GB2320125A - Controlling the characteristics of an audio signal - Google Patents

Abstract

A characteristic of an audio signal produced by an electronic musical instrument is modified in response to manual operations applied to a controlling device. The manual operations are preferably effected by operation of a foot pedal having two degrees of freedom such that it may be rocked about a horizontal axis (fig. 6) and rotated about a vertical axis 406. For each degree of freedom there is a resonance circuit, having a coil 601, 1001 and a cooperating member 203, 206 such that the inductance, and thus the resonant frequency of the circuit, is varied by relative movement between the coil and the cooperating member. The relationship between manual operation and audio signal modification is linearised by a compensating characteristic determined by the shape of the coil.

Description

Title: CONTROLLING CHARACTERISTICS OF AN AUDIO SIGNAL The present invention relates to an apparatus and a method for controlling audio signal processing equipment.

INTRODUCTION Electronic musical instruments and audio effects provide a variety of sounds for musicians. Techniques for creating and manipulating audio signals continue to be developed, particularly in the field of digital audio. The cost of digital techniques, often referred to as digital signal processing, is decreasing. Key digital technologies including conversion between the analogue and digital domains, standards for digital audio interconnections and data storage continue to mature, allowing advances to be made to the sophistication of digital musical equipment.

The evolution of the midi standard and associated computer software for controlling musical events and processes via midi has provided many ways for musicians to control the increasing range of sounds which have become available.

The midi sequencer is a tool which enables musicians to transcend manual playing techniques and assemble music of arbitrary complexity. Midi may also be used to communicate live performance data to a synthesiser or effects processor.

Performance data may be generated by a midi device such as a modulation wheel, breath controller or foot pedal. However, the amount of data which may be generated during a live performance is limited compared to that which may be accumulated over several passes of a piece of music when using a sequencer.

SUMMARY OF THE INVENTION According to a first aspect of the present invention, there is provided apparatus for controlling audio signal processing means configurable to modify a characteristic of an audio signal in response to a manual operation applied to said controlling apparatus, wherein a said controlling apparatus includes a resonance circuit including inducting means having an inductance adjustable in response to said manual operation, such that the resonant frequency of said resonance circuit is adjustable in response to said manual operation.

Preferably, the inducting means comprises a coil and a co-operating member, wherein said inductance is varied by relative movement between said coil and said co-operative member. Preferably, the shape of the cdil is defined tp 0 produce substantially linear changes in the resonant frequency of said resonance circuit in proportion to said manual operation. Altematively, the shape of the cooperating member may be defined to produce substantially linear changes in the resonant frequency of said resonance circuit in proportion to said manual operation.

According to a second aspect of the present invention, there is provided a method of modifying a characteristic of an audio signal in response to manual operation, comprising steps of connecting a control device to an audio signal processing means, wherein said control device includes a resonance circuit having a capacitance means and an inducting means adjustable in response to said manual operation, such that the resonant frequency of said resonance circuit is adjustable in response to said manual operation.

In a preferred embodiment, the resonance circuit includes a high gain inverter and a tuned circuit arranged to generate a square wave at the resonant frequency of said tuned circuit. Preferably, the inductance is connected to an output of said inverter via a resistance and said capacitance includes a first capacitor and a second capacitor, said first capacitor is connected from said input of said inverter to a fixed voltage; said second capacitor is connected from said fixed voltage to the junction of said inductance with said resistance; and said capacitors are considered as being arranged in series thereby substantially forming the capacitance in a parallel tuned circuit which includes said inductance.

The inverter may be an HCMOS inverting buffer and the resonance circuit may resonate at radio frequencies.

In a preferred embodiment, variations in resonant frequency are conveyed to a frequency measuring means by a process of pulse width modulation. The pulse widths may be determined by counting the frequency of resonance and a plurality of variable inductances may be provided within respective resonance circuits. Preferably, the plurality of resonance circuits are individually selectable for activation in response to a multiplex control signal and a resonance circuit may be activated in response to a signal supplied to an additional input of said inverter.

Preferably, pulse widths are multiplexed in time to convey respective frequencies of said resonance circuits.

Preferably, manual operations are effected by operation of a user's foot and the controlling device may have a plurality of modes of freedom and different audio signal characteristics may be modifiable in response to different foot movements. In a preferred embodiment, two modes of freedom are provided which may be derived by the controlling device being configured so as to rock back and forth about a substantially horizontal axis and to rotate about a substantially vertical axis. Preferably, the horizontal axis is located in-front of said vertical axis, so as to facilitate preferred modes of freedom of said foot.

Preferably, means are provided for counting two resonant frequencies altemately to generate high and low pulse widths on a shared multiplex, wherein said high and low pulse widths respectively convey frequencies of respective resonance circuits. The altemate high and low pulse widths may be supplied as a multiplex control signal to selectively activate the resonance circuits.

BRIEF DESCRIPTION OF THE DRAWINGS Figure 1 shows an sound recording studio equipped with recording equipment, synthesisers, audio processors, a midi computer sequencer, and a foot pedal controller and a foot pedal processor; Figure 2 details a side view of the foot pedal controller shown in Figure 1, including a pedal, a moveable sub-chassis and a vertical mode of rotation; Figure 3 details a top view of the foot pedal controller shown in Figure 1, including a horizontal mode of rotation; Figure 4 details construction of the moveable sub-chassis and foot pedal facilitating the vertical mode of rotation shown in Figure 2, including a vertical sensor block and a support block; Figure 5 details a rear view of the construction shown in Figure 4; Figure 6 details construction of a coil in the vertical sensor and interaction with the support block shown in Figures 4 and 5, at a first vertical angle; Figure 7 shows the apparatus shown in Figure 6 at a second vertical angle; Figure 8 details an aspect of the construction of the foot pedal controller shown in Figure 1 for facilitating limited horizontal rotation, shown from above; Figure 9 further details aspects of the construction shown in Figure 8, including an electrical connection to the vertical coil sensor shown in Figures 6 and 7; Figure 10 details construction of a horizontal coil sensor and its interaction with the moveable sub-chassis at a first horizontal angle; Figure 11 shows the apparatus of Figure 10 at a second horizontal angle; Figure 12 details an electronic resonating circuit of the type used with the vertical and horizontal coil sensors shown in Figures 6 and 10; Figure 13 details frequency ranges obtainable from the circuit shown in Figure 12 in conjunction with the coil sensors shown in Figures 6 and 10; Figure 14 details the circuit of the foot pedal controller, including resonator circuits of the type identified in Figure 12; Figure 15 details a wavefonn generated by the circuit shown in Figure 14; Figure 16 details circuitry contained in the foot pedal processor shown in Figure 1, for reconstructing the waveform of the type shown in Figure 15 from signals supplied from the foot pedal controller circuit shown in Figure 14 over several metres of electrical cable; Figure 17 details other circuitry contained in the foot pedal processor shown in Figure 1, including a digital signal processor; Figure 18 details operations performed by the digital signal processor shown in Figure 17 when calculating positions of the foot pedal in both dimensions, including processes for calculating a vertical position and calculating a horizontal position; Figure 19 details the process for calculating the vertical position and the process for calculating the horizontal position shown in Figure 18, including applying a scaling constant, an offset constant and a look-up table; Figure 20A details processes for calibrating the foot pedal, including a process of calculating a scaling constant and offset constant shown in Figure 19 for each dimension; Figure 20B details the process of calculating constants shown in Figure 20A; Figure 21 details the horizontal and vertical look-up tables applied in Figure 19; Figure 22 details data flow between processes operating on the digital signal processor shown in Figure 17, including supplying pedal position data to a DSP function; Figure 23 represents a first application of supplying foot pedal data to a DSP function shown in Figure 22; and Figure 24 represents a second application for supplying foot pedal data to a DSP function shown in Figure 22.

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS The invention will now be described by way of example only with reference to the accompanying drawings identified above.

A sound recording studio is shown in Figure 1. A mixing desk 101 receives audio signals from a first rack unit 102, a second rack unit 103, a microphone 104 and a synthesiser 105. Most of the cabling has been omitted for the sake of clarity.

The first rack unit 102 includes individual nineteen inch rack-mounted modules for generating and processing sounds. The second rack unit 103 includes two synchronised eight track digital audio tape recorders 106 and 107. These are able to receive as well as supply analogue audio signals from and to the mixing desk 101. Thus sixteen tracks are available on which audio signals may be recorded, and tracks may be selected individually for recording or playback via the mixing desk 101. The tape recorders 106 and 107 are remotely controlled by a tape controller 108, which facilitates convenient control of the tape machines from a position in front of the mixing desk 101.

The second rack unit 103 further contains a power amplifier 109, which receives a stereo audio signal from the mixing desk 101. The amplified stereo signal is supplied to a left monitor loudspeaker 110 and a right monitor loudspeaker 111.

The synthesiser 105 may be used to supply analogue audio signals to the mixing desk 101. Additionally, playing the synthesiser or adjusting a synthesiser control 112, may result in midi codes being generated. These may be supplied to a computer 113, running sequenceing software. Midi and audio signals may be generated simultaneously by the synthesiser 105. Midi signals from the synthesiser 105 may also be supplied to midi controllable modules in the first rack unit 102. Such modules include a synthesiser module 114 and a reverb effect module 115. Thus midi signals supplied from the synthesiser 105 may be used to control additional tone generation resources in the synthesiser module 114, or control effect attributes of the reverb effect module 115.

Midi signals generated by the synthesiser 105 may be recorded using the computer sequencer 113, and subsequently replayed automatically. Additionally, the computer sequencer 113 enables changes to be made to recorded midi events either by editing or by algorithmic modification or enhancement. Additional midi events may be superimposed, or overdubbed, onto existing midi recordings, thereby accumulating expressive information in the form of keyboard 116 events and manipulation of synthesiser controls 112.

Signals from the several audio sources described above are supplied to the mixing desk 101. The mixing desk 101 has a number of channels 117, to which an individual sound source may be connected. This sound source may be a track on a tape, or a piece of studio equipment such as a synthesiser or the microphone 104. Each channel 117 has a fader which controls the overall level of the sound which will be supplied to the final stereo signal. Altematively, when recording, the fader may also be used to control the level of signal supplied to a tape track. Each channel has the ability to supply a proportion of its incoming signal to several audio busses, which are connected to various audio effects processors mounted in the first rack unit 102. In addition to the reverb unit 115, other effects include compressors 118 and 119, and digital multi-effects units 120 and 121. Signals generated by these processors are supplied back to channels on the mixing desk, which have adjustable levels for inclusion in the overall stereo mix.

The computer sequencer 113 is synchronised to the tape recorders 106 and 107, such that activation of tape transport controls 108 also results in corresponding actions within the midi sequencer. Thus, having assembled a number of recorded audio and midi tracks on the tape recorders 106, 107 and the computer 113 respectively, a final mix may be effected by appropriate manipulation of mixing desk controls, effects processors and synthesisers.

The studio as described so far includes a variety of methods for creating and manipulating sound. In a studio a high degree of control is available through the use of computers. In addition to the computer sequencer 113, digital effects units employ computer-driven interfaces which provide the user with a variety of sound processing options. Most computer based musical equipment may be controlled in real time using midi control sources and or a midi sequencer.

Midi control sources such as those provided on a synthesiser control panel 112, or the velocity of keys 116, provide an important degree of expression, which may be recorded in a midi sequence. However, there are a large number of parameters that are available for control via midi. For example, tone, reverb, stereo position, filtering and volume may all be simultaneously controllable.

However, operating all five of these simultaneously through separate controls on various items of equipment is unlikely to yield an expressive result compared to that which might be desired, so such an effect must be achieved by overdubbing.

However, many of these parameters are controllable when playing acoustic instruments. It is this degree of expression which known studio equipment theoretically provides for electronic sounds but in practice cannot, due to the lack of immediacy of the computer interfaces available. It is therefore necessary to rely on computer sequencing or mixing techniques to modify a recorded performance.

This is an inherently less expressive and more time consuming process.

The studio shown in Figure 1 further includes a foot pedal controller 151 which is operable in two dimensions rather than in the usual one dimension. The foot pedal controller 151 supplies signals to a foot pedal processor 152, which includes a circuit for converting signals received from the foot pedal controller 151 into numerical values. These are supplied internally to control digital signal processing functions. In addition, the numerical values may be supplied to a midi output of the foot pedal processor 152, thereby facilitating control via two dimensions of any item of midi-compatible studio equipment.

It should be appreciated that simultaneous control of two parameters by the foot pedal controller 151 is more powerful than simultaneous control by, say, two individual foot pedals operable in single dimensions. The parameters which are controlled in each dimension may be selected such that movement in the two dimensions facilitates an exploration of the available sound space. The sound space may be considered as being defined by the permutations of parameters mapped to the two pedal dimensions. This sound space corresponds to a clearly identifiable physical space, as defined by the range of possible pedal positions.

This is highly intuitive, and therefore useful in musical expression.

The foot pedal controller 151 shown in Figure 1 is detailed in side view in Figure 2. An operator's foot 201 rests upon a foot pedal 202. The pedal 202 is mounted on support blocks 203 and 204, only one of which is visible from this view. A horizontal axle 205 passes through and rotates with the support blocks 203 and 204. The horizontal axle also passes through a moveable sub-chassis 206. Movement of the sub-chassis 206 facilitates horizontal rotation of the foot pedal 202. The sub-chassis 206 rests on a foot pedal base 207. A plug 208 is shown connected to the foot pedal base 207, connected to a cable 209, which supplies position signals to the foot pedal processor 152, shown in Figure 1.

The foot pedal 202 is vertically rotatable about a centre defined by the horizontal axle 205, and has a maximum extent of fifty degrees. This is a much larger freedom of rotation than that provided by many known single dimensional foot pedals and is facilitated by the cut-away shape of the sub-chassis 206, in combination with the technique employed for converting angular position into an electrical signal. The moveable sub-chassis 206 rests on the pedal base 207 and is able to rotate horizontally. In Figure 3, the foot pedal controller 151 is viewed from above and is shown having a maximum extent of horizontal rotation of seventy degrees.

The sub-chassis assembly shown in Figure 2 is further detailed in Figure 4.

The top surface of the foot pedal 202 is covered by a roughened rubber surface 401, which is necessary to provide enough friction to rotate the pedal horizontally.

The foot pedal 202 is made of flat aluminium plate which is 0.25 inches thick. Bolts 402 and 403, concealed from above by the roughened rubber surface 401, fix the support block 203 to the foot pedal 202. Additional bolts fix the other support block 204. A grub screw 404 passing through the support block 203 fixes the support block 203 to the horizontal axle 205, and thereby the foot pedal 202 also.

The horizontal axle 205 passes through the sub-chassis 206, and is free to rotate. The axle 205 is made of aluminium. The sub-chassis 206 is also made of aluminium, and a plastic lining is used to ensure that the axle 205 is able to rotate in its locating holes in the sub-chassis 206 without excessive friction or wear. The support block 203 passes in front of a sensor block 405, which is fixed to the moveable sub-chassis 206. The sub-chassis is horizontally rotatable about a vertical axis defined by a locating bolt 406. In order to minimise friction between the sub-chassis and the surface of the pedal base 207, a self lubricating surface interface is provided. Thus, the base of the moveable sub-chassis 206 has a selflubricating plastic layer 407, which is selected with the surface of the pedal base 207 to minimise friction and wear resulting from horizontal rotation of the pedal. In the present embodiment a layer of polytetrafluoroethylene (PTFE) forms layer 407 and the top surface of the pedal base 207 has a layer of the substance commonly sold under the trade name "Formica".

Both axes of rotation for the pedal are shown in Figure 4. The horizontal axis, defined by the axle 205, is offset forward from the vertical axis, defined by the bolt 406. Thus the planes of rotation are offset from each other. This design reflects the natural rotation planes of the human foot. Rotation of the foot in a horizontal plane about a vertical axis may be considered as having a centre defined by the heel, and proceeding vertically up towards the knee. Rotation of the foot in a vertical plane may be considered as occuning about an imaginary horizontal axis running from left to right in front of the heel. If the horizontal axis is placed further forward with respect to the vertical axis than the position shown in Figure 4, it becomes difficult to comfortably rotate the pedal simultaneously in both dimensions. Thus, the present embodiment positions the axes at positions considered optimal for the required modes of movement.

The apparatus shown in Figure 4 is shown viewed from the back in Figure 5. For clarity, the pedal 202 is shown in a flat position. Here the two support blocks 203 and 204 may be seen separately. The support blocks 203 and 204 are separated from the sub-chassis 206 by plastic washers 501 and 502. These provide sufficient friction to maintain the pedal at its vertical angle when no pressure is applied from the foot 201. The sensor block 405 is positioned adjacent to the support block 203. The sensitive side 503 of the sensor block 405 is positioned very close to the support block 403. The sensor block 405 is fixed to the sub-chassis 206 by two screws, one of which 504 is shown in this view.

Design and operation of the vertical sensor is shown in Figures 6 and 7. In Figure 6 the sensor block is shown with a coil 601 mounted in close proximity to the support block 203. The support block 203 is made of aluminium therefore movement of the foot pedal results in a change in the amount of aluminium in close proximity to the sensor coil 601. The sensor coil 601 forms part of a tuned circuit. The inductance of the coil 601 varies depending on the amount of aluminium in dose proximity to it. The resonant frequency of the tuned circuit is thereby affected by the vertical angle of the pedal. Figure 7 details the relative positioning of the coil 601 and support block 203 for a different vertical angle. The coil is formed by winding fifty turns of 0.125 mm enamelled copper wire on an 18 mm diameter former, and then squashing it to form the shape shown. The inductance of the coil is approximately 20 micro henries and the shape and size of the coil is chosen to optimise linearity and to maximise the variation in resonance frequency for the range of pedal positions to be measured, in this case covering fifty degrees of movement. Altematively a circular coil could be used, with a carefully chosen shape for the part of the support block which passes in front of it.

This latter arrangement is more difficult to optimise experimentally, but may be more suitable in a manufactured product.

Further details of the sub-chassis relating to horizontal rotation are shown in Figure 8. The pedal base 207 beneath the sub-chassis 206 has a cut-out shape 801. A limiting bolt 802 passes through the sub-chassis and is located somewhere in the cut-out area 801. Horizontal rotation of the sub-chassis is limited by the extremes to which the limit bolt 802 can move within the cut-out 801. Thus, the seventy degree limit of horizontal rotation is defined in this way. Electrical connections to the vertical coil sensor block 405 shown in Figures 6 and 7 may be made to circuitry located in the pedal base via the cut-out 801.

A side view of the arrangement shown in Figure 8 is shown in Figure 9. The limit bolt 802 is shown having a shock-absorbing sleeve 903, through which wires 901 may pass, thereby forming a convenient route for the electrical connection between the vertical coil sensor 601 and the pedal controller circuitry located in the foot pedal base 207. The foot pedal base is constructed from two layers of chipboard 904 and 905, the top layer 904 having the cut-out 801 shown in Figure 8 and a top surface of Formica. The lower layer 905 is largely cut away, except at its edges, thus facilitating a space in which the circuit board is located. A thin aluminium base plate 906 provides covering and screening for the circuitry on the underside of the pedal base 207.

Design and operation of the horizontal coil sensor is detailed in Figures 10 and 11. The sub-chassis is horizontally rotatable about the vertical bolt 406 and is made of 0.25 inch thick aluminium. A coil 1001 is embedded in the top surface of the foot pedal base 207. The shape of the coil 1001 is chosen to maximise the linearity of the relationship between the resulting resonant frequency and the horizontal angle of the sub-chassis. The shape is also chosen to maximise the range of resonant frequencies for the angle of rotation which is to be measured, which is seventy degrees. The coil is constructed by winding 25 tums of 0.125 mm enamelled copper wire on a 2.2 inch diameter circular former, and then squashing to the shape shown. Again, the same effects of linearity may be achievable by a suitably-shaped volume of aluminium in the base of the moveable sub-chassis, but this is difficult to obtain experimentally and so has not been used in the present embodiment. The coil has an approximate inductance of 20 micro henries.

Figure 11 shows the relative positions of the sub-chassis and horizontal coil sensor 1001 for a different horizontal pedal angle.

In summary, the coil sensor shaping may be considered as linearising the relationship between resonant frequency and pedal angle by modifying the relationship between the effective permeability of the coil core and the physical distance between the coil and the moveable core. This may be achieved by selecting a particular shape for the coil, by selecting a particular shape for the core or by a combination of the two.

The design of known pedals suffers from several problems, depending on the angle-sensing technology employed. Cheaper pedals use geared potentiometers. Potentiometers, usually of carbon track construction, eventually become dirty and wear out, resulting in considerable electrical noise when the pedal is moved, and false readings for stationary positions of the pedal. These problems are known to users of guitar effects pedals in the form of loud electrical noise imposed on the amplified guitar signal.

A known improvement on potentiometer-based foot pedals is to use a Halleffect sensor. A Hall effect sensor generates a voltage in proportion to the strength of a local magnetic field, provided by a movable magnet coupled to the position of the pedal. Hall effect sensors are therefore immune to the problems of wear and electrical noise associated with potentiometers. However, Hall effect sensors are expensive and require a stabilised power supply of 20mA at 5V, typically. Problems of temperature-related drift emerge when attempting to design a pedal having a wide angle of movement, made worse by the significant power dissipation which can cause self heating in the sensor chip. Steps may be taken to reduce these problems, using techniques such as ovening. However, at this point in the design cycle, it becomes clear that a Hall effect sensor does not provide the ideal solution.

A lesser problem with both the above solutions is that a voltage is generated, which must then be converted to a digital value. Many midi applications will be satisfied with an angular resolution of 128 steps, or even substantially less. However, neither potentiometers nor hall effect transducers provide a simple solution for measuring angles to a much higher precision, for example to 4096 steps, which would represent substantially continuous measurements. To do this, an expensive twelve bit analogue to digital converter would be required.

A third known solution is optical quadrature encoding. Optical quadrature sensors are widely used in mouse pointers for personal computers, and have become affordable for use in foot pedals as a result. An optical grating passes before two infra-red light emitting diodes. Detector photodiodes in line with their respective light emitting diodes may receive light or not, depending on the position of the grating. As the grating moves, it is possible to decode the direction of travel by quadrature detection, and the magnitude by the number of pulses received by a photodiode. The precision of this technique is restricted by the physical size of the components and the size of the pattem of the optical grating. Known implementations have resolutions of as few as fifty steps.

In Figure 12 a schematic diagram is shown for a resonating circuit of the type used with sensor coils 601 and 1001. An inverting HCMOS logic buffer 1201 supplies a signal to a resistor 1202. A sensor coil 1203 is connected from the resistor to the input 1204 of the buffer 1201. Capacitors 1205 and 1206 are effectively in series, and form a parallel mode tuned circuit with the sensor coil 1203. The output 1207 of the buffer 1201 is a square wave 1208 having a frequency given by the equation shown. The precise inductance L of the sensor coil 1203 depends on the position of the aluminium which passes in front of the coil. Thus the frequency of the square wave 1208 can be used to determine the position of the pedal.

The resistor 1292 is selected to minimise the current consumption of the circuit, and to minimise radio frequency radiation. Too high a value for the resistor will result in unstable or non-resonant oscillation. The value of 10 K ohm is considered optimal. The power supply for the HCMOS buffer is 3.3 volts. Different combinations of values for components and power supply voltages do not necessarily sustain resonant oscillations.

The resonant frequency of the tuned circuit formed by the sensor coil 1203 and the capacitors 1205 and 1206 is largely immune to fluctuations in power supply voltage, and other circuit conditions. Thus these three components will d

Two circuits of the type shown in Figure 12 could be used to supply signals to frequency counter circuits for converting frequencies into precise numerical values. However, if the signals are to be transmitted along a cable, interference between the two may result. Furthermore the frequencies of oscillation are in the medium wave transmission band, and might cause interference with radio receivers or other sensitive equipment. At the very least it would be preferable to reduce the frequencies of oscillation to much lower values, and therefore avoid the problems of driving a high frequency square wave down a long cable. Reducing the frequency does not reduce accuracy of measurement. The square wave may be reduced by a factor of a thousand, and then a pulse width measurement may be canied out, rather than a frequency measurement. This still leaves the problem of interference between the signals, however.

A circuit which solves all these problems is shown in Figure 14. The horizontal oscillator circuit is formed around an HCMOS NOR gate 1401, and the 'vertical oscillator circuit is formed around another NOR gate 1402. Each NOR gate has two inputs, one of which may be used for the resonating circuit, the other of which may be used to disable oscillations by supplying a logic 1 signal, in this case corresponding to 3.3 volts. A third NOR gate 1403 is wired to act as a simple inverter. The inverting NOR gate 1403 facilitates selective activation of either the horizontal oscillator or the vertical oscillator. When the input to the inverter 1403 is high, the horizontal oscillator is enabled. When the input to the inverter 1403 is low, the vertical oscillator is enabled.

When an oscillator is disabled, its output is forced to remain low. Thus, by supplying both oscillator outputs to the inputs of a fourth NOR gate 1404, the output of this NOR gate will provide an inverted form of the square wave from whichever oscillator is activated. This may not appear to be useful, as there appears to be no way of distinguishing which of these oscillators is selected just from looking at the output from the fourth NOR gate 1404. This signal is supplied to a counter 1405, which divides the frequency of the square wave by 1024.

Considering the situation when the output 1406 of the counter 1405 is low, this output selects the vertical oscillator circuit for activation. The counter output 1406 remains low until the vertical oscillator has changed state 512 times. Then the counter output 1406 goes high, selecting the horizontal oscillator, and remains high until the horizontal oscillator has changed state 512 times. Thus the high period of the counter output 1406 depends on the amount of time taken for the horizontal oscillator to go through 512 state changes, and the low period of the counter output 1406 depends on the amount of time taken for the vertical oscillator to go through 512 state changes.

The output 1406 of the divider 1405 is also supplied to a resistor 1407, which drives the base of a Darlington pair transistor 1408. The resistor 1407 is required to reduce current consumption to a minimum. A capacitor 1409 ensures that the Darlington transistor 1408 is able to switch on and off at high speed. The collector of the Darlington transistor 1408 is connected to a line balancing resistor 1410, which, in conjunction with resistors at the other end of the connecting cable, prevents line reflections from distorting the waveform. Signals are supplied from the resistor 1410 to a socket pin 1411, which forms the output of the foot pedal controller 151. A power regulating circuit 1412 supplies a 3.3 volt one milliamp power supply to the two HCMOS logic chips. The four NOR gates 1401 to 1404 are part of the same standard HCMOS integrated circuit, a 74HC02. The counter 1405 is also a standard widely available integrated circuit, the 74HC4040. Both chips have an extremely low power consumption, thereby avoiding problems of circuit heating which might cause oscillator drift. Three connections are made to the circuit shown in Figure 14. The output,1411, an unregulated low power +5V to +15V input 1413, and a zero volt or ground connection 1414. These three connections can be made by a standard high quality XLR cannon type plug 208 shown in Figure 2.

The waveform at the output 1406 shown in Figure 14 is shown diagrammatically in Figure 15. This waveform has an approximate frequency of 500Hz. The horizontal width and the vertical width have a duration of approximately two milliseconds, which can be measured accurately using pulse width measurement techniques. By supplying a multiplexed signal along a single connection, problems of interference are avoided.

The foot pedal processor 152 shown in Figure 1 receives signals from the foot pedal controller 151 via the standard XLR-equipped three core cable. The length of this cable may run to several metres, and so it is important to balance the impedance at each and of cable, and thereby avoid problems of ringing and reflection, which might otherwise reduce the accuracy of pedal position measurements. A circuit in the foot pedal processor 152 for receiving signals, and supplying a suitable power connection to the foot pedal controller 151, is shown in Figure 16. A resistor 1601 limits the current suppliable to the power input 1413 of the pedal, thus protecting the equipment from accidental short-circuit. The signal from the foot pedal output 1411 is supplied to two resistors, which effectively match the impedance of the resistor 1410 in the pedal circuit, when a diode drop across the transistor 1604 is taken into account. A diode 1605 prevents reverse base-emitter voltage breakdown, again in case of a faulty connection being made to the pedal. A resistor 1606 pulls the output 1607 low when the transistor is switched off. The reconstructed waveform is a square wave having a magnitude of five volts, making it suitable for being supplied to a pulse width measuring circuit.

The remainder of the circuitry contained in the foot pedal processor 152 is shown in Figure 17. The reconstructed pedal waveform is supplied to the timer pin of a Motorola DSP56000 digital signal processor, details of which are available from Motorola Literature Distribution; P.O. Box 20912; Phoenix, Arizona 85036.

The timer pin is designated TIO, for Timer Input or Output. The digital signal processor 1701 has associated memory 1702, which includes static, non-volatile and read-only memory. The read-only memory stores instructions for the functions of the foot pedal processor.

The digital signal processor is able to receive digital audio signals from an analogue to digital converter 1703, and supply digital audio signals to a digital to analogue converter 1704. Analogue audio signals are thereby transferable to and from other studio equipment, thus enabling effects processing to be performed on such signals by the digital signal processor 1702. A user interface 1705 includes a menu-driven display and various buttons, so that the user can navigate through a number of options and determine the type of audio processing which is to be performed on the digitised audio signals. Furthermore, it is possible to determine how movements of the foot pedal control characteristics of a particular process.

Although certain audio processes may be performed by the foot pedal processor 152, other processing devices can be controlled by the foot pedal via midi. Thus the digital signal processor is capable of receiving and transmitting midi signals via a midi interface 1706.

Instructions residing in the memory 1702 control the processor to perform certain operations. One of these operations is conversion of the signal supplied to the timer pin TIO into numerical values suitable for control signal processing functions or transmission via midi. Processes performed by instructions for converting the foot pedal waveform into numerical values are summarised in Figure 18. The digital signal processor operates substantially in an interrupt driven processing environment. In other words, processing is performed in response to electrical events. Timing circuitry associated with digital to analogue conversion generates a first interrupt at the audio sample rate of 44.1 kHz. Thus, this interrupt transmits a left and a right audio sample to the digital to analogue converter 1704, and receives a left and a right sample from the analogue to digital converter 1703.

Left and right samples are constructed in batches of 32. A second intenupt signal is generated once every 32 samples, to request processing and calculation to generate the next batch of 32 samples. This interrupt signal has a frequency of 44100/32 or about 1.38 kHz.

In the present embodiment this 1.38 kHz interrupt provides a suitable point to sample the TIO pin state, as this is expected to change at less than half this frequency. At step 1801 in Figure 18, the digital signal processor is instructed to wait for the next 1.38 kHz intenupt. When this interrupt is detected, control is directed to step 1802, where a check is made to see whether the state of the TIO pin has changed. If the TIO pin has not changed, control is directed to step 1808, where other 1.38kHz interrupt-driven tasks are performed, such as signal processing of the next batch of 32 left and right samples.

Altematively, if the TIO pin has changed state, control is directed to step 1803. At step 1803 the timer control bit INV is inverted. The DSP56002 timer is operated in timer mode 5 (period measurement mode). This is described in detail in the DSP56002 User's Manual, available from Motorola. A rising or falling edge of the TIO pin results in a 24 bit counter value being transferred to a register. This counter is incremented at twenty megahertz. If the control bit INV is high, the counter value will be transferred on the falling edge of TIO, or if INV is low the counter value will be transferred on the rising edge of TIO. Thus, by inverting the INV bit after each transition, both the high periods and low periods of the pedal waveform may be measured consecutively.

At step 1804 the difference between the current timer value (the amount stored in the register) and the previous timer value is calculated. The timer value may overflow, and this must be taken into account in order to avoid incorrect negative values being used. At step 1805 a question is asked as to whether the TIO pin is now high. If it is, this indicates that the period just measured was the low cycle of the pedal waveform, corresponding to the vertical measurement. Thus, control is thereafter directed to step 1806 where the vertical position is calculated.

Otherwise it may be assumed that the horizontal cycle of the waveform has just been measured, and control is directed to step 1807, where the horizontal position is calculated. After step 1806 or 1807, control is directed to step 1808, where other 1.38kHz intenupt-driven tasks are performed.

It is possible that the foot pedal controller 151 has not been connected to the foot pedal processor 152. This condition may be detected by the fact that TIO does not change over more than two cycles. If this is the case, middle position values for the pedal may be used, rather than random amounts which are contained within the memory 1702 at power-up.

The steps for calculating the horizontal position 1807 or the vertical position 1806 from the timer difference value are detailed in Figure 19. At step 1901 the timer difference is converted by applying scale and offset constants and a linearlyinterpolated 128 step look-up table. The effect of the scaling and offset constants may be understood with reference back to Figure 15. The range of frequencies, and hence timer difference values, produced by each dimension is restricted to a certain range of the available spectrum. It is desirable to normalise this frequency range to a fractional range between zero and one, which is convenient for further processing in the DSP56002 processor 1701. Thus the offset constant for a particular dimension corresponds to the minimum timer difference value for that dimension, reducing it to zero. The scaling constant is then applied to expand the range to cover zero to one. The offset and scaling constants are different for each dimension, and, to simplify manufacture, may be different for each pedal, due to component tolerances and so on. A calibration cycle is used to determine these values once a pedal is installed with a particular foot pedal processor. This calibration cycle is extremely simple.

The look-up table applied in step 1901 defines a linearising curve for the dimension under conversion. Unlike offset and scaling constants, the shape of the curve may be considered as being invariant for a particular design of product, thereby simplifying the type of calibration which must be performed by the purchaser of a pedal. Process 1901 results in a 24 bit fractional value being calculated, representing a highly accurate measurement of the angle of the pedal in whichever of the two dimensions is being converted. However this accuracy is not as high as the 24 bits of resolution of the processor. Thus, it is necessary to perform a type of filtering in order to avoid jittering of numerical values.

In step 1902 a question is asked as to whether the pedal has changed direction. In other words, in the dimension currently under consideration, has the direction of any movement actually changed. If the direction has not changed, control is directed to step 1905 where the pedal position is updated. This is safe to do, as effectively a value is stored which is the position plus maximum jitter or minus minimum jitter. Altemativeiy, if the direction has changed, control is directed to step 1903, where a question is asked as to whether the position has changed by greater than the backlash constant. The backlash constant is selected as being just greater than the amount of jitter. Thus a position in a new direction will not be stored until the new direction is known to be the result of a deliberate movement, and not just jitter. Thus, if the pedal movement is sufficient, the pedal direction is updated in step 1904, for that dimension, and the position updated in step 1905.

Altematively, if the direction change is due to a change of value less than the backlash constant, this value is ignored, and control is directed back to step 1808 in Figure 18. This filtering technique is known as a backlash algorithm.

The calibration cycle for determining the offset and scaling constants for each dimension is detailed in Figure 20A. In step 2001, the user is requested, via the display panel of the user interface 1705, to position the pedal away from extremes in either dimension. The user indicates when this has been done, and at step 2002 a measurement is taken of the pedal at this position, resulting in raw timing values for X and Y dimensions (horizontal and vertical dimensions respectively). Minimum and maximum values for each dimension are set as being equal to this initial measured time. At step 2003 the user is requested to move the pedal to its extremes. In other words, to the left and right extremes horizontally, and the up and down extremes vertically. At step 2004 minimum and maximum values are cumulatively recorded. At step 2005 control is directed back to step 2004 until the user signifies that the pedal has been moved to its extremes, and that calibration is complete.

At step 2006 the minimum and maximum timer values for each dimension are used to determine the offset and scaling constants which are used during normal operation, in step 1901 shown in Figure 19. The four constants are called X~SCALE, X~OFFSET, Y~SCALE and Y~OFFSET. These are stored in nonvolatile memory in step 2007.

The calculations for determining the offset and scale constants performed in step 2006 are detailed in Figure 20B. It is known that a small amount of oscillation drift will occur due to ambient temperature affecting the capacitors and coil sensors of the foot pedal controller circuit shown in Figure 14. Generally it will be preferable to ensure that maximum and minimum values are obtainable from the pedal at the extremes of position, rather than find that temperature drift has made the top or bottom few percent of values inaccessible. Thus an inactive border is defined. In Figure 20 the value BORDER is a fraction, having the value .025, corresponding to inactive borders of 2.5% in each dimension. Thus X~OFFSET, X~SCALE, Y~OFFSET and Y~SCALE are obtained from equations containing this value. When applied to timer measurements in step 1901, pedal values in the inactive border regions will go negative or higher than unity. These values are limited to zero and unity respectively, before being supplied as addresses to the linearly interpolated look-up table.

The look-up tables applied in step 1901 in Figure 19 are represented graphically in Figure 21. Normalised timer values are supplied as addresses, and an angle scaled to the range of zero to unity is the data result. The established technique of linear interpolation is used to provide a high accuracy from a small look-up table of 128 steps.

The accuracy of conversions achieved by the processes outlined above is very high. A linear resolution of 4096 steps has been obtained in the present embodiment. The limits of resolution are determined by the clock frequency of the timer in the digital signal processor 1701. In this case the clock frequency is twenty megahertz. Another restriction is oscillator stability. It has been found that 50Hz mains hum can interfere with the oscillators, and this has limited the present embodiment to the resolution of 4096 steps. However, in an altemative embodiment, the coils and connections are screened using thin coatings of conductive paint, so as not to interfere with the resonant frequency. It is anticipated that the resolution can be significantly increased by this method.

Competing sensor technologies, such as potentiometers or Hall effect transducers are unable to match this resolution, are more expensive and are unreliable.

The sample rate for each dimension is about 250 Hz. This is considered to be adequate for a foot pedal. This rate could be increased without loss of resolution, if the timer clock rate is increased proportionately. Altematively, a trade off between resolution and sample rate may be chosen.

The flow of information in the digital signal processor 1701 shown in Figure 17 is shown in Figure 22. The processes of pulse width conversion 2201, as summarised in Figures 18 and 19, supply numerical values for pedal positions in both dimensions to digital signal processor (DSP) functions 2202, and midi interpreter, controller and merger functions 2203. In some instances it will be useful to insert midi pedal values into a midi stream from another source, such as the synthesiser 105. In order to do this, the incoming midi codes must be interpreted to a degree, so that midi pedal commands may be inserted between incoming midi commands, and not corrupt midi data. The numerical values supplied from the pulse width conversion process 2201 to the digital signal processing functions 2202 enable the foot pedal to be used to control various characteristics of sound, depending on the DSP function which has been selected.

An example of a DSP function controlled by the pedal is shown in symbolic form in Figure 23. A state variable filter 2301 has controls for frequency 2302 and resonance 2303. The frequency control 2302 defines a cut-off frequency, above which audio frequencies are attenuated by 12dB per octave. The resonance control 2303 defines the level to which frequencies near the cut-off frequency are amplified. This type of filter is used widely in sound processing, and particularly in sound synthesis. In the embodiment, the vertical position of the foot pedal may be used to control the cut-off frequency and the horizontal position of the foot pedal to control the resonance. Thus analogue audio signals from any source in the studio environment may be processed in this way.

In Figure 24 a different type of DSP function is shown. A surround sound processor 2401 includes controls for front and back 2402, and left and right 2403.

The surround sound processor is able to encode a stereo signal in such a way that, when decoded with compatible surround sound equipment connected to three or five loudspeakers, sounds may be located anywhere in space. The vertical foot pedal position may be used to control the front to back sound location, and the horizontal pedal position used to control the left to right sound location.

In an altemative embodiment, audio processing is excluded, and a processor is contained within the foot pedal controller 151. Midi connections are provided on the foot pedal base, with controls and an interface for selecting various operational configurations, such as midi channel number and so on. Due to the low power consumption of the pedal circuit shown in Figure 14, it is possible to operate the pedal without a power supply, power being derived from a midi signal supplied to the midi input on the pedal base.

In a further alternative embodiment, foot pedal positions are communicated using the XMIDI protocol. This protocol is back compatible with the widely established midi standard, and inserts additional information into the eight bit midi word by means of a third, reversed logic state where the midi signal state would be perceived as a zero. The XMIDI protocol provides increased bandwidth and precision by adopting temary rather than binary signalling, and is therefore suitable for efficiently handling the high precision measurements made by the circuitry and transducers of the invention.

In an additional further embodiment the multiplexing arrangement shown in Figure 14 is extended to selectively control a larger number of tuned oscillators.

The coil in each such circuit is arranged to detect the displacement of a foot switch. A microcontroller selects an oscillator, and measures the period until the next pulse is generated by the circuit, before selecting the next oscillator.

The displacement of a switch is indicated by this period. When such an arrangement is used for on-off switches, a reduced level of resolution may be tolerated. For example, a resolution of sixteen steps is sufficient to provide hysteresis points at one third and two thirds the distance of travel of the foot switch. When eight such switches are multiplexed, the cost of the circuit is competitive with electrical contact switches, while providing the inherent advantages of expensive contactless switches, such as those which use hall effect sensing techniques.

In the switching embodiment, it may be assumed that all switches are in the off position at power-up. Thus a degree automatic calibration may be performed at this time. The minimum frequencies for each oscillator may be determined thereby defining an offset. An initial frequency range may be assume, which may be extended as a result of measurements made during use, thereby defining a scale factor. Thus a scaling factor and an offset may be determined for each switch during normal use, thereby avoiding the need for a manual calibration procedure.

Claims (82)

1. Apparatus for controlling audio signal processing means configurable to modify a characteristic of an audio signal in response to a manual operation applied to said controlling apparatus, wherein said controlling apparatus includes a resonance circuit including inducting means having an inductance adjustable in response to said manual operation, such that the resonant frequency of said resonance circuit is adjustable in response to said manual operation.

2. Apparatus according to claim 1, wherein said inducting means comprises a coil and a co-operating member, wherein said inductance is varied by relative movement between said coil and said co-operating member.

3. Apparatus according to claim 2, wherein the shape of said coil is defined to produce substantially linear changes in the resonant frequency of said resonance circuit in proportion to said manual operation.

4. Apparatus according to claim 2, wherein the shape of said cooperating member is defined to produce substantially linear changes in the resonant frequency of said resonance circuit in proportion to said manual operation.

5. Apparatus according to any of claims 1 to 4, further comprising storage means and processing means, wherein said storage means stores values associated with a range of frequencies generated by said resonance circuit in response to manual operations; said processing means is arranged to process said stored values with a frequency value generated in response to a frequency received from said resonance circuit; and said processing results in the generation of a value which represents the position of said controlling apparatus with respect to a normalised range of position values.

6. Apparatus according to claim 5, wherein said stored values include an off-set value and a scaling value.

7. Apparatus according to claim 6, wherein said off-set value is determined in response to the resonant frequency of said resonance circuit when said controlling apparatus is at a first known position.

8. Apparatus according to claim 7, wherein said scaling value is determined in response to the reciprocal of the range of resonant frequencies of said resonance circuit for positions of said controlling apparatus between said first known position and a second known position.

9. Apparatus according to any of claims 6 to 8, wherein the range of resonant frequencies is modified so as to define an active main region and an inactive border region for positions of said controlling apparatus.

10. Apparatus according to any of claims 5 to 9, including a look-up table, wherein said processing means is further arranged for addressing said look-up table in response to said processing of said stored values with said frequency value, such that data from said look-up table provides a substantially linear quantification of said manual operation.

11. Apparatus according to any of claims 1 to 10, wherein said resonance circuit includes a high gain inverter and a tuned circuit which includes said inductance means and capacitance means, arranged to generate a square wave at the resonant frequency of said tuned circuit.

12. Apparatus according to claim 11, wherein said inductance means is connected to an output of said inverter via a resistor and said capacitance means includes a first capacitor and a second capacitor; said first capacitor is connected from said input of said inverter to a fixed voltage; said second capacitor is connected from said fixed voltage to the junction of said inductance means with said resistor; and said capacitors are considered as being arranged in series thereby substantially forming the capacitance in a parallel tuned circuit which includes said inductance means.

13. Apparatus according to claim 11 or claim 12, wherein said inverter is an HCMOS inverting buffer.

14. Apparatus according to any of claims 1 to 13, wherein said resonance circuit resonates at radio frequencies.

15. Apparatus according to any of claims 1 to 14, wherein variations in resonant frequency are conveyed to a frequency measuring means by a process of pulse width modulation.

16. Apparatus according to claim 15, wherein pulse widths are determined by counting the frequency of resonance.

17. Apparatus according to any of claims 1 to 16, including a plurality of variable inductance resonance circuits.

18. Apparatus according to claim 17, wherein said plurality of resonance circuits are individually selectable for activation in response to a multiplex control signal.

19. Apparatus according to claim 18 when dependant on claim 11, wherein a resonance circuit is activated in response to a signal supplied to an additional input of said inverter.

20. Apparatus according to claim 18 when dependant on claim 15 or when dependent on claim 16, wherein pulse widths are multiplexed in time to convey respective frequencies of said resonance circuits.

21. Apparatus according to any of claims 1 to 20, configured to operate with a user's foot, and said manual operation consists of movement of said foot.

22. Apparatus according to claim 21, when dependant on claim 17, wherein said controlling apparatus has a plurality of modes of freedom and different audio signal characteristics are modifiable in response to different foot movements.

23. Apparatus according to claim 22, having two modes of freedom.

24. Apparatus according to claim 23, configured to rock back and forth about a substantially horizontal axis and configured to rotate about a substantially vertical axis.

25. Apparatus according to claim 24, wherein said axes are offset, with said horizontal axis located in-front of said vertical axis, so as to facilitate preferred modes of freedom of said foot.

26. Apparatus according to claim 24 when dependent on claim 15, including means for counting the two resonant frequencies altemately to generate high and low pulse widths on a shared multiplex, wherein said high and low pulse widths respectively convey frequencies of respective resonance circuits.

27. Apparatus according to claim 26 when dependent on claim 18, wherein said altemate high and low pulse widths are supplied as said multiplex control signal to selectively activate said resonance circuits.

28. Apparatus according to any of claims 1 to 27, wherein said audio signal processing means processes musical tones.

29. Apparatus according to any of claims 1 to 27, wherein said audio signal processing means modifies a characteristic of a musical tone by supplying parameters to a DSP algorithm.

30 Apparatus according to claim 29 when dependent on claim 17, wherein said DSP algorithm has a plurality of simultaneously controllable audio characteristics controllable via a plurality of respective parameters derived from manual operation of said controlling apparatus.

31. Apparatus according to claim 30, wherein said DSP algorithm is a resonating filter and two of said respecting parameters control the pitch and resonance of said filter.

32. Apparatus according to claim 30, wherein said DSP algorithm is a surround sound encoding algorithm and two of said respective parameters control the left-right and the front-back location of a sound source.

33. Apparatus according to claim 30, wherein said DSP algorithm is a virtual acoustic synthesis algorithm, wherein two of said parameters respectively control aspects of playing style on a virtual acoustic instrument.

34. Apparatus according to any of claims 1 to 33, wherein said modifications are recorded in a digital sequencer.

35. Apparatus according to any of claims 1 to 20, having selection means arranged to modify said adjustable inductance in said resonance circuit.

36. Apparatus according to claim 35 when dependent on claim 17, wherein said controlling apparatus has a plurality of selection means, wherein each of said selection means has a respective resonance circuit such that the application of pressure results in movement of said selection means, thereby producing a change in the resonant frequency of the respective resonant circuit.

37. Apparatus according to claim 36, wherein each of said selection means has sufficient resolution to provide two hysteresis points.

38. Apparatus according to any of claims 1 to 37, wherein signals derived from said manual operation are supplied to said audio signal processing means in accordance with MIDI or XMIDI protocols.

39. Apparatus according to any of claims 1 to 38, wherein signals derived from said manual operation are supplied to said audio signal processing means by communication between computation processors on a common silicon chip.

40. Apparatus according to claim 26, wherein said high and low pulse widths are conveyed to a remote processing means.

41. A method of modifying a characteristic of an audio signal in response to manual operation, comprising steps of connecting a control device to an audio signal processing means, wherein said control device includes a resonance circuit having an inducting means adjustable in response to said manual operation, such that the resonant frequency of said resonance circuit is adjustable in response to said manual operation.

42. A method according to claim 41, wherein said inducting means comprises a coil and a co-operating member, wherein said inductance is varied by relative movement between said coil and said co-operating member.

43. A method according to claim 42, wherein the shape of said coil is defined to produce substantially linear changes in the resonant frequency of said resonance circuit in proportion to said manual operation.

44. A method according to claim 42, wherein the shape of said cooperating member is defined to produce substantially linear changes in the resonant frequency of said resonance circuit in proportion to said manual operation.

45. A method according to any of claims 41 to 44, further comprising steps of storing values associated with a range of frequencies generated by said resonance circuit in storage means; processing said stored values with a frequency value generated in response to a frequency received from said resonance circuit, wherein said processing results in the generation of a value which represents the position of said controlling apparatus with respect to a normalised range of position values.

46. A method according to claim 45, wherein said stored values include an off-set value and a scaling value.

47. A method according to claim 46, wherein said off-set value is determined in response to the resonant frequency of said resonance circuit when said controlling apparatus is at a first known position.

48. A method according to claim 47, wherein said scaling value is determined in response to the reciprocal of the range of resonant frequencies of said resonance circuit for positions of said controlling device between said first known position and said second known position.

49. A method according to any of claims 46 to 48, wherein the range of resonant frequencies is modified so as to define an active main region and an inactive border region for positions of said control device.

50. A method according to any of claims 45 to 49, wherein said processing means is arranged to address a look-up table in response to said processing of said stored values with said frequency value, such that data from said look-up table provides a substantially linear quantification of said manual operation.

51. A method according to any of claims 41 to 50, wherein said resonance circuit includes a high gain inverter and a tuned circuit arranged to generate a square wave at the resonant frequency of said tuned circuit.

52. A method according to claim 51, wherein the inductance is connected to an output of said inverter via a resistance and said capacitance includes a first capacitor and a second capacitor, said first capacitor is connected from said input of said inverter to a fixed voltage; said second capacitor is connected from said fixed voltage to the junction of said inductance with said resistance; and said capacitors are considered as being arranged in series thereby substantially forming the capacitance in a parallel tuned circuit which includes said inductance.

53. A method according to claim 51 or claim 52, wherein said inverter is an HCMOS inverting buffer.

54. A method according to any of claims 41 to 53, wherein said resonance circuit resonates at radio frequencies.

55. A method according to any of claims 41 to 54, wherein variations in resonant frequency are conveyed to a frequency measuring means by a process of pulse width modulation.

56. A method according to claim 55, wherein pulse widths are determined by counting the frequency of resonance.

57. A method according to any of claims 41 to 56, including a plurality of variable inductance resonance circuits.

58. A method according to claim 57, wherein said plurality of resonance circuits are individually selectable for activation in response to a multiplex control signal.

59. A method according to claim 58 when dependent upon claim 51, wherein a resonance circuit is activated in response to a signal supplied to an additional input of said inverter.

60. A method according to claim 58 when dependent upon claim 55 or when dependent upon 56, wherein pulse widths are multiplexed in time to convey respective frequencies of said resonance circuits.

61. A method according to any of claims 41 to 60, wherein manual operations are effected by operation of a user's foot.

62. A method according to claim 61, when dependent upon claim 57, wherein said controlling device has a plurality of modes of freedom and different audio signal characteristics are modifiable in response to different foot movements.

63. A method according to claim 62, having two modes of freedom.

64. A method according to claim 63, wherein said controlling device is configured to rock back and forth about a substantially horizontal axis and configured to rotate about a substantially vertical axis.

65. A method according to claim 64, wherein said axes are off-set, with said horizontal axis located in-front of said vertical axis, so as to facilitate preferred modes of freedom of said foot.

66. A method according to claim 64 when dependent upon claim 55, including means of counting the two resonant frequencies altemately to generate high and low pulse widths on a shared multiplex, wherein said high and low pulse widths respectively convey frequencies of respective resonance circuits.

67. A method according to claim 66 when dependent upon claim 58, wherein said alternate high and low pulse widths are supplied as said multiplex control signal to selectively activate said resonance circuits.

68. A method according to any of claims 41 to 67, wherein said audio signal processing means processes musical tones.

69. A method according to any of claims 41 to 67, wherein said audio signal processing means modifies a characteristic of a musical tone by supplying parameters to a DSP algorithm.

70. A method according to claim 69 when dependent upon claim 57, wherein said DSP algorithm has a plurality of simultaneously controllable audio characteristics controllable via a plurality of respective parameters derived from manual operation of said controlling apparatus.

71. A method according to claim 70, wherein said DSP algorithm is a resonating filter and two of said respective parameters control the pitch and resonance of said filter.

72. A method according to claim 70, wherein said DSP algorithm is a surround sound encoding algorithm and two of said respective parameters control the left-right and the front-back location of a sound source.

73. A method according to claim 70, wherein said DSP algorithm is a virtual acoustic synthesis algorithm, wherein two of said parameters respectively control aspects of plain style on a virtual acoustic instrument.

74. A method according to any of claims 42 to 73, wherein said modifications are recorded in a digital sequencer.

75. A method according to any of claims 41 to 60, having selection means arranged to modify said adjustable inductance in said resonance circuit.

76. A method according to claim 75 when dependent on claim 57, wherein said controlling apparatus has a plurality of selection means, wherein each of said selection means has a respective resonance circuit such that the application of pressure results in movement of said selection means, thereby producing a change in the resonant frequency of the respective resonant circuit.

77. A method according to claim 76, wherein each of said selection means has sufficient resolution to provide two hysteresis points.

78. A method according to any of claims 41 to 77, wherein signals derived from said manual operation are supplied to said audio signal processing device in accordance with MIDI or XMIDI protocols.

79. A method according to any of claims 41 to 78, wherein signals derived from said manual operation are supplied to said audio signal processing device by communication between computation processors on a common silicon chip.

80. A method according to claim 66, wherein high and low pulse widths are conveyed to a remote processing device.

81. Apparatus for controlling audio signal processing means substantially as herein described with reference to the accompanying figures.

82. A method of modifying a characteristic of an audio signal substantially as herein described with reference to the accompanying figures.