Why should instrumentalists have all the fun? For years guitarists, bassists, and keyboardists have had arsenals of tone-twisting tools at their disposal, enabling them to stomp and slide their way to sonic paradise. Singers, on the other hand, have had few options for changing their vocal sound onstage. Sure, a little reverb or delay could be added from the front-of-house mixer, but that was about it.

Thanks to advances in digital processing, however, there's now a growing list of signal processors designed especially for vocalists. These devices allow singers to enhance, alter, pitch-correct, harmonize, and, in some cases, change the gender of their voices.

To give you an idea of the possibilities, we chose six of the hottest live-performance voice processors on the market and tested them under performance conditions using male and female singers. We picked models that offer at least some form of pitch shifting or pitch correction. Unlike guitar, bass, and rack effects — which tend to resemble each other in features — these devices are as different as the singers who use them, in terms of both feature set and physical design. Choosing which processor is right for you depends largely on your performance style and the needs of your particular gig.

The six units we looked at are the Akai HV-1 DuoBuddy, the Antares Vocal Producer AVP-1, the Boss VT-1 Voice Transformer, the DigiTech Vocal 300, the DigiTech Vocalist Workstation EX, and the TC-Helicon VoicePrismPlus. But before delving into the product reviews, let's look at some of the important concepts of voice processing.

HARMONIES IN A BOX

Harmony generation is one of the coolest features on a voice processor. (Most of the devices we tested have some type of harmony capability.) Depending on a unit's features, you can generate two-part or even multiple-part, musically correct vocal harmonies.

The most basic scheme used for producing harmonies is called fixed-interval, or chromatic, pitch shifting (similar to what you'd find on an instrument processor with pitch-shifting capabilities), whereby the input note is shifted up or down by a preset number of semitones. In live applications, that type of effect is useful for creating octaves and can also be effective on parallel intervals such as fifths.

But fixed intervals can have limitations regarding creating key-correct harmonies. Take, for example, a major-third harmony in the key of C. With a pitch shifter that's set to a fixed interval of a third (four half-steps), the harmony would be correct for the notes C (E), F (A), and G (B) but would be a half-step sharp for the notes D (F#), E (G#), A (C#), and B (D#).

However, a processor with “intelligent harmonization” features could be set to generate the correct notes, because these devices can be set to conform each harmony note to a specified scale (scalar) or chord (chordal).

For those musical situations in which pre-programmed chords and scales are too static for the material at hand, intelligent harmony processors have another trick called manual input. By inputting MIDI note data in real time from a keyboard (or other) controller into a voice processor's MIDI In port, it's possible to specify exact harmony voicings and scales.

You can assign manual input to harmonies or to the lead voice (in the latter case, you can even talk into the microphone and play a melody with the keyboard). That can allow you to hit notes that would otherwise be impossible for your voice to reach. Some notes might not sound entirely natural, but the technique is effective nonetheless.

You can also program a sequencer to input the correct note and chord information into your voice processor. Of course, this option requires that your band play in sync with the sequence.

VOICE-CHANGE OPERATION

The pitch of a person's voice is just one of its defining characteristics. The resonant frequencies produced by your vocal tract are called formants, and they help determine the qualities, or the character, normally associated with gender and size.

Many voice processors offer formant shifting (aka formant shaping), a feature that lets you control these resonant qualities. Formant shifting is therefore a key parameter when you are trying to create a realistic sound that falls outside of your voice's natural range. Also known as formant-corrected pitch shifting, this process allows the formant characteristics of your voice to be preserved as the pitch is shifted, thereby helping it to maintain its original character and avoid the “chipmunk” (or “Darth Vader”) effect.

One of the most dramatic applications of formant shifting is “gender bending,” in which a male voice can be made to sound like a female and vice versa. You can also use formant shifting to change the quality of a voice within its own range. It's an effective tool for extending the character of your voice beyond its own limitations.

Formant shifting is especially valuable for building realistic harmonies because it reduces the “all me” effect you'd get from harmonizing all in the same voice. By assigning each voice in a harmony its own formant profile, you can more closely approximate the sound of multiple singers.

TC-Helicon — the joint venture between TC Electronic and the Canadian company IVL Technologies (IVL also licenses or has licensed its voice-processing technology to a number of companies, including DigiTech, Mackie, Korg, Yamaha, and Steinberg) — offers yet another level of realism with its voice-modeling technology, available on its VoicePrismPlus and Voice One processors. Based on actual human voices, it allows you to add individual attributes, such as vibrato, timbre, and even stylistic expression, to a voice.

RANDOM ACTS OF SINGING

When real humans sing together, their natural body rhythms and phrasing cause minute timing differences. That's one of the qualities that gives a choir its sonic breadth. Some voice processors can randomize elements of the harmony voices in relation to the input signal. Basic parameters include delay and pitch, but advanced models also let you randomize stylistic elements such as scoop (bending up to the note), vibrato, and more.

GETTIN' IN TUNE

Automatic pitch correction is a relatively new technology and requires a fairly robust digital processor in order to be effective. A pitch corrector analyzes an incoming signal and conforms it to the nearest note in a specified scale. The closer your input note is to the correct scale value, the more seamless the pitch correction will be. Pitch correctors can be set to both chromatic and diatonic scales. In the latter case, all of the notes will fall within a specific key. Although diatonic pitch correction can be more accurate than the chromatic variety (which corrects to the nearest semitone), it robs you of the ability to add accidentals (notes that fall outside of the specified key signature).

When the input note is halfway between two scale values, a pitch corrector can have some troubles. The processor may send the “corrected” note to the wrong pitch. For example, let's say you have the processor set to a chromatic scale. If you start a note 50 cents flat (with the intention of scooping it up to the next correct note), the pitch corrector may read the note as sharp instead of flat and send it a half-step below your intended destination.

Several parameters allow you to account for the pitch fluctuations that are part of a natural performance. The pitch corrector's speed parameter can slow its response to allow you to scoop up and down to notes. A slower response time preserves more of your natural inflections, but it can also reduce the effectiveness of the correction. Slow times are especially useful when you want to ensure that long sustained notes stay in pitch. On the opposite end of the spectrum, fast response times can create a stair-step effect that can give your voice a vocoder-like quality.

You can also adjust the sensitivity of a pitch corrector, allowing it to track and correct notes without altering too much of your performance. Lower sensitivity settings let you have some flexibility in expressing your note, while higher sensitivity offers stricter pitch correction.

Unless you're experiencing severe pitch problems, you'll most likely need the pitch correction for only a few notes. If your technique includes scoops and vibrato, you may even find a slow pitch correction response intrusive. Some devices allow you to specify which notes in any specific scale will receive correction and which will be left alone. (Useful if you have a high or low note that you're having trouble singing in tune.)

SIX PROCESSORS

Now that we've covered the basic concepts, let's look at the processors themselves. Each of these devices has a unique feature set and is optimized for a different application.

Akai HV-1 DuoBuddy

The DuoBuddy is a stompbox-style harmony generator. Because many of the unit's functions can be controlled with its three large footswitches, it's ideal for guitarists and bass players who sing, or for any vocalist who stands onstage.

Connections include XLR mic and ¼-inch line inputs, an XLR mic output, a ¼-inch mix out, and a separate harmony output that splits the lead and harmony signals when active.

The DuoBuddy offers two-voice (your voice plus a single harmony) intelligent harmonization with chromatic, scalar, and chordal modes. It also offers pitch correction, which can either be used on the lead voice (by setting the harmony voice to unison) or on the harmony voice. You also get formant shaping, and, to add realism to the harmony voice, random delay and pitch change. Front-panel knobs let you control the blend between lead and harmony signals.

There's no MIDI input, but you can choose from 12 preset scales and 9 user scales (only 4 scales can be accessed at any one time). The footswitches can be assigned to change presets or to change the root key of a particular scale — an especially useful feature in a live situation. You can also assign a variation to each scale as part of a preset and trigger that with the footswitch without changing presets. That feature will, for example, let you alter the third from major to minor where appropriate without having to change the overall scale. Nice. As you'd expect, you can bypass the DuoBuddy by foot, although I noticed a popping sound whenever I did so.

The DuoBuddy's small, two-character LED display is somewhat limited, and the owner's manual is poor. (Some of the programming routines just didn't jive with their descriptions in the manual.) Still, the footswitch operation, harmony features, price, and compact size will appeal to many musicians. (Akai will soon be shipping the DeccaBuddy [$499], a rackmount harmony processor that lets you generate four separate harmony parts, which can result in a total of ten voices.)

Pros: Compact. Foot operated. Harmony and formant shifting with pitch correction. Footswitch can recall presets or change keys. Independent lead and harmony outputs. Footswitch lets you select an alternate harmony on the fly. Mic and line inputs.

Cons: Bypass switch pops. Limited display and poor manual makes for cumbersome editing. Only one voice of harmony.

Antares Vocal Producer AVP-1

Like DigiTech, Antares has been a pioneer in vocal processing. Many people credit (or blame) Antares's AutoTune pitch corrector — which first appeared as a Pro Tools TDM plug-in and later was ported to other software formats and to hardware devices — for revolutionizing the recording industry by allowing marginal singers to sound good. The single-rackspace AVP-1 differs from the rest of the processors we reviewed in two major ways: it has no mic input (you patch it in to your vocal channel's inserts), and it offers no type of pitch shifting.

The AutoTune pitch-correction feature, however, is worth the price of admission. You can set it to operate chromatically, or you can set it to any one of 24 diatonic scales, in major and minor keys. From there, the AVP-1 gives you plenty of control over how the pitch correction takes effect. You can control timing and sensitivity, turn individual notes on and off within a scale, and defeat the AutoTune (as well as other parameters) in real time, using MIDI or a footswitch. Overall, I found the AutoTune feature to be very effective.

But the AVP-1 brings much more to the table. It offers a compressor/gate, a de-esser, and EQ. Further, it has a feature that none of the other processors in this roundup have: microphone modeling. This effect can imbue the output of a common stage mic with sonic characteristics from different mics, such as a large-diaphragm condensers.

The mic-modeling section also includes an effective emulation of a tube-mic preamp. I especially like the way it can add clarity and punch to a dynamic mic to evoke the sound of a studio condenser — without introducing the problems, such as feedback and rumble, that would make one of these mics a nightmare onstage.

The AVP-1 consistently enhanced the sound of my voice. Oddly enough, despite the mic-modeling features, the AVP-1 is not designed to work as a mic preamp. It has a single unbalanced line input and both a mono output and one labeled Double Track. (The Double Track function generates a separate, detuned version of the lead vocal that can be routed to the Double Track output or mixed in the mono output.)

Operation is easy, and the manual presents complex subject matter clearly and with good humor. The presets are also very effective and are geared for a range of musical applications. There are even presets for a few instruments — such as guitar, bass, and drums — that take advantage of the mic modeling, EQ, and dynamics control.

Pros: Outstanding and versatile pitch correction. Mic modeling lets you get condenser sound with out facing condenser problems (like feedback) onstage. Excellent sound. Dynamic effects, EQ, and de-essing. MIDI input can control effects parameters in real time. Good manual.

Cons: No mic inputs. Unbalanced line inputs. No dedicated power switch.

DigiTech Vocal 300

The Vocal 300 floor-mounted processor is the vocalist's equivalent to a guitarist's or bassist's multi-effects stompbox. It accepts mic and line inputs and offers both line and mic outputs, meaning that you can place it between your mic and the house mixer's mic inputs. There are also two ⅛-inch jacks: a headphone out and an input for an external source like a CD player.

The chassis sports four foot controls: program up, program down, bypass switches, and an expression pedal for controlling volume, pitch bend, and other effects parameters.

Features include a compressor, a mic preamp (which also has some basic voice-altering features), a 3-band EQ, a noise gate, delay, reverb, and a number of modulation effects such as chorus, flange, phase shifting, tremolo, vibrato, doubling, envelope, detune, and pitch shift (chromatic). The unit also offers some special effects like Strobe (which creates chopperlike effects), Pixelator (which creates a grainy effect), and Whammy (in which the expression pedal controls the pitch in real time).

The Vocal 300 has 40 factory presets (as well as 40 user presets), which cover basic vocal sounds and a number of unusual sounds suitable for a horror-movie soundtrack. Editing is relatively straightforward, thanks to a matrix system that shows you the current edit parameters. You can select and change values with the five data knobs, which also operate as EQ and level controls in standard play mode.

While the Vocal 300 is less complex than some of the other processors here — it has no pitch correction or intelligent harmonization — it's a useful device for performers who need to change vocal sounds on the fly. If two singers alternate on one mic onstage, you can use the Vocal 300 to optimize settings for each. Using your vocal mic for an instrument such as harmonica or horn is also cool; trumpeter Wallace Roney used a Vocal 300 on Herbie Hancock's recent Future 2 Future tour. You can add effects to the instrument and then go back to a pristine vocal sound for your singing. The sturdy chassis should stand up to plenty of abuse. On the negative side, the manual is extremely sketchy. It explains the controls but gives you very little in the way of applications.

Pros: Foot operated. Sturdy construction. Wide range of effects. Built-in expression pedal. Operates on mic and line-level signals.

Cons: Poor owner's manual.

DigiTech Vocalist Workstation EX

In 1989, DigiTech became the first company to offer affordable vocal harmony processing and intelligent pitch shifting, with the Vocalist VHM5. The Vocalist Workstation EX reflects the mature development of a successful product line. It has an XLR mic input, ¼-inch balanced line input, stereo ¼-inch balanced line outputs, a headphone out, MIDI jacks, and an input that lets you use an external footswitch to control bypass, program advance, and other parameters.

This compact tabletop unit features advanced pitch processing, including pitch shifting (with speed control), chromatic pitch shift, scalar pitch shift (you can use preset scales or create your own custom scales and chords), chordal pitch shift, and manual input. Formant control lets you set the gender for the harmony voices, and you can set the scoop parameter to create more realistic bends between notes.

The programming interface is very user-friendly. You get sliders to set lead and harmony level as well as reverb. More important, there's a one-octave minikeyboard on the front panel that lets you input key and chord information directly. (You can also do this using MIDI, which offers a greater range of control.) Soft keys under the display allow you to select the harmony's voicings. That is one of the Workstation's coolest features: you can change the intervals as you sing to experiment with the harmonies in real time. Song mode lets you string together a number of harmony profiles that you can switch with a momentary footswitch (optional).

The reverb sounds good, but it's the harmonies that steal the show. Tracking is excellent, and the shifted voices generally sound sweet and natural. The interactive features make this an effective live tool. And happily, in contrast to the Vocalist 300, the manual is outstanding.

Voice Processor Specs

	Akai HV-1 DuoBuddy	Antares Vocal Producer AVP-1	DigiTech Vocal 300	DigiTech Vocalist Workstation EX	Roland (Boss) VT-1 Voice Transformer	TC-Helicon VoicePrismPlus
Type	floor-style	rackmount (1U)	floor-style	tabletop	tabletop	rackmount (2U)
Dimensions	8.8" (W) × 2.3" (H) × 6.4"(D)	19"(W) × 1.75" (H) × 5" (D)	13" (W) × 2.25" (H) × 8.5" (D)	10.5" (W) × 6.6" (H) × 1.7" (D)	7" (W) × 2.06 (H) × 6.3 (D)	19" (W) × 3.5" (H) × 8.2" (D)
Weight	1.9 lb.	4.5 lb.	3.75 lb.	1.5 lb.	1.1 lb.	7.8 lb.
A/D/A Conversion	20-bit/48 kHz	18-bit/44.1 kHz A/D 24-bit/44.1 kHz D/A	24-bit/44.1 kHz	18-bit/48 kHz	A/D/A spec is not available	24-bit/44.1 kHz; 48 kHz
Mic Input	XLR	none	XLR	XLR	¼"	XLR
Line Input	(1) ¼" unbalanced	(1) ¼" unbalanced	⅛" stereo CD input; ¼" unbalanced	(1) balanced ¼"	RCA	(1) ¼" TRS balanced; (1) ¼" TRS aux
Digital I/O	—	—	—	—	—	AES/EBU; S/PDIF
Mic Output	XLR	—	XLR	XLR	¼"	—
Line Outputs	(2) ¼" mix; ¼" harmony	(2) ¼" main; double tracker	(1) ¼" TRS stereo	(2) ¼" TRS balanced	(2) RCA	(2) ¼" TRS balanced
MIDI	—	In, Out, Thru	—	In, Out, Thru	—	In, Out, Thru
Available pitch shifting or harmonization	chromatic, scalar, chordal	—	chromatic	chromatic, chordal, scalar, manual	chromatic, (continuously variable)	chromatic, chordal, scalar, manual
Generated Harmony Voices	1	—	1	4	1	5 (4 harmony plus lead doubler)
Pitch Correction	yes	yes (on lead voice)	—	yes	—	yes (on the harmony voice)
Voice Character	formant shaping random delay, random delay with pitch change	—	formant shaping, detune, vibrato	formant shaping, detune, scoop, vibrato	formant shaping, detune, vibrato	voice modeling, formant shaping, random delay scoop, vibrato
Additional Effects	—	gate, compressor, EQ, mic modeler, de-esser, double tracker	compressor, mic preamp, gate, EQ, modulation, delay, reverb expression pedal	reverb	reverb	gate, compressor, EQ, modulation, delay, reverb
Presets	21 (12 factory/9 user: 4 available at one time)	35 rewritable/restorable	80 (40 factory/40 user)	100 (50 factory/50 user)	40 (20 factory/20 user; 8 available at one time)	128 rewritable/restorable
Controls	footswitches; front-panel knobs; expression pedal	front-panel buttons	footswitches; expression pedal; knobs	top-panel sliders; buttons and knob	top-panel sliders; buttons	front-panel buttons and knobs
MSRP	$399.00	$599.00	$249.95	$599.95	$395.00	$1,598.00

Pros: Good value. Extensive harmonization features. Pitch correction. Formant shifting and other humanizing parameters. MIDI input of harmony notes. Front-panel minikeyboard for controlling harmony. Parameter control from front-panel sliders. Friendly user interface. Excellent sound. Mic input. Song mode lets you chain presets to suit your song. Excellent manual.

Cons: No effects other than reverb.

Roland (Boss) VT-1 Voice Transformer

A product that will appeal heavily to the DJ market, the VT-1 is a tabletop unit that features sliders that control pitch (continuously variable from one octave above the input to one octave below), a formant slider (also continuously variable), a mix balance control for setting the relationship between the original and pitch-shifted signal, and a slider to control the built-in digital reverb. There's also a Robot button that restricts all notes to a monotone, creating a vocoder-like effect.

The VT-1 has a ¼-inch mic input (I wish it was XLR), a ¼-inch mic output, and stereo RCA outputs. With a footswitch, you can toggle the voice character effect on and off while keeping the reverb active. The VT-1 can hold eight presets at a time (four of them user definable).

Despite — or maybe because of — its relative simplicity, the VT-1 is a gas to play with. I loved creating scratchlike sounds, gender-bending, and being able to change voice parameters with the sliders as I sang. That you can slide the pitch, mix, and formant parameters independently increased the fun. However, the VT-1 offers no chordal or scalar harmonization, nor does it allow for MIDI input, so it's not a tool you'd use for generating pop-type harmonies.

Although it has fewer features than the other processors in this story, the VT-1 does what it does — creating dance-music-style vocal effects — very well and with good sound quality.

Pros: Simple to operate. Continuously variable pitch and format shifting plus reverb. Sliders offer real-time control of all parameters. Small.

Cons: Limited feature set and applications. RCA outputs only.

TC-Helicon VoicePrismPlus

The VoicePrismPlus is the most feature-rich — and the most expensive — device of the six reviewed here. The two-rackspace unit is as complete a voice processor as you're likely to find, offering intelligent harmonization, formant shifting, pitch-corrected harmonies, advanced voice modeling that can alter the character of a lead voice in great detail. It also gives you MIDI input control and a full signal path including a microphone input, EQ, compression, and spatial and modulation effects. (TC-Helicon's VoicePrism [$999] has most of the same features, except it lacks advanced voice modeling and has a less flexible digital I/O.)

The VoicePrismPlus offers a wide range of inputs: a front-panel XLR mic input (with switchable 48V phantom power); a bunch of rear-panel connections, including a balanced ¼-inch line input; a separate balanced ¼-inch aux input that bypasses the harmony section but routes signal through the VoicePrism's effects section; stereo ¼-inch TRS outputs; and S/PDIF and AES/EBU digital I/O.

As a harmony generator/pitch shifter, the VoicePrism covers the gamut, including chromatic, scalar, chordal, and manual-input pitch shifting. Up to four voices are available. Manual shifting can be based on MIDI input, and you can assign each voice in the harmony to its own MIDI channel, which allows you to play in a specific harmony for each (otherwise, the harmony voices may overlap).

Although the VoicePrismPlus's harmony features are impressive, they tell only part of the story. The unit goes beyond basic formant shifting, offering extensive modeling of both lead and harmony voices. You can not only define the gender and size of the singer, but TC-Helicon's Human Voice Modeling lets you impart qualities such as breathiness, growl, rasp, scoop, pitch accuracy, vibrato, and more. There are presets that are based on the vocal characteristics of well-known singers including Stevie Wonder and Louis Armstrong. I got a kick out of the setting that turned my own low tenor into the voice of a soprano opera singer. I compensated for the ensuing identity crisis by dialing up a preset reminiscent of Barry White.

As with the DigiTech Vocalist Workstation EX, the VoicePrismPlus's MIDI implementation is extensive. You can assign continuous controllers to effects parameters and use the keyboard to input notes for the harmonies. You can also organize presets into song chains that can be accessed with a momentary footswitch.

The editing interface is generous and expansive, and there's enough information provided that you rarely need to consult the manual. I like the fact that you can keep the sound-shaping elements of a preset, such as compression, EQ, and reverb settings, while changing the harmonies.

The VoicePrismPlus's effects are nothing short of outstanding: lush reverbs, sparkling delays, musical EQ and compression, and more. I found that the presets not only sounded good but were resistant to feedback. I don't have space to go into all of this unit's features, but suffice it to say that its price tag is justified. (Note that TC-Helicon also offers the VoiceOne, which uses similar technology but is geared more toward studio than live use — hence its exclusion from this story. The VoiceOne offers a dazzling feature set, albeit at a fairly high price.)

Pros: Mammoth feature set includes interval, harmony, chordal, and manual pitch shifting in a number of modes. Unique voice modeling goes beyond mere formant shifting. Extensive real-time control. Sleek editing interface. Excellent effects. Complete selection of analog and digital I/O.

Cons: Expensive.

IT'S A WRAP

The purpose of this story was to provide you with a wide-angle view of what options are available by looking at a broad range of vocal processors. However, because of the significant difference in the prices and the features of the six processors, we didn't feel that it would be fair or relevant to declare a “winner.” They all have their strengths, and which one is right for you depends on your needs.

If you don't want to spend too much, need only two-part harmonies, and desire hands-free operation onstage, the Akai DuoBuddy is a good bet. If controlling vocoder-style sounds, gender bending, and other wild effects with real-time sliders appeals to you, consider the Boss VT-1.

If your aim is to purchase a multi-effects processor for your voice, and you don't need intelligent harmony features, take a close look at DigiTech's Vocalist 300. If you'd like multiple voices of harmony generation, MIDI control, and a lot of processing power for a relatively low price, the DigiTech Vocalist Workstation EX might be your best bet.

However, if flexible and accurate pitch correction is your highest priority, and you don't need harmony generation or pitch shifting, the Antares AVP-1 is at the top of the heap. (And of course, its mic modeling and dynamics control features aren't too shabby either.) Finally, if your budget can handle it, the TC-Helicon VoicePrismPlus is the Rolls-Royce of the six units we looked at, offering top-notch harmony generation, a wide range of mult