The recorder is a marvellous instrument. The fingering system allows a keyless instrument to be played chromatically over more than two octaves. The thumb-hole acts as a speaker-hole making it possible to play the second and third registers without blowing harder (as must be done on most whistle class instruments). Criticism of the recorder in the past has centred on its lack of dynamic variation.

If you want to make a flute class instrument such as the recorder play a louder note you have to increase the volume of air you are supplying. By this I mean the actual quantity of air, not the pressure or speed. If you are supplying 1 litre from your lungs every 10 seconds then supplying 2 litres in the same time will make the instrument louder!

Since the recorder has a fixed duct through which you blow the only way to get more air into the instrument is to blow harder and push more air through by increase the speed of the airstream. Unfortunately it is also a feature of flute class instruments that the pitch of a note will rise if the air speed increases.

Through careful design of the windway and voicing it is possible to make a recorder where the pitch is quite stable over a range of breath pressure allowing for a greater range of musical dynamics. Some of the dynamic variation is brought about by a change in tone, which suggests a change in actual loudness, but some genuine changes in volume can be achieved. Although the overall change in volume is still quite small it is musically very useful, as it is through windway and voicing design that the baroque recorder achieves such a beautiful and variable sound.

However, since the recorder is played with other instruments we are faced with a dilemma. If you were able to make a recorder where the pitch of each note is independent of the breath pressure you are then unable to adjust the pitch of these notes. In practical music making this is a very real problem. Good players are constantly adjusting the pitch of notes to match notes played by other instruments. A flexible pitch instrument like a violin can always play better in tune than a fixed pitch instrument such as a piano. The errors in equal temperament tuning are much more noticeable on a pure sustained tone instrument such as a recorder than they are on a piano (where equal temperament tuning actually works very well). The classic example of this is tuning thirds - a chord with an equal temperament third is unsatisfactory when played on recorders.

Since the recorders of professional quality are still cheap by the standards of other musical instruments many players get around these problems by having different instruments for consort playing and for solo playing. However, with a solo instrument you are still compromising your ability to play in tune by extending the dynamic range of the instrument.

There are techniques to allow the recorder to play with a greater range of dynamics.

The basic approach is to alter the fingering in some way to alter the intonation slightly and then adjust the breath pressure to play the note in tune, but at a different dynamic level.

The first method is to change the standard fingering of a note for a so-called "dynamic fingering". For a loud note you choose an alternative fingering that is slightly flat and for a quiet note you choose a fingering that is slightly sharp. It is unfortunate that these fingerings can be quite specific to a particular instrument, but the effect can be impressive, particularly as the alternative fingerings can have a diferent tonal colour, heightening the effect of a dynamic change. This technique requires preparation, and a good example of a complete movement marked up with alternative fingerings can be found in Chapter 3 of Anthony Rowland-Jones' book "Playing Recorder Sonatas" (Clarendon Press, Oxford 1992).

Walter Van Hauwe lists a good basic set of dynamic fingerings in Book 3 of his series "The Modern Recorder Player", and Michael Vetter lists a good deal more in his book "Il flauto dolce ed acerbo" (Moeck Verlag 1969).

The second method is to keep the standard fingering but to slightly cover the next open hole on the instrument, or to slightly leak the lowest finger, depending on whether you want to flatten or sharpen the note in question. This can be very effective on long notes, allowing crescendo and decrescendo on a single long note, but is less effective on rapid music due to the difficulty in accurately placing the shading or leaking fingers at speed. To increase the control it is often adviseable to refinger the note for a forked fingering if possible, as the tuning changes will be less extreme and more controllable (especially when leaking fingers). A good description of the approach can be found in Anthony Rowland-Jones' book "Recorder Technique" (OUP 1986).

These are not new techniques: Ganassi writes: "An instrument can imitate the human voice by varying the pressure of the breath and shading the tone by means of suitable fingering" in his Fontegara (1535 Venice [trans. Dorothy Swainson 1956 Robert Lienau Musikverlag]).

In addition to these techniques where the loudness is actually altered the player has a number of musical devices to assist the effect of a change in volume. Note duration is often used by recorder players to simulate an echo - a slight staccato makes the notes seem quieter and lighter. The articulation weight and speed can also help to colour the notes, as can note placement slightly ahead or behind the beat.

As you can see from the descriptions above techniques aimed at extending the dynamics available to the recorder player are not easy. The player must commit a good deal of time to learning tables of alternative fingerings and to practicing some difficult finger techniques to balance the pitch change from the breath pressure alteration against leaking or shading fingerholes. It is not surprising, therefore, that a number of devices have been thought up to allow a more straightforward control of pitch during dynamic changes.

One group of devices works by having a small hole drilled through the sidewall or block close to the rectangular window. During normal playing the hole is closed by a key and the recorder plays as normal. When the player wants to play a passage at lower volume the small hole is opened by operating the key and the recorder sharpens slightly. Playing with reduced breath pressure then brings the pitch back down to the correct level. The device apparently works well, even if the amount of sharpening is greater for notes with fewer fingers down than for notes with more fingers down. It might be possible to extend the use of such a device by using a needle valve rather than a simple flap valve, so that partial opening would be controllable.

A number of manufacturers make devices based on this principle, and you can see a photo of a recorder equipped with a "lip key" on the Dolmetsch website.

Maarten Helder in conjunction with Mollenhauer has developed a "harmonic" tenor and alto which also incorporates a dynamic device. The main thrust of the development appears to be an adjustment of the normal tonehole arrangement so that the overtones of a particular fingering are in a harmonic series. This should radically alter the tone of the instrument, and the strengthening of the low notes allows the voicing to be adjusted to give a three-octave range. Truly a radical design! I have not heard one of these instruments yet and I've only seen photos, but I am really keen to try one out! There is a webpage on Nick Lander's Recorder Homepage which has some photos of the Mollenhauer/Helder instrument.

The second approach is startlingly simple. By using electronic amplification the volume can be altered simply by adjusting the gain of the amplifier. The microphone can either be attached externally to the recorder, mounted on a microphone stand, or mounted through the sidewall or block. Although the internally mounted microphones will produce a different tone to those mounted externally, the tone can be readily changed using equalisation equipment. Recorder maker Philippe Boulton has done a good deal of work in this area, and a brief description of his instruments which incorporate internal microphones appeared in the March 1998 issue of Recorder Magazine. I cannot find a website with pictures of this instrument (though it only looks like a recorder with a wire coming out of the side of the instrument near the window).

 A Flexible Embouchure - eliminating the fixed windway 

If these pitch-correction devices work reasonably how come they have not caught on? Are they actually solving a problem that the vast majority of recorder players do not see as a problem? Certainly the devices run contrary to the current thinking in the early music community about authenticity of instrument design - if the majority of advanced players are renaissance and baroque music specialists then I can understand why dynamic devices are not popular. However, the recorder is also an instrument used in contemporary music. Perhaps in these applications dynamics are arranged electronically if required?

The answer may also be in the fact that one of the attractions of the recorder is its simplicity (and hence cheapness). The add-on devices are seen as add-ons, and not an integral part of the instrument. Even if the devices are easier than the refingering techniques you still have to learn how much the pitch changes for each fingering, and unfortunately this is likely to vary with different instrument designs.

Several years ago it struck me that if the recorder player was going to commit to learning a new technique to acquire control over dynamics, why learn to control an external mechanism? The reason that the recorder pitch goes up as you blow harder is that the windway is of a fixed size. On a transverse flute you can increase the size of the aperture between the lips as you blow harder so that the speed of the air jet does not increase. You pass more air so the flute plays louder, but the pitch does not go up as the speed of the air jet does not increase.

So I started work on an instrument that would combine the elegance of the recorder's fingering system with a more flexible embouchure system. I wanted to produce an instrument that would allow a straightforward chromatic scale to be played without keywork and be capable of expressive variation in tone and loudness. Because I wanted to retain as much of the recorder's tone and character I started by looking at end blown designs.

If you cut off the beak of the recorder (no - don't do it it is a bit more complicated than that!) you get what is called a notch-flute. The japanese Shakuhachi and the Andean Quena are two examples of this class of flute. It is played by putting the head against the lower lip and chin so that the edge is level with the gap between the lips. Formation of the lip shape, angle of blowing and air pressure must be learnt, as with the transverse flute, but the fingering can be made the same as for a recorder.

I took as my first example the cylindrical soprano instrument found in Dordrecht, and thought to date from the mid fifteenth century. I chose this instrument as it was simple to prototype, and as I had already made a number of recorders based on this design I was familiar with its tuning. After a number of prototypes to develop a working set of dimensions for the notch I produced this instrument:

The tone is powerful and "open" and the intonation is good. The tone was a bit of a surprise, sounding far more like a flute than a recorder. However, the notch gave a good control over loudness, and tone could also be varied (even though I was very inexpert at doing so). This was sufficiently promising for me to think about making the musically more interesting lower instruments in the recorder family as notch flutes.

I next tried to develop a treble instrument. I knew from previous work that simply scaling up the Dordrecht instrument to a treble size was not going to work, as the toneholes get unacceptably large and far apart. I therefore based the instrument on a modern "baroque" style recorder with double holes for the lower two right hand holes. In addition to using a tapered bore I needed to make a number of other changes. The soprano instrument is so light that friction from the right thumb against the underside keeps it in place. However since the treble receives less support from the lip than a normal recorder I found it necessary to have a thumb rest. I formed this as an integral part of the instrument:

The end of the soprano instrument was of a sufficiently small diameter that it sealed easily against the upper part of the lower jaw. The instrument could be tilted very slightly to bring the plane of the ramp parallel to the direction of blowing. On the treble instrument I needed to shape the end of the instrument to make this tilting unnecessary, as the instrument was difficult to seal, and tilting broke the seal very quickly:

As can be seen from the photo the end of the tube is at right angles to the plane containing the notch.

Interestingly this seems to be a fundamental difference between quena and shakuhachi playing technique. Shakuhachi are routinely played with the lower "corner" of the tube resting against the chin, and the upper portion of the tube (containing the notch) held away from the lips. Quena are usually played with the end of the tube pressed against the chin. This has a marked effect on the ease of playing. When the tube is sealed against the chin the returning pressure pulse is forced back through the window, making it much easier to sustain a tone.

I had finally produced a workable end blown instrument with recorder fingering. However, I was not at all happy with the tone, and I didn't know if the fault was with the instrument or with my playing technique. I realised that to take the design any further I would need to invest a large amount of time to learn to play the notch flute properly, so that I could asses further changes to the design. I could not afford to take the time away from flute and recorder practise to do this, so I shelved the design to be picked up at a later date.

When the photographs above first appeared on this website I was contacted by a couple of quena makers who kindly passed on some advice about improving the tone of the instrument. If I return to the end blown flute approach I will try out new notch geometry and see if this helps.

Several months later I reviewed the design and thought that perhaps there was another way around this problem.

I was struck by a phrase in the introduction to Ardal Powell's translation of "The Keyed Flute" by JG Tromlitz:

I have been making and playing baroque flutes for a number of years, so this was familiar territory. Perhaps by using the headjoint from a baroque or early classical flute coupled with a recorder bore I could retain the recorder-like sound, with the tone and volume flexibility of the baroque flute? A "Recorder-System Flute" perhaps?

After some initial experimentation to establish the basic proportions for the headjoint I set out to see what sort of embouchure would best suit this new instrument.

Initially I tried embouchures derived from my work on modern one-key flutes. These allow the one key flute to produce more power than the typical embouchure cut that was used during the baroque period. The dynamic range is therefore greater and high notes are easier. If care is taken in the design of the embouchure the pitch flexibility (essential to maintain good intonation on a baroque flute) can be greater allowing for small tuning movements, leading to stronger forked chromatic notes. The tone is brighter and sounds less "covered" than a typical eighteenth century English flute.

The tonal alteration caused considerable problems in this instrument, probably due to the steeper body taper of the recorder body. The transition between the first and second register was accompanied by a marked change in tone. The tone of a treble recorder typically brightens as you rise through the first register, and the first notes of the second register are usually a little more covered in tone. The change between g'' and a'' can often be noticed in passage work for this reason. However, in brightening the overall tone I found that I had accentuated this natural tendency in the recorder bore.

I found that a more conservative embouchure cut allowed a smooth transition across the registers, and a softer more recorder like tone. A side-effect of this conservative cut is the range is limited to that of a normal treble recorder, whereas I had hoped to be able to extend the top end. If I want to extend the range upwards I will almost certainly need to change the instrument so that its bore is more similar to a baroque flute bore.

The position of the holes for the left hand also provided me with a good deal of difficulty, as they dictate the shape that the left hand adopts whilst playing. The left hand must be free from tension yet provide sufficiently good support for the instrument so that finger movement is not limited. If the fingers are helping to grip the instrument then there is a delay in lifting them so smooth rapid playing is difficult, and the flute is poorly supported playing f#'' (-/12-/----) and g'' (-/-2-/----).

The position of the thumb is difficult - it is unlike any other flute I have seen as the thumb lies above the first finger. Typically on a flute the thumb will lie somewhere between the position of the first and second fingers, falling into a nice neutral and relaxed position. Also, as the thumb covers a hole rather than a key there is less latitude in the position that the thumb can adopt and still seal the hole. The best compromise I found was to roll the hole around to the side of the flute slightly so that it could be closed with the side of the thumb, and to offset the second and third fingerholes.

From the picture above you can just make out that the fingerholes are not in a straight line down the flute, and the thumbhole is shifted around slightly to the side of the flute. This allowed a more curved palm shape and tucks the knuckle of the left-hand index finger a little under the flute so it helps to support the weight of the instrument.

The lowest two holes are made as double holes in the normal manner to allow easy playing of f'# and g'#. I did not offset the holes for the right hand as I don't have any trouble reaching them. As far as I can tell it is far less common to see the lower holes offset on a recorder, whereas the upper holes on plastic instruments are often slightly offset.

 Because this is a new instrument I have enjoyed myself and designed a more modern turned appearance than is normal for a recorder. You can see the complete flute below:

I now have a recorder-system flute that is in tune with itself, at modern pitch, and ergonomically good. Since this is a new flexible embouchure instrument I need to get used to it before I can judge the overall tone. This means that I need to spend a couple of months solid playing to give it a good work out.

Future development of the alto may involve alteration of the bore to reduce the taper which will increase the range and smooth the register changes even further, but I may run into problems with finger reach.

I'd also be interested in making a soprano based instrument, as it would share many fingerings with a high D whistle and might be a useful chromatic instrument for Irish music (and allow recorder players to join in a session with an instrument that sounds right).

I hope you have enjoyed this description of the current state of the project. If you have any suggestions for improvements or further work then please contact me using the link on the home page.

Back to home page