WO1999019863A1

WO1999019863A1 - Method of assisting in tuning of a musical instrument and tuning aid therefor

Info

Publication number: WO1999019863A1
Application number: PCT/GB1998/002984
Authority: WO
Inventors: Robin Spence; Roger Abraham Bunce; Christopher Arie Thompson; David Stewart Palmer
Original assignee: The University Of Birmingham
Priority date: 1997-10-10
Filing date: 1998-10-06
Publication date: 1999-04-22
Also published as: GB9721438D0

Abstract

A method of assisting in the tuning of a musical instrument and a musical instrument tuning aid for carrying out the method. The method comprises the steps of playing a first note on a musical instrument (hereinafter called 'the first played note'), calculating the pitch (fundamental frequency) of the first played note and calculating the pitch of a note (hereinafter called 'the correct note'). For a defined scalic structure, the correct note is a note in a scale of half tones based on a reference note, and is the note in such a scale having a pitch closest to the pitch of the first played note. The method further includes the step of generating an audible output of said correct note.

Description

METHOD OF ASSISTING IN TUNING OF A MUSICAL INSTRUMENT AND TUNING AID THEREFOR

The present invention relates to a method for assisting in the tuning of a musical instrument and to a tuning aid for carrying out such a method.

Tuning musical instruments can be difficult for both students and experienced musicians, particularly when two or more types of instrument are involved, for example, when tuning a saxophone to a piano, or using a tuning fork or tuning whistle to tune a guitar. This is because of tonal differences between different instruments.

Electronic tuners are available which analyse the pitch of a note played on an instrument and indicate on a meter (see for example WO 90/03638) or via lights if the note is sharp or flat. However, they do not train the musicians to listen to and compare differences in musical pitch. This is particularly important for stringed, brass and woodwind instruments where the accuracy of pitch does not depend merely on the initial tuning but on how the musician plays each note. Clearly, tuning and pitch control skills are vital when playing with other musicians.

One way of acquiring these skills is for a teacher to play a note on one instrument, while a student listens and simultaneously plays a second note on a second instrument. This process is repeated until the second note is in tune with the first. Clearly, tuning to a similar instrument is easier because there are no major tonal differences. Furthermore, tuning may be effected by eliminating the beat frequency which occurs when two similar notes are played. This skill can take a long time to acquire simply because the teacher or other musician needs to be present. It is an object of the present invention to provide a method for assisting in the tuning of a musical instrument and a tuning aid for carrying out such a method, which allow musicians to tune their instruments accurately, and enable them to acquire skills in tuning and playing in tune with other instruments, without the need for other musicians to be present.

According to a first aspect of the present invention, there is provided a method of assisting in the tuning of a musical instrument, said method comprising the steps of:

(i) playing a first note on a musical instrument (hereinafter called "the first played note");

(ii) calculating the pitch (fundamental frequency) of the first played note;

(iii) calculating the pitch of a note (hereinafter called "the correct note") which, for a defined scalic structure, is a note in a scale of half tones based on a reference note, said correct note being the note in such a scale having a pitch closest to the pitch of the first played note; and

(iv) generating an audible output of said correct note.

In practice, depending on the skill with which the musical instrument is played, there may be variations in the pitch of the first played note, in which case, the pitch calculated in step (ii) is an average value. Similarly, there may be variations in the volume of the first played note, in which case, the audible output in step (iv) may be adjusted to be constant in volume.

Said reference note may be, for example, A = 440Hz (concert pitch). Preferably, however, the method includes an additional step of selecting the reference note prior to step (iii), so that any note may be chosen as the reference. The scalic structure referred to in step (iv) can be untempered (natural) or tempered (mean-tone or equal). Preferably, the method includes a step of choosing the temperament of the scalic structure.

For an equally tempered scale (usually used in Western orchestral music), the scale of half tones will be a scale of diatonic semitones, the interval between each semitone being 100 cents. In the case of a natural scale (often used, for example, by string quartets), the musical intervals between adjacent half tones are not equal and vary according to the key of the scale. Thus, if the scalic structure is untempered, the method preferably includes a step of selecting a key upon which to base the scale.

The tonal quality of the correct note generated in step (iv) may be the same as, or different to, that of the first played note. Preferably, the tonal quality of the correct note corresponds to a musical instrument.

The method may include the step of choosing the tonal quality of the correct note, thereby allowing a user to tune any musical instrument to itself, or to a predetermined selection of other musical instruments.

In one embodiment of the method, the correct note has the same tonal quality as the first played note and is produced by recording (preferably digitally) the first played note during step (i) and generating the audible output of the correct note in step (iv) by replaying the recording of the first played note at an altered speed, said altered speed being calculated such that at that speed, the pitch of the replayed note has the value calculated in step (iii).

Preferably, the method includes the step of visually indicating (graphically and/or alphanumerically) the difference in pitch between the first played note and the correct note.

The audible output of step (iv) may occur at a predetermined time after step (i), preferably said predetermined time being selected by the user.

Alternatively, step (iv) may be initiated by the playing of a second note on the musical instrument (hereinafter called "the second played note"). If the second played note is substantially a repeat of the first played note, a simultaneous aural comparison of the second played note and the correct note is enabled. A significant tuning benefit is obtained for those instruments (e.g. strings, woodwind and brass) which are capable of modulating the note played, since the user can modulate the second played note during the audible output of the correct note in step (iv), so that the second played note has the same pitch as the correct note (i.e. the note is being played in tune), thereby developing both hearing and playing skills.

Most preferably, said method comprises the steps of:

(i) selecting a reference note;

(ii) selecting a scalic structure (and a key if a natural scale is selected);

(iii) playing a first note on a musical instrument;

(iv) calculating the pitch of the first played note;

(v) calculating the pitch of a correct note which, for the selected scalic structure, is a note in a scale of half tones based on a reference note, said correct note being the note in such a scale having a pitch closest to the pitch of the first played note;

(vi) generating an audible output of said correct note;

(vii) visually indicating the difference in pitch between the first played note and the correct note; and (viii) attempting to reproduce the correct note by playing a second note on the musical instrument; wherein step (viii) is initiated during step (vi).

According to a second aspect of the present invention, there is provided a musical instrument tuning aid for performing the method of said first aspect, said musical instrument tuning aid comprising:-

(i) a microphone for receiving sound from said musical instrument;

(ii) frequency analysing means connected to said microphone and arranged to calculate the pitch of the first played note;

(iii) processing means adapted to calculate the pitch of the correct note; and

(iv) a tone generator connected to the processing means, and arranged to generate an audible output of the correct note.

Preferably, the tuning aid also comprises means for recording the first played note. The recording means may be activated manually by, for example, provision of a switch. Alternatively the recording means may be sound-activated, i.e. when the microphone picks up the first played note. In either case, means may be provided (e.g. a light) to indicate that the duration of the first played note is sufficient for correct functioning of the tuning aid. Most preferably, said recording means is a digital recording means.

Preferably, the frequency analysing means includes means for reducing harmonic content of the first played note, so as to enable a more accurate determination of the pitch of the first played note.

In a first embodiment, the tone generator is arranged to generate the audible output of the correct note by playing back the recorded first played note at a different speed, said speed being determined by playback control means. In said first embodiment, it will be understood that the correct note has the same tonal quality as the first played note (i.e. the correct note will have the same variations in volume and pitch as the first played note).

In said first embodiment, means may be provided to compensate for any fluctuations in pitch and volume of the first played note, such that the correct note has a constant pitch and volume, said constant pitch being an average of the pitch of the first played note.

In a second embodiment, a stored table of sound forms (wave table) of differing tonal qualities corresponding to a selection of musical instruments is provided and connected to the tone generator. In use, a user selects a sound form to be used by the tone generator, such that the audible output of the correct note has the tonal quality of the sound form (i.e. instrument) selected.

In a third embodiment, the features of said first and second embodiments are combined, thereby allowing any instrument to be tuned to itself and to instruments whose sound form is stored in the wave table.

Preferably, the tuning aid includes a visual display which is capable of indicating graphically or alphanumerically the difference in pitch between the first played note and the correct note (i.e. how many cents the first played note is sharp or flat). The absolute frequency of the second note and/or its name may also be displayed.

The frequency analysing means, the processing means and the tone generator may be (or include) a host computer operating under appropriate control software, the host computer being connected to the other components of the tuning aid. Alternatively, the components of the tuning aid may be incorporated into a single (preferably portable and more preferably hand held) device.

Embodiments of the invention will now be described with reference to the accompanying drawings in which:-

Fig 1 is a block diagram of a tuning aid in accordance with said second aspect of the present invention;

Fig 2 is a block diagram showing the tuning aid of Fig 1 in more detail;

Fig 3 is a circuit diagram of the sample rate timer shown in Fig 2;

Fig 4 is a flow diagram illustrating the steps performed by the host computer;

Fig 5 is a block diagram of the tuning aid of Example 2; and

Fig 6 is a block diagram showing the tuning aid of Figure 5 in more detail.

EXAMPLE 1: Tuning Aid using Host Computer

Referring to Figure 1 , the tuning aid comprises a microphone 10 and a microphone amplifier 12 connected to a sampling unit 14, a host computer 16 connected to the sampling unit 14 via a serial RS232 link 14a, and a tone generator comprising a power amplifier 18 and a loudspeaker 20 connected to the sampling unit 14. The sampling unit 14 has digital recording and playback capabilities and includes playback control means in the form of a sample rate timer 36 (Fig 2). The host computer 16 is programmed to enable it to analyse the frequency of digitised sound (samples) received from the sampling unit 14, and to process this information to determine the pitch of the correct note.

The tuning aid also comprises a 3dB/segment LED recording level bargraph 22 (Fig 2), an LED recording indicator 24 (Fig 2), power and record switches (not shown) for operation of the tuning aid by a user and a visual display for indicating properties of a played note (not shown).

In use, the user switches the tuning aid on and presses the record button. A note to be checked is played continuously (first played note) until the LED recording indicator 24 (Fig 2) extinguishes (after about 4 seconds). The bargraph display 22 showing the sound level helps the user to maintain a volume within the dynamic range of the microphone amplifier 12. In a simpler embodiment (not shown) the recording level is set automatically.

After a short pause, the name of the nearest semitone ("correct note") in a chromatic scale based on A = 440Hz and its octave position is displayed on the visual display. Also displayed is the error in pitch of the first note played by the user in cents sharp or flat. The user then replays the original note, triggering playback of the recorded note at a speed which changes the pitch of the recorded note to that of the correct note, as determined by the host computer 16 (described hereinafter). The amount of 'discomfort' or beat frequency heard between this second note played and the correct note is used as a guide to tuning the instrument. If during playback, the user plays a note for longer than 3 seconds, playback of the corrected note is retriggered. In alternative embodiments (not shown), the duration of replay of the corrected note may be determined by either (i) the duration of the original note, (ii) manual operation of a switch, or (iii) an electronic timing circuit.

Referring to Fig 2, the sampling unit 14 is shown in more detail, and has digital recording and playback functions. Components necessary for digital recording comprise a peak hold circuit 26, an anti-alias filter 28 and an analogue-to-digital converter (ADC) 30. Components necessary for playback comprise a digital-to-analogue converter (DAC) 32 and a reconstitution filter 34. The sample rate timer 36 (clocked by a crystal oscillator 44) controls the rate at which samples are both recorded and played back. The functions of the sampling unit 14 are coordinated by a microcontroller 37 operated by a control program stored in a read only memory (ROM) 38. Sample data is stored in a random access memory (RAM) 39. Transfer of data between the sampling unit 14 and the host computer 16 is effected via RS232 drivers 40.

In use, the note to be recorded is picked up by the microphone 10 and amplified by the amplifier 12. The amplifier output is monitored by the peak hold circuit 26 which drives the 3dB/segment LED bargraph display 22. The peak hold circuit 26 incorporates a comparator which switches when the input signal level to the amplifier 12 exceeds a preset threshold. The comparator output is monitored by the microcontroller 37 which initiates recording when a state change is detected. Thus, recording is sound activated, with a margin of safety against false triggering due to background noise. During the recording phase, the input signal is digitised via the anti-alias filter 28 and the ADC 30 under control of the microcontroller 37. After each ADC strobe, an internal 16 bit counter is incremented, and sampling is terminated when the counter overflows, thereby ensuring that RAM 39 does not overflow. The sampling rate is determined by the sample rate timer 36.

The data samples are stored sequentially in the RAM 39 and the microcontroller 37 is controlled by the control program executed from ROM 38. The ROM 38 and RAM 39 occupy the same address space and are decoded using the program memory enable output of the microcontroller 37. During the recording period, blocks of 4096 data samples are transmitted to the host computer 16, via the RS232 drivers 40. The host computer 16 calculates the pitch of the note and the correct note (described with reference to Fig 4).

In playback mode, the recorded samples are written to the DAC 32 at a rate determined by the sample rate timer circuit 36 (described with reference to Fig 3). The DAC output is then filtered by the reconstitution filter 34. The correct note is then played through the loudspeaker 20 which is driven via the power amplifier 18.

Referring to Fig 3, the sample rate timer 36 comprises a 12 bit, binary, self- reloading, presettable down counter 41,42,43 clocked by a crystal oscillator 44, and shift registers 46 and 47.

In use, the aim of the sampling rate timer 36 is to provide a highly accurate sampling interval for controlling the pitch of the replayed sound (correct note). For a given sampling rate, the appropriate number, provided by the microcontroller 37 on line 45, is latched into the counter 41 ,42,43 via the shift registers 46,47 and a line 49. The counter 41,42,43 is then enabled via line 48 and the underflow output is used to generate an interrupt in the microcontroller 37. The microcontroller 37 will then either read the ADC 30 in record mode or write to the DAC 32 in replay mode. The counter 41 ,42,43 is clocked at 25MHz providing a sampling interval resolution of < 40ns. In this embodiment, the frequency of the replayed sound can be controlled to within 2 cents. However, the resolution can be improved by increasing memory capacity. In record mode, the counter 41 ,42,43 is loaded with a default value which corresponds to the nominal sampling rate of 16.005kHz. In Fig. 4, a flow diagram representing the steps performed by the host computer 16 and its control program is shown. The functions of the host computer 16 can be divided into three main steps:-

1. Frequency analysis: determination of fundamental frequency of the first played note

The serial port of the host computer 16 is initialised to match the baud rate and protocol used by the sampling unit 14 (Block 51). A look up table of the frequencies of all the notes from bottom A (A1) to top C (C8) on the piano corresponding to concert pitch is generated. Each entry in the table is a structure comprising the note name, its frequency and its octave number. The values for a 512 point Bartlett window (see Rorabaugh, Digital Filter Designer's Handbook) are generated and stored in an array (Block 52).

The host computer 16 waits for data (recorded samples processed as described above by the sampling unit 14) to appear on the serial port. When activity is detected, the next 4096 bytes are read and stored in an array. The first 512 bytes are then copied into a separate array (Block 53) and then multiplied by the Bartlett window function to reduce spectral leakage caused by truncation of the input data, although other window functions such as a Hanning window function may be preferred. An array of values representing the spectral power density function of the audio waveform is obtained by performing a Fast Fourier Transform (FFT) on the windowed data so as to obtain the sampled frequency spectrum and particularly the frequency of the first peak. Such a 512 point FFT gives a frequency resolution of 31.35Hz, however a longer FFT would result in enhanced frequency resolution. An algorithm then establishes the approximate frequency of the fundamental component of the spectrum subject to a maximum error equal to the spectral resolution of the FFT (Block 54). The first significant peak is assumed to be the fundamental component, but a more sophisticated algorithm could be employed to reduce the possibility of errors at very low frequencies due to the close proximity of the spectral components.

Digital filter coefficients are then calculated for a filter cut-off frequency equal to the approximate fundamental frequency (Block 55). The Frequency Sampling Method (see Rorabaugh, Digital Filter Designer's Handbook) is used. The number of pass band samples is calculated and incremented by one to ensure that the pass band encompasses the fundamental component. A transition band sample value of 0.5 is included to reduce stop band ripple. Alternatively, full optimisation may be performed automatically to give more accurate coefficients. To accurately measure the fundamental frequency of the stored waveform, the harmonic content must first be substantially reduced. To achieve this the stored samples are filtered by a 251 tap Finite Impulse Response low pass digital filter (using the filter coefficients calculated above) (Block 56). The cut-off frequency of the filter is set to the approximate fundamental frequency obtained above, resulting in an array of samples of sinusoidal waveform substantially equal to the fundamental component of the original audio signal. The resultant output values are stored in a separate array.

To measure the fundamental frequency to the required accuracy, an averaging technique over several cycles is used (Block 57). Frequency F, accurate to within 0.06% (2 cents) is given by:

F = Fsamp*Fapprox/10*n

where Fsamp is the sampling frequency, Fapprox is the approximate frequency of the fundamental obtained from the FFT and n is the number of data samples in (Fapprox/10) complete cycles of the sinusoid.

2. Determination of correct note

The look up table of piano notes generated above (Block 52) is scanned and each value compared with the calculated fundamental frequency. The frequency of the semitone nearest to the fundamental frequency of the note played is retrieved from the table together with the note name and associated octave number which are displayed on the host computer screen (Block 58). The frequency error is displayed as a cent value sharp or flat. An analogue indicator which visually shows the error may also be incorporated.

3. Calculation of playback rate

To correct the pitch of the stored waveform, the samples are replayed at a rate different to the one at which they were originally recorded (Block 59). From the frequency error, the host computer 16 calculates the required replay rate and converts this value into a 12 bit binary word which is then loaded into a sample rate timer 36 via the RS232 serial link 14a. The value to be loaded into the sampling rate timer 36 is calculated from:

n = (Fsamp *(Fa-Fc)/Fc) + Fsamp

where n is the new timer value, Fsamp is the sampling frequency, Fc is the correct frequency and Fa is the fundamental frequency of the note played by the user.

It will be understood that the replayed (correct) note is merely a pitch shifted version of the original note. Therefore, any frequency variation in the original note will be reproduced. This has the advantage of the user tuning to what is in effect the sound of a real musical instrument, thereby helping the user to learn to produce a better quality tone. However, such variation may cause difficulty in tuning the instrument precisely, particularly at higher frequencies. If required, this can be overcome by averaging the frequency of the sampled sound over several time slices and then correcting the pitch of each slice before playback, although it will be understood that the correct note will not necessarily have the same tonal characteristics of the originally played note. Such averaging also provides a more accurate means of determining the frequency of the played note.

Other embodiments of the present invention (not shown) may comprise one or more of the following features:

1) A manual switch allowing the user to alter the reference pitch on which the look up table is based, or additional software allowing reference pitch alterations to be set on the tuning aid.

2) A manual switch (and/or additional software) allowing the user to select a particular scalic structure, for example equally tempered or natural. It should be noted that the frequency of the notes within the latter scalic structure depend upon the key of the piece of music being played. Hence, provision would also need to be made to allow the user to select the appropriate key.

3) An additional look up table, often referred to as a wave table, to allow playback of the sound of an instrument other than the instrument being tuned.

4) A manual switch (and/or additional software) allowing the replay of a complete musical scale, corrected in pitch as described. EXAMPLE 2: Stand-alone Tuning Aid

Referring to Fig 5, the stand-alone tuning aid is a handheld device comprising an external microphone 61 , a microphone amplifier 62, a digital sampling and processing unit 63 (described in more detail with reference to Fig 6), and a liquid crystal character display 64. The device is connected to an external, active loudspeaker 65.

In use, the user switches the device on and plays the note to be checked. This note must be held for at least one second. After a short period, the name of the nearest semitone on a chromatic scale and its octave position are displayed on the display 64. Also displayed is the pitch error in cents sharp or flat. If the user is still playing the original note when the display 64 is updated, then the device will replay a recording of the original sound at the corrected pitch. If the user has finished playing the original note before the display 64 is updated, then the user must play the note again to initiate replay. The beat frequency between the user's note and the replayed note can be used as a guide to tuning the instrument. This process may be repeated until correct tuning is achieved.

Figure 6 shows the sampling and processing unit 63 in more detail. The unit comprises a signal compander 71 to provide the device with a wide volume dynamic range. The unit also comprises an anti-alias filter 72, an 8-bit ADC 73, a high speed microcontroller 74, ROM 75 which contains the device control program, RAM 76 used to store data samples, an 8-bit DAC 77, and a low pass filter 78.

The functions of the sampling/processing unit can be divided into five main steps:- 1. Recording the first played note

After the device is switched on, the volume level is monitored by software stored in the ROM 75. On detecting a level above a preset threshold, a delay is initiated after which the level is again monitored. If the level is still above the threshold, then recording is activated. This helps prevent false triggering of the recording process.

After the recording process is activated, a delay is initiated to allow time for the attack phase of the note to end. Recording then begins and continues until 4096 samples have been stored. This takes about 0.5 seconds at the sampling rate of 8kHz.

The note to be checked is picked up by the microphone 61 , amplified by amplifier 62 and compressed by the compander 71. The signal is digitised by the ADC 73 after passing through the anti-alias filter 72 and the result read by the microcontroller 74. The sampling rate is controlled by a 16-bit counter (not shown) internal to the microcontroller 74. Another 16-bit counter internal to the microcontroller 74 is used to count the number of samples taken. The data samples are stored sequentially in the RAM 76.

2. Frequency analysis: determination of fundamental frequency of first played note

The method of frequency analysis is similar to that used in Example 1, except for the following differences:-

(i) The analysis is performed by the microcontroller 74 rather than the host computer 16;

(ii) A 256 point FFT is used rather than a 512 point FFT; and

(iii) The harmonic content is reduced using a digital Infinite Impulse

Filter (I IF) (not shown) with a 5^th order Butterworth low pass filter characteristic. 3. Determination of correct note

The method used is similar to that used in Example 1, except that the look up table of notes from A1 to C8 is stored permanently in ROM 75.

4. Calculation of playback rate

The method is similar to that used in Example 1 but with the following modification. The replay of the pitch corrected note involves looping a small section of the recorded sound which introduces an unwanted pitch shift into the signal. This is corrected by readjusting the sample replay rate so that there are an integer number of samples within a time equal to the fundamental period of the signal.

5. Playback

Having established the required playback rate, the sample rate timer is loaded with the appropriate value and the audio input level is again monitored. On detecting an input note, the pitch corrected note is replayed until the device is reset. The loop samples are written out sequentially to the DAC 77. The DAC output is filtered by the low pass filter 78 and expanded by the compander 71. The compander output is fed to the external speaker 65.

EXAMPLE 3 : Tuning Aid using Digital Signal Processing Integrated Circuit

In a third example (not shown) the tuning aid is in the form of a hand held instrument which is even smaller than that of Example 2. The tuning aid incorporates an internal loudspeaker and the functions of the host computer 16 are replaced by a digital signal processor (DSP). The DSP provides better performance such as much reduced computation time, better sound reproduction and smaller physical size of the device. However, miniaturisation limits the size of the loudspeaker, which in turn limits sound output at low frequencies. This can be particularly problematic when tuning instruments such as a string bass or baritone saxophone. The problem can be overcome by:- (i) providing a single earphone through which the correct note can be heard at the same time as the played note of the instrument through the other ear, or (ii) providing two ear phones which allow passage of ambient sound, or (iii) providing means to electronically mix the correct note and played note.

Claims

1. A method of assisting in the tuning of a musical instrument, said method comprising the steps of:

(i) playing a first note on a musical instrument, hereinafter called "the first played note";

(ii) calculating the pitch of the first played note;

(iii) calculating the pitch of a note hereinafter called "the correct note" which, for a defined scalic structure, is a note in a scale of half tones based on a reference note, said correct note being the note in such a scale having a pitch closest to the pitch of the first played note; and

(iv) generating an audible output of said correct note.

2. A method in accordance with Claim 1 , wherein the pitch calculated in step (ii) is an average value.

3. A method in accordance with Claim 1 or Claim 2, further including the step of selecting the reference note prior to step (iii).

4. A method in accordance with any preceding Claim, wherein the correct note is produced by recording the first played note during step (i) and generating the audible output of the correct note in step (iv) by replaying the recording of the first played note at an altered speed, said altered speed being calculated such that at that speed, the pitch of the replayed note has the value calculated in step (iii).

5. A method in accordance with any preceding Claim, wherein the audible output of step (iv) occurs at a selectable predetermined time after step (i).

6. A method in accordance with any one of Claims 1 to 4, wherein step (iv) is initiated by the playing of a second note on the musical instrument, hereinafter called "the second played note".

7. A musical instrument tuning aid for performing the method of any one of Claims 1 to 6, said musical instrument tuning aid comprising:-

(i) a microphone (10;61) for receiving sound from said musical instrument;

(ii) frequency analysing means (16;74) connected to said microphone

(10;61 ) and arranged to calculate the pitch of the first played note;

(iii) processing means (10;61) adapted to calculate the pitch of the correct note; and

(iv) a tone generator (18,20;65) connected to the processing means

(10;61), and arranged to generate an audible output of the correct note.

8. A tuning aid in accordance with Claim 7, wherein the tuning aid also comprises means (14;63) for recording the first played note.

9. A tuning aid in accordance with Claim 8, wherein the recording means (14;63) is sound-activated.

10. A tuning aid in accordance with any one of Claims 7 to 9, wherein the frequency analysing means (16,74) includes means for reducing harmonic content of the first played note.

1 1. A tuning aid in accordance with any one of Claims 7 to 10, wherein the tone generator (18,20;65) is arranged to generate the audible output of the correct note by playing back the recorded first played note at a different speed, said speed being determined by playback control means.

12. A tuning aid in accordance with Claim 1 1 , wherein compensation means are provided to compensate for any fluctuations in pitch and/or volume of the first played note, such that the correct note has a constant pitch and/or volume, said constant pitch being an average of the pitch of the first played note.

13. A tuning aid in accordance with any one of Claims 7 to 12, wherein a stored table of sound forms of differing tonal qualities corresponding to a selection of musical instruments is provided and connected to the tone generator.

14. A tuning aid in accordance with Claim 13 when appended to Claim 11, wherein selection means are provided for selecting whether the audible output is generated (i) by playing back the recorded first played note or (ii) from the stored table of sound forms.