A method of modeling (fig. 4, fig. 5) a characteristic of a non-linear system (21), comprises feeding test input signals (Sig6, Sig7) into the non-linear system (21) to obtain test output signals (Resp6, Resp7), wherein the test input signals (Sig6, Sig7) include a first test input signal (Sig6) and the test output signals (Resp6, Resp7) include a first test output signal (Resp6), identifying occurrences when an output level in at least one specific frequency band of the first testing output signal significantly changes under the first test input signal (e.g. identifying the frequencies at which there are overtones) so as to obtain a first profile (Profile B: frequency index versus input level, fig. 7), and modeling the characteristic (e.g. frequency response of pre-amplifier stage, fig. 8) based on the first profile (fig. 20).

Various test input signals may be used : Amplitude sweep (fig. 6a-c), frequency sweep (fig. 10), white noise or chirp signal. The breakup point of overtones onsets in the table of frequency bins versus input signal is determined.

The application of this profile to model the non-linear amplifier stage of a music instrument, for instance a guitar, is also defined.

A musical-performance-information transmission method using a first instrument and a second instrument, wherein the first instrument produces sounds in accordance with a user's musical performance and generates musical-performance data in accordance with the produced sounds and the second instrument produces sounds by receiving the musical performance data via a communication means. In the musical-performance-information transmission method, a mixed-sound signal is generated in accordance with a mixture of sounds produced by the first instrument and sounds that are different from the sounds produced by the first instrument, a reference signal is generated in accordance with the sounds produced by the first instrument, and on the basis of the mixed-sound signal and the reference signal, the reference signal is removed from the mixed-sound signal, generating a separated signal, and sound is emitted on the basis of said separated signal.

A voice analysis method comprises generating a time series of a relative pitch (R), which is a difference between a pitch (PB) generated from music track data (XB) designating respective notes of a music track in time series, and a pitch (PA) of a reference voice. The music track is divided into unit sections (UA) of a predetermined duration, and singing characteristics data (Z) is generated, which includes, for each of a plurality of statuses (St) of a model (M), classification information for classifying the unit sections (UA) into a plurality of sets and variable information defining a probability distribution of the time series of the relative pitch (R) within each of the classified unit sections (UA). The classification information is generated based on a condition relating to an attribute of the note and based on the condition relating to an attribute of the each of the unit sections (UA).

A voice synthesis method comprises generating a relative pitch transition (CR) based on synthesis-purpose music track data (YB) and singing characteristic data (Z). The singing characteristics data (Z) comprises a first singing characteristics data (Z1) including a first decision tree T1[n] and a second singing characteristics data (Z2) including a second decision tree T2[n]. The first singing characteristics data (Z1) and the second singing characteristics data (Z2) are mixed. The relative pitch transition (CR) is generated corresponding to the synthesis-purpose music track data (YB) and the mixed singing characteristics data based on a model (M). The first decision tree (T1[n]) and the second decision tree (T2[n]) differ in at least one of size, structure, and classification.

A sound signal analysis apparatus 10 includes sound signal input means adapted for inputting a sound signal indicative of a musical piece, tempo detection means adapted for detecting a tempo of each of sections of the musical piece by use of the input sound signal, judgment means adapted for judging stability of the tempo, and control means adapted for controlling a certain target in accordance with a result judged by the judgment means.

The waveform data generating apparatus has a waveform data generating circuit WP which inputs a digital signal formed of a plurality of bits which form a control signal for controlling an external apparatus, and generates waveform data indicative of a waveform of a control tone which corresponds to the input digital signal, is formed of tones corresponding to respective values of the bits of the input digital signal, and is formed of frequency components included in a certain high frequency band. The waveform data generating circuit WP has a basic waveform data extraction portion WP7 which extracts a part or a whole of the intermediate portion which is situated at an intermediate portion of the waveform data, and corresponds to the intermediate portion of the digital signal whose bit pattern coincides with a certain bit pattern as basic waveform data.

Waveform data stored in an external memory (125), like a NAND-type flash memory, are transferred from the external memory to an internal waveform memory (126) via a buffer and read out from the waveform memory to thereby reproduce a tone. Transfer instruction is generated each time data readout from the waveform memory progresses by one page, and the transfer instruction is registered into a transfer queue (521). Thus, in response to the transfer instruction from the queue, the waveform data stored in the external memory are read out on page by page and stored into the waveform memory. The external memory stores therein the waveform data in pages with error correction code attached per page. As the waveform data of a page are transferred from the external memory to the waveform memory, an error is detected using the error correction code, and, if the error is correctable, the error is corrected. If the error is uncorrectable, volume of a tone being generated is rapidly attenuated, or a warning is issued.