-roll” display familiar from MIDI sequence editors to the manual pitch-adjustment of monophonic audio recordings. After loading in an audio track, the pitches are displayed as waves (note-by-note) in a piano-roll screen, and, just as with MIDI data, each note can be dragged up or down to any pitch (automatic correction is also available). For better expressiveness, even slides and glides between notes are visually represented, and can be edited as well.
With all this control, not only can errant notes be shifted to the intended (nearest) pitch, but the melody can be changed and experimented with at will, and custom harmonies can be created with an unusually accessible visual environment. Melodyne’s excellent formant-adjusting transposition algorithms, while not unique for this type of processing, are widely acknowledged as among the best around.
Of course, digital generation of harmonies has been with us for quite a while (hardware from DigiTech and TC Electronic), but thanks to Melodyne’s visual interface and resulting ease of use, many people feel that this constitutes the next level in pitch control. Other manufacturers would seem to agree, as pitch processors with a similar look and feel have just been announced or released from Waves (Tune) and MOTU (functionality built into Digital Performer 4.6). Like Melodyne, they are not simple plug-ins, but are accessed easily enough from within the DAW environment, and offer the same combination of pitch correction and melody/harmony editing and creation.
Is it Live, or is it Melodyne?
Naturally, as with the more basic pitch correction tools, questions have arisen about the use of such computer technology in place of simply working out and recording harmonies and additional vocal tracks with real live singers. And again, there’s both an upside and a downside.
In the plus column, processors like these allow very quick generation of doubled vocal and harmony tracks, even after the singer has gone, which facilitates creative experimentation—I’ve done this myself in sessions where there either wasn’t time to figure out and record harmonies the traditional way, or when a producer came up with some great ideas for background vocals, but only after dismissing the session vocalist.
But on the downside, generating a lot of backing tracks from one or two leads doesn’t offer the variety of vocal tones you’d get from having different singers come in and record the different harmonies, and even with subtle doubling algorithms to slightly vary the rhythm of the artificial harmony/doubling tracks, digital processors excel more at creating very tight harmony parts, and are less effective when a looser, more individual feel would be better for each of the various elements in a vocal arrangement.
All things in moderation, my son
Still in all, a judicious combination of high technology and good old-fashioned musicianship, as always, will provide the best and most musical results. And today’s crop of pitch tools, when used with a deft and subtle hand, can take the art and process of vocal recording closer to a state of nirvana than ever before—so enjoy!
Joe Albano (firstname.lastname@example.org) is a recording engineer and producer in New York. You can check out his website at www.rooftopproductions.com.