7/28/2019 Beatboxing Techniques.

1/4

Characteristics of the beatboxing vocal styleTechnical report C4DM-TR-08-01

Dan Stowell and Mark D. Plumbley

Centre for Digital Music

Department of Electronic Engineering

Queen Mary, University of London

dan.stowell@elec.qmul.ac.uk

19th February 2008

1 Introduction

Beatboxing is a tradition of vocal percussion whichoriginates in 1980s hip-hop, and is closely connectedwith hip-hop culture. It involves the vocal imitationof drum machines as well as drums and other percus-sion, and typically also the simultaneous imitation ofbasslines, melodies, and vocals, to create an illusionof polyphonic music. It may be performed a capellaor with amplification. In this report we describe somecharacteristics of the beatboxing vocal performancestyle, as relevant for music signal processing and re-lated fields. In particular we focus on aspects of beat-boxing which are different from other vocal styles orfrom spoken language.

Beatboxing developed well outside academia, andseparate from the vocal styles commonly studied byuniversities and conservatories, and so there is (to ourknowledge) very little scholarly work on the topic,either its history or its current practice. Beatbox-ing is mentioned in popular histories of the hip-hopmovement, although rarely in detail. An under-graduate thesis looks at phonetic aspects of somebeatboxing sounds [Lederer, 2005]. Some technical

work is inspired by beatboxing to create (e.g.) avoice-controlled drum-machine [Hazan, 2005a,b, Ka-pur et al., 2004, Sinyor et al., 2005], although theseauthors dont make explicit whether their work hasbeen developed in contact with practising beatbox-ers.

In the following we describe characteristics of beat-boxing as contrasted against better-documented tra-ditions such as popular singing [Soto-Morettini, 2006]or classical singing [Mabry, 2002]. Because of the rel-ative scarcity of literature, many of the observationscome from the first authors experiences and obser-vations: both as a participant in beatboxing commu-nities in the UK and online, and during user studiesinvolving beatboxers as part of the first authors PhDstudy.

We describe certain sounds narratively as well as

in International Phonetic Alphabet (IPA) notation[International Phonetic Association, 1999] (see also[Fukui, 2003]), which will be demarcated by slashes// . The IPA representation may be approximate,since the notation is not designed to accommodateeasily the non-linguistic and extended techniquesounds we discuss.

2 Extended vocal technique

Perhaps the most fundamental distinction betweenthe sounds produced while beatboxing and those pro-duced during most other vocal traditions arises from

beatboxings primary aim to create convincing im-personations of drum tracks. (Contrast this againstvocal percussion traditions such as jazz scat singingor indian bol, in which percussive rhythms are imi-tated, but there is no aim to disguise the vocal originof the sounds.) This aim leads beatboxers to do two

1

mailto:dan.stowell@elec.qmul.ac.ukmailto:dan.stowell@elec.qmul.ac.uk

7/28/2019 Beatboxing Techniques.

2/4

things: (1) employ a wide palette of vocal techniquesto produce the desired timbres; and (2) suppress some

of the linguistic cues that would make clear to an au-dience that the source is a single human voice.

The extended vocal techniques used are manyand varied, and vary according to the performer.Many techniques are refinements of standard linguis-tic vowel and consonant sounds, while some involvesounds that are rarely if at all employed in naturallanguages. We do not aim to describe all commontechniques here, but we will discuss some relativelygeneral aspects of vocal technique which have a no-ticeable effect on the sound produced.

2.1 Non-syllabic patterns

The musical sounds which beatboxers imitate maynot sound much like conventional vocal utterances.Therefore the vowel-consonant alternation which istypical of most use of voice may not be entirely suit-able for producing a close auditory match. Instead,beatboxers learn to produce sounds to match thesound patterns they aim to replicate, attempting toovercome linguistic patternings. Since human listen-ers are known to use linguistic sound patterns as onecue to understanding a spoken voice [Shannon et al.,1995], it seems likely that avoiding such patterns mayhelp maintain the illusion of non-voice sound.

As mentioned above, vocal traditions such as scator bol do not aim to disguise the vocal origin of thesounds. Hence in those traditions, patterns are oftenbuilt up using syllable sounds which do not stray farfrom the performers languages.

2.2 Use of inhaled sounds

In most singing and spoken language, the vast major-ity of sounds are produced during exhalation. (Manylanguages do allow a minor linguistic role for inhaledphonation [Ladefoged and Maddieson, 1997]. Somevocal performance traditions feature inhaled sounds,e.g. Inuit throat-singing games [Nattiez, 2008].)

A notable characteristic of beatboxing is thewidespread use of inhaled sounds. We propose thatthis has two main motivations. Firstly it enables acontinuous flow of sounds, which both allows for con-tinuous drum patterns and also helps maintain theauditory illusion of the sounds being imitated (since

the sound and the pause associated with an ordinaryintake of breath are avoided). Secondly it allows for

the production of certain sounds which cannot beproduced equally well during exhaling. A commonly-used example is the inward clap snare /l/ 1.

Inhaled sounds are most commonly percussive. Al-though it is possible to phonate while breathing in,the production of pitched notes while inhaling doesnot seem to be used much at all by beatboxers.

Although some sounds may be specifically pro-duced using inward breath, there are many soundswhich beatboxers seem often to be able to producein either direction, such as the closed hi-hat sound/t^/ (outward) or /^/ (inward). This allows some de-

gree of independence between the breathing patternsand the rhythm patterns.

2.3 Vocal modes/qualities

Laver [1980] provides the classic phoneticians de-scription of the different voice qualities or phona-tory settings that an individual can produce, includ-ing falsetto, creaky voice, harsh voice, breathy voice,and ventricular voice. (The term modal voice isalso employed, to refer to the most common vocalquality against which these others are to be distin-guished.) These qualities may be consciously manip-ulated by a speaker, may be part of linguistic distinc-

tions between vowels, or may be indicative of vocalpathology. In study of the singing voice, too, dif-ferent vocal modes are distinguished [Soto-Morettini,2006], including head voice, chest voice, belt, twangyvoice, growl, breathy voice and creaky voice. Notethe (incomplete) overlap between the categories usedby the two communities.

Beatboxers make use of different vocal quali-ties to produce specific sounds. For example,growl/ventricular voice may be used to produce abass tone, and falsetto is used as a component ofsome sounds, e.g. vocal scratch, synth kick. Inthese cases the vocal qualities are employed for their

timbral effects, not (as may occur in language) toconvey meaning or emotional state.

Some beatboxing techniques involve the alterna-tion between voice qualities. If multiple streams

1http://www.humanbeatbox.com/inward snares

2

http://www.humanbeatbox.com/inward_snareshttp://www.humanbeatbox.com/inward_snares

7/28/2019 Beatboxing Techniques.

3/4

are being woven into a single beat pattern, thiscan involve rapid alternation between (e.g.) beats

performed using modal voice, vocals or sound ef-fects performed in falsetto, and basslines performedin growl/ventricular voice. The alternation betweenvoice qualities can emphasise the separation of thesestreams and perhaps contribute to the illusion ofpolyphony.

2.4 Trills / rolls / buzzes

Beatboxers tend to use a variety of trills to produceoscillatory sounds. (Here we use the term trill inits phonetic sense, as an oscillation produced by arepeated blocking and unblocking of the airstream;not in the musical sense of a rapid alternation be-tween pitches.) The IPA explicitly recognises threetrill types:

/r/ (alveolar trill or rolled R)

// (voiced bilabial trill)

// (uvular trill)

These have a role in beatboxing, as do others: trillsinvolving the palate, inward-breathed trills and click-trills.

The frequency of vocal trills can vary from subsonic

rates (e.g. 2030 Hz) to low but audible pitches (e.g.100 Hz) [Ladefoged and Maddieson, 1997, chapter7]. This leads to trills being employed in two dis-tinct ways: (1) for rapidly-repeated sounds such asdrum-rolls or dalek sound (the gargling effect ofuvular trill); and (2) for pitched sounds, particularlybass sounds. In the latter category, bilabial trill (lipbuzz) is most commonly used, but palatal trills andinward uvular trills (snore bass) are also used.

Notably, beatboxers improve the resonant tone ofpitched trills (particularly //) by matching the trillfrequency with the frequency of voicing. This re-quires practice (to be able to modify lip tension suit-

ably), but the matched resonance can produce a verystrong bass tone, qualitatively different from an or-dinary voiced bilabial trill.

A relatively common technique is the click roll,which produces the sound of a few lateral clicks inquick succession: /{{{