Recorder Home Page

Click to return to Recorder Home Page

Vibrato and Tremolo on the Recorder

by Nicholas S. Lander


"Vibrato" is a collective term covering oscillations of pitch, intensity or timbre that singly or in combination serve to enrich the musical sound (Neumann 1978: 521). In Early Music at least, Donington (1974: 229) suggests that we should reserve the word tremolo for fluctuation of intensity, often but not necessarily amounting to a reiteration of the note, and the word vibrato for a fluctuation in pitch not amounting to a change of note. When two notes a tone or semitone apart alternate we speak of a trill. When the notes are more than a tone apart we speak again of tremolo or tremolando. All of these were regarded as specific ornaments in the baroque period.

According to Donington, it is mainly tremolo, not vibrato, with which good singers bring their tone to life, and it is vibrato with which string-players enhances their tone.

Donington (1982: 35) relates that there are good acoustic as well as historical reasons for including vibrato in proper moderation.

"Recent researches put the time-span after which it is possible for our own faculties to perceive a new aural event as such, and not merely as an undifferentiated continuation, at about one-twentieth to one-eighteenth of a second. Any absolutely unvarying persistence of the same aural signal beyond this time-span rapidly fatigues that band of fibres in the basilar membrane of the ear which is involved in detecting it: there is then a subjective decline both in the volume and in the colorfulness of the sound perceived. It seems to go a little dead on us; and this is the acoustic consideration which makes vibrato a natural rather than an artificial recourse on melodic instruments. The vibrato just mitigates that deadening persistence."

Vibrato and tremolo are amongst the few expressive devices available to recorder players and are thus of fundamental importance. Up


Most modern methods for wind instruments recommend the production of vibrato by rhythmical muscular contraction which intercepts the regular flow of air. Rowland -Jones (1971) describes its production on the recorder "from the back of the throat, or, when a throbbing cantabile effect is desired, from the source of air-flow, that is to say, the diaphragm". He represents the action by the syllables "hu-hu-hu ..." aspirated deep in the throat without ceasing the actual flow of air. Similar accounts are given by Charlton (1981), Kottick (1975), Linde (1975), Vetter (1973), Waitzman (1978), Weisber (1975), and Wollitz (1978) who variously suggest the syllables "ah-ah-ah...", "uh-uh-uh", or "hu-hu-hu" for its production.

By far the most cogent instruction on the production of such an aspirated vibrato on the recorder is by Waitzman (1978). Waitzman notes that a pure throat vibrato is unacceptable on account of its extreme rapidity, jagged profile, and essentially uncontrollable nature. Contrariwise, a diaphragmatic vibrato alone is unsuitable due to its inherent slowness. Thus he argues that on the recorder and flute vibrato is normally generated by the action of the throat assisted by and coupled to the pressure and muscular action of the diaphragm. The throat serves as a valve, the player causing the throat musculature to assume the proper degrees and attitudes of tension, relaxation, and constriction so that the flow of air impelled by the diaphragm causes the effective diameter of the trachea alternatively to expand and contract many times in succession thus causing a regular variation in the strength of the air-stream. A useful analogy is that of an inflated balloon when the neck is stretched to make a squealing noise. The diaphragm serves mainly to cushion and smooth out the oscillation produced in the throat and is free to move in sympathy with them. Thus, acting in conjunction with one another the throat and diaphragm control the frequency, amplitude, regularity, and contour of vibrato. As Waitzman wryly remarks: "The techniques for this control are more easily learned than analyzed."

Now, this process when applied to the recorder certainly does produce an undulating intensity not unlike that of a pipe-organ's "tremulant" stop, but when played at any but the slowest speed it affects pitch hardly at all, approaching a tremolo in Donington's sense. Applied to reed-instruments the same process produces a hybrid tremolo/vibrato in which the proportion of each remains much the same regardless of the frequency of aspiration since such instruments respond to rapid aspiration (or articulation of any kind, for that matter) far less readily than does the recorder.

Some considerable light on the tremolo/vibrato controversy is thrown by Winckel (1967) from which it is evident that we must distinguish between acoustical phenomena and the psychology of their perception (psycho-acoustics). Winckel discusses experiments which demonstrate that the human ear is most sensitive to a vibrato of 4 pitch changes per second. Beyond 6 per second one notices only the unequivocal pitch of the initial tone with intensity fluctuations of the period of the vibrato. Measurements of good professional voices show that the vibrato rate of 7 per second is considered to produce the most aesthetically pleasing tone. This same periodic change of 7 per second holds for the violin. At this periodicity only a single pitch is heard, therefore no gliding of the pitch, but an oscillation of loudness, i.e. a tremolo in Donington's sense. All this means that for the violin and the voice a vibrato with a period change greater than 6 per second is in fact perceived as a tremolo and is actually more aesthetically pleasing than a slower vibrato! this auditory impression has led singing teachers (and Donington) to the erroneous conclusion that it is a fluctuation in volume that singers use to bring their tone to life.

As we have seen, on the recorder whilst response to aspiration remains much the same as its frequency is increased the extent of pitch change decreases dramatically. This acoustical phenomenon is complemented by the human perception of a rapid vibrato as a tremolo (Winckel 1967). Significantly, Weisber (1975) recommends a vibrato from 4-6 or 7 oscillations per second for wind instruments, suggesting that it should be faster for higher instruments than for lower. Toff (1996), writing of vibrato on the flute, recommends from 4-6 oscillations per second. Similarly, Waitzman (1978) suggests 4 oscillations per beat at a metronome setting of 76-80 (ie, ca 5 oscillations per second).

In general, contemporary recorder-players appear to utilise an aspirated tremolo/vibrato as described above rather than true vibrato to maintain a live and vibrant tone. Whilst a small, constant undulation of this kind certainly does much to add life and character to recorder tone its use should never become unconscious and indiscriminate or nervous tremolo is to be avoided at all costs. In particular, a tremolo/vibrato should be made conspicuous only to some definite purpose such as:

Waitzman (1978) cautions that tremolo/vibrato should be used to enhance a beautiful tone rather than mask a poor one. He also notes that sustained tones require a low amplitude of tremolo/vibrato, short tones a fast amplitude. Weisber (1975) notes that a faster tremolo/vibrato should be used for higher instruments than for lower ones. Up


A rather more obvious tremolo involving definite reiteration of the note can be produced on many wind instruments either by "flutter-tonguing" (vibrato linguale) or enunciating an uvular [R]. This has been used to special effect in romantic and modern music. From the recorder's repertoire Benjamin Britten's Noyes Fludde (where it imitates the cooing of doves) and Manuel Jacob's Eclogue come to mind. It is used extensively by avant-garde composers.

A very delicate, "tongued tremolo" can be be produced by articulating the syllables "yer-yer-yer..."(Rowland-Jones 1978; Vetter 1973). This, too, has been employed by protagonists of the avant-garde.

Thus for the recorder player there are four kinds of tremolo, each with its own distinctive character:

Note that the term "tremolo"as such occurs but once in the published recorder repertoire, namely in the Canzona con il tremolo "La Grimaneta" for "flautin & fagotto" from Giovanni Battista Riccio's Il terzo libro delle divine lodi musicali of 1620 (Griscom & Lasocki 1994: 243). VibratoAs early as 1536 Agricola characterised the wind-player's vibrato as "trembling breath". Praetorius (1619) described a vibrato produced by the action of the diaphragm, and Mersenne (1636) refered to "certain tremolos which intoxicate the soul."

Nowadays vibrato is considered as being on both sides of a note or on the sharp side. This can be produced on reed instruments by manipulating the muscles of the lips and jaw as well as the diaphragm, but on the recorder only by allowing breath pressure to fluctuate in a regular manner controlled by the diaphragm. The production of vibrato in this way on the recorder is the subject of an article by Davis (1974) who recommends its application to music of the baroque period. Played slowly this gives a wide vibrato but, as we have seen, as its frequency is increased the extent of the pitch change automatically decreases until it becomes an aspirated tremolo. Since it involves a very noticable oscillation in volume as well as pitch, a slow diaphragmatic vibrato on both sides of the note tends in practice to become a distressing warble. Thus its use on the recorder is probably best avoided, despite historical evidence in support of its use on the baroque flute (Sterne 1977).

A vibrato performed by finger movements alone on a wind instrument is always on the flat side. Although this was possible on baroque flutes and oboes it is not, for the most part, on their modern descendants. The recorder is one of the few wind-instruments on which this style is still feasible, and it should be used wherever stylistically appropriate, and not just in the performance of French baroque music and its derivatives in other countries. Indeed, it is called for in treatises by Ganassi (Venice 1535), Blankenburgh (Amsterdam 1654), Hudgebut (London 1679), Salter (London 1683), Loulie (Paris ca 1700), Hotteterre (Paris 1707), Corrette (ca 1735), Mahaut (Paris/Lyon 1750), Quantz (Berlin 1752), and Tromlitz (Leipzig 1791). See Brüggen (1996), Dickey (1978), Haynes (1997a & b); Lasocki (in Hotteterre 1707/1968: 66, footnote), Mather (1973: 65-66), Moens-Haenen (1988), Neumann (1978: 511-522; 1993), Veilhan (1979: 36-37) and Zimmermann (1994).

The table of fingerings for these flattements, as they are called, which I have adapted from Hotteterre (1707) to suit modern (neo-baroque) recorders may be of interest; it will certainly convey some idea of the complexity of this style of playing. Note Hotteterre's indication that a flattement is done with the technique of the trill, but that only the edge of the hole is struck.

These notes concern both alto and soprano recorders with modern ('English') fingering systems. For a given fingering the note produced on an alto will be indicated first followed in parenthesis by that produced on a soprano (or tenor, although it will sound an octave lower).

Fingering will be indicated as follows:

0 123 4567 thumb and all fingers on
- -2- ---- second finger of left hand on alone
/ thumb-hole pinched
0 1-3 45/- finger 6 half-holed
0 1/3 4567 finger 2 partially closed
4* trill with this finger
(5) optional fingering
/ 1-3 4-6/ X bell hole closed against one's knee

Thus / 123 45/*- indicates the fingering for a flattement on a'' on an alto and e''' on a soprano, made by striking the edge of half-hole 6 with the finger.

alto soprano flattement
f' c'' Not feasible. See below.
f#'/gb' c#''/db'' ditto
g' d'' 0 123 456/*
g#'/ab' d#''/eb'' 0 123 456/*
a' e'' 0 123 45/*-
a#'/bb' f'' 0 123 4/*67
b' f#''/gb'' 0 123 /*56-
c'' g'' 0 123 /*---
0 123 -/*--
c#''/db'' g#''/ab'' 0 12/* 45--
d'' a'' 0 12- 4*---
d#''/eb'' a#''/bb'' 0 1-3 4/*--
e'' b'' 0 1-/* ----
f'' c''' 0 -2- 4*---
f#''/gb'' c#'''/db''' - 12- 4*---
g'' d''' - -2/* ----
g#''/ab'' d#'''/eb''' - 123 456/*
a'' e''' / 123 45/*-
a#''/bb'' f'' / 123 /*5--
b'' f#'''/gb''' / 123 /*5--
c''' g''' / 123 /*---
/ 123 -/*--
c#'''/db''' g#'''/ab''' / 123 4/*--
/ 12/* 4---
d''' a''' / 12/* ----
d#'''/eb''' a#'''/bb''' / 12/* ----
e''' b''' / 12- 45/*-
f''' c'''' / 1-- 45/*-
g''' d'''' / 1-3 4-6/*

For f' (c'') and f#'/gb' (c#''/db'') Hotteterre recommends shaking the instrument. It is hard to see what this could achieve in itself other than what might be termed "visual vibrato", a cue to the audience that a vibrato is intended. But perhaps here is an authentic use for the so-called "lip vibrato" described by Brüggen (1997) in which the fluctuation of breath pressure is generated by arm movement, "by pressing the mouthpiece (ie windway entrance) slightly to the side or toward the upper lip, temporarily decreasing the amount of air going through the instrument." One could, of course, use aspirated tremolo/vibrato instead.

Feldon (1996, 1997), presents evidence to the effect that the placement of flattement in French baroque music was deployed to articulate key rhetorical elements in a musical structure in a flexible manner. Thus it was applied to heighten the expression by signalling a change, high point, instability, or ambiguity or used to lend emphasis to contrasting or shifting movements of rhythmic stress.

Contemporary composers for recorder have sometimes made use of a vibrato effected by shading the window of the recorder with the palm of the lower hand. This is termed "labium vibrato" by Bowman (1997) and by Brüggen (1997). By this means the fluctuations of pitch can be over a very wide range, though one is limited to the notes available to the upper hand alone (unless one enlists the aid of an assistant, perhaps). If the air flow remains constant then the overall pitch is lowered. Thus the effect of labium vibrato can be very dramatic since it tends to be combined with a powerful tone. Examples of its use include Hans Martin Linde's Music for a Bird and Marcus Zahnhaussen's Viva Vivaldi.

Thus for the recorder player there are four kinds of vibrato, each with its own particular character:

Timbre Fluctuations

Vibrato effected by periodic fluctuation of timbre seems to have attracted no particular name. It has been given little more than passing mention in the literature on the recorder, e.g. by Brüggen (1997) who nowhere states quite how such an effect might be achieved. As the quote from Donington (1982) at the head of this article hints, and as Brüggen himself concedes, it is inevitable that slight changes of timbre are coupled with vibrato or tremolo however these are achieved. Indeed Toff (1996), writing about vibrato on the flute, equates timbral vibrato explicitly to fluctuations in pitch + fluctuations in intensity. But since such "timbral vibrato" can hardly be independently controlled, it need not concern us further, though it is important to be aware of its co-existence with other forms of vibrato and tremolo. Greater possibilites for timbral vibrato might be offered by the manipulation of tonal colour possible with innovative instruments such as Michael Barker's "MIDIfied recorder" or Philippe Bolton's electro-acoustic recorder.


We have seen that on the recorder tremolo grades almost imperceptibly into vibrato and that aurally the following are but convenient points on a continuum: However, technically it is possible to combine certain of these effects so that they can reinforce or work against each other. Thus one might combine finger vibrato and aspirated vibrato. However, an attempt to combine flutter-tonguing with an uvular tremolo might land you in hospital! Further, the effects discussed here can be combined with others to give new and exotic sounds, e.g. combining flutter-tonguing with trilling or tremolando, used to striking effect by Richard Harvey of Gryphon.

Notwithstanding the hints given above on several particular uses of aspirated tremolo/vibrato, it is impossible to lay down general rules for the distribution of either tremolo or vibrato. Although Hotteterre (1708) bids us: "Observe that it is necessary to make flattements on almost all long notes, and to do them ... slower or quicker according to the tempo and character of the piece" one should keep in mind his earlier dictum (Hotteterre 1707) that: "It is taste and practice which teaches their appropriate use, rather than theory." This latter view is echoed by Rowland-Jones' comment that:

"The detailed application of vibrato and volume variation to phrasing cannot be explained in words; it stems from musicianship, that quality partly inborn, partly bred of enthusiasm, self-criticism, study, and a sense of tradition."
Inherent in all such opinions is the message that much can be learnt by experimenting oneself and by listening critically to authoritative playing on any instrument whose tone is not produced mechanically. Up


Click to return to Recorder Home Page

© Copyright 1996-2005, N.S. Lander
Page last updated Sat, 09 Jul 2005 14:0
Webmaster: Nicholas S. Lander