EP1671315A1 - Procede et dispositif pour caracteriser un signal audio - Google Patents

Procede et dispositif pour caracteriser un signal audio

Info

Publication number
EP1671315A1
EP1671315A1 EP05735854A EP05735854A EP1671315A1 EP 1671315 A1 EP1671315 A1 EP 1671315A1 EP 05735854 A EP05735854 A EP 05735854A EP 05735854 A EP05735854 A EP 05735854A EP 1671315 A1 EP1671315 A1 EP 1671315A1
Authority
EP
European Patent Office
Prior art keywords
sequence
sub
designed
period length
sequences
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Granted
Application number
EP05735854A
Other languages
German (de)
English (en)
Other versions
EP1671315B1 (fr
Inventor
Markus Cremer
Christian Uhle
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Fraunhofer Gesellschaft zur Forderung der Angewandten Forschung eV
Original Assignee
Fraunhofer Gesellschaft zur Forderung der Angewandten Forschung eV
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Fraunhofer Gesellschaft zur Forderung der Angewandten Forschung eV filed Critical Fraunhofer Gesellschaft zur Forderung der Angewandten Forschung eV
Publication of EP1671315A1 publication Critical patent/EP1671315A1/fr
Application granted granted Critical
Publication of EP1671315B1 publication Critical patent/EP1671315B1/fr
Ceased legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Classifications

    • G—PHYSICS
    • G10—MUSICAL INSTRUMENTS; ACOUSTICS
    • G10L—SPEECH ANALYSIS TECHNIQUES OR SPEECH SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING TECHNIQUES; SPEECH OR AUDIO CODING OR DECODING
    • G10L25/00—Speech or voice analysis techniques not restricted to a single one of groups G10L15/00 - G10L21/00
    • G10L25/48—Speech or voice analysis techniques not restricted to a single one of groups G10L15/00 - G10L21/00 specially adapted for particular use
    • G—PHYSICS
    • G10—MUSICAL INSTRUMENTS; ACOUSTICS
    • G10H—ELECTROPHONIC MUSICAL INSTRUMENTS; INSTRUMENTS IN WHICH THE TONES ARE GENERATED BY ELECTROMECHANICAL MEANS OR ELECTRONIC GENERATORS, OR IN WHICH THE TONES ARE SYNTHESISED FROM A DATA STORE
    • G10H2210/00—Aspects or methods of musical processing having intrinsic musical character, i.e. involving musical theory or musical parameters or relying on musical knowledge, as applied in electrophonic musical tools or instruments
    • G10H2210/031—Musical analysis, i.e. isolation, extraction or identification of musical elements or musical parameters from a raw acoustic signal or from an encoded audio signal
    • G10H2210/071—Musical analysis, i.e. isolation, extraction or identification of musical elements or musical parameters from a raw acoustic signal or from an encoded audio signal for rhythm pattern analysis or rhythm style recognition

Definitions

  • the present invention relates to the analysis of audio signals, and in particular to the analysis of audio signals for purposes of classifying and identifying audio signals to characterize the audio signals.
  • the extraction of fingerprints is particularly important when analyzing audio signals, ie signals that include music and / or speech.
  • the aim is also to “enrich” audio data with metadata in order to retrieve metadata for a piece of music, for example, on the basis of a fingerprint.
  • “Fingerprint” thus denotes a comp- generated from a music signal limited information signal, which does not contain the metadata, but is used for referencing the metadata, for example by searching a database, for example in a system for identifying audio material (“AudioID”).
  • Music data usually consists of superimposing partial signals from individual sources. While there is typically a relatively small number of individual sources in pop music, namely the singer, the guitar, the bass guitar, the drums and a keyboard, the number of sources for an orchestral piece can be very large.
  • An orchestral piece and a pop music piece for example, consist of an overlay of the tones emitted by the individual instruments.
  • An orchestral piece or any piece of music thus represents a superposition of partial signals from individual sources, the partial signals being the tones generated by the individual instruments of the orchestra or pop music ensemble, and the individual instruments being individual sources.
  • groups of original sources can also be understood as individual sources, so that at least two individual sources can be assigned to a signal.
  • An analysis of a general information signal is shown below using an orchestral signal as an example.
  • An orchestral signal can be analyzed in a number of ways. For example, there may be a desire to recognize the individual instruments and to extract the individual signals of the instruments from the overall signal and, if necessary, to convert them into a musical notation, the musical notation functioning as “metadata”. There are further possibilities for analysis in extracting a dominant rhythm, whereby rhythm extraction is better on the basis of the percussion instruments than on the basis of the more tone-giving instruments, which are also referred to as harmonic sustained instruments. While percussion instruments are typically timpani, Drums, rattles or other percussion instruments include, the harmoniously held instruments include all other instruments, such as violins, wind instruments, etc.
  • the percussion instruments also include all those acoustic or synthetic sound generators that contribute to the rhythm section due to their sound characteristics (e.g. rhythm guitar).
  • rhythm extraction of a piece of music it would be desirable to extract only percussive parts from the entire piece of music and then perform rhythm recognition on the basis of these percussive parts without the rhythm recognition being “disturbed” by signals from the harmonically sustained instruments.
  • melodic fragments in contrast to the rhythmic structure, mostly do not occur periodically. For this reason, many methods of searching for melodic fragments are limited to finding their occurrence individually. In contrast to this, in the area of rhythmic analysis, the focus is on finding periodic structures.
  • Methods for identifying melodic topics are only suitable to a limited extent for identifying periodicities present in a sound signal, since, as has been explained, musical themes are recurring, but they do not so much describe a basic periodicity in a piece of music, but rather if at all superordinate periodicity information. In any case, methods for identifying melodic themes are very complex, since the various variations of the themes have to be taken into account when searching for melodic themes. It is known from the music world that subjects are usually varied, for example through transposition, reflection, etc. The object of the present invention is to provide an efficient and reliable concept for characterizing an audio signal.
  • the present invention is based on the knowledge that an efficiently calculable characteristic of a sound signal that is meaningful in terms of a large amount of information can be determined as a characteristic on the basis of a sequence of times of use by determining the length of the period, dividing into sub-sequences and summarizing into a combined sequence.
  • a single sequence of times of use of a single instrument ie a single sound source along the time
  • at least two sequences of times of use of two different sound sources that occur in parallel in the piece of music are considered.
  • a common period length is determined using the sequences of operating times of the two sound sources, which is based on at least two sound sources. According to the invention, each sequence of times of use is then divided into respective sub-successes, for a length of a sequence is equal to the common period length.
  • the characteristic extraction then takes place on the basis of a combination of the sub-sequences for the first sound source into a first combined sub-sequence and on the basis of a combination of the sub-sequences for the second sound source in a second combined sub-sequence, the combined sub-sequences as a characteristic for serve the sound signal and can be used for further processing, such as for extracting semantically meaningful information about the entire piece of music, such as genre, tempo, time signature, similarity to other pieces of music etc.
  • the combined sub-sequence for the first sound source and the combined sub-sequence for the second sound source thus form a drum pattern of the sound signal if the two sound sources, which have been taken into account on the basis of the sequence of times of use, are percussive sound sources, such as drums, other percussion instruments. Instruments or any other percussive instruments that are characterized by the fact that it is not their pitch, i.e. their pitch, that decides, but that their characteristic spectrum or the rise and fall of an output tone and not the pitch are of greater musical importance.
  • the procedure according to the invention thus serves for the automatic extraction of preferably drum patterns from a preferably transcribed, that is to say e.g. B.
  • Notes representation of a music signal can be in MIDI format or can be determined automatically from an audio signal using methods of digital signal processing, such as, for example, with independent component analysis (ICA) or certain variations thereof, such as, for example, non-negative independent component analysis, or generally with concepts which are known under the keyword “blind source separation” (BSS).
  • ICA independent component analysis
  • BSS blind source separation
  • the extraction of a drum pattern is first carried out by recognizing the note inserts, that is to say starting times, for each different instrument and for each pitch in the case of tonal instruments.
  • a reading of a note can take place, which reading can consist in reading a MIDI file or in scanning and image processing of a notation or in accepting manually typed notes.
  • a grid is then determined, according to which the note insertion times are quantized, whereupon the note insertion times are then quantized.
  • the length of the drum pattern is then determined as the length of a musical bar, as an integral multiple of the length of a musical bar or as an integral multiple of the length of a musical counting time.
  • a pattern histogram is then used to determine a frequency of occurrence of a specific instrument per metric position.
  • the pattern histogram can be processed as such.
  • the pattern histogram is also a condensed representation of the musical events, ie the notation, and contains information about the degree of variation and preferred beats, a flatness of the histogram indicating a strong variation, while a very "mountainous" histogram indicates a more stationary signal in the sense of a self-similarity.
  • preprocessing In order to subdivide a signal into characteristic regions of the signal that are similar to one another and to extract a drum pattern only for regions in the signal that are similar to one another and another for other characteristic regions in the signal Determine drum pattern.
  • the present invention is advantageous in that a robust and efficient way of calculating a characteristic of a sound signal is obtained, in particular due to the subdivision carried out, which is very robust according to the period length which can also be determined with statistical methods and can be carried out equally for all signals is. Furthermore, the concept according to the invention is scalable in such a way that the meaningfulness and accuracy of the concept can easily be increased at the price of a higher computing time by the fact that more and more sequences of occurrences of more and more different sound sources, ie instruments, are included in the determination the common period length and be included in the determination of the drum pattern, so that the calculation of the combined sub-sequences becomes more and more complex.
  • an alternative scalability also consists in calculating a specific number of combined sub-sequences for a specific number of sound sources, in order then to rework the received sub-sequences depending on the further processing interest and thus to reduce their significance as required. Histogram entries below a certain threshold can e.g. B. be ignored. However, histogram entries can also be quantized per se or can only be binarized in general, depending on the threshold value decision, in such a way that a histogram only contains the statement that there is a histogram entry in the combined sub-sequence or not.
  • the concept according to the invention is a robust method due to the fact that many sub-sequences are "merged" into a combined sub-sequence, but can nevertheless be carried out efficiently since no numerically intensive processing steps are required.
  • percussive instruments without pitch which are also called drums in the following, play an essential role, especially in popular music.
  • Much information about rhythm and musical genre is contained in the "notes" played by drums, which could be used, for example, in an intelligent and intuitive search in music archives in order to be able to carry out classifications or at least preclassifications.
  • the notes played by drums often form recurring patterns, which are also called drum patterns.
  • a drum pattern can serve as a compressed representation of the played notes by extracting a note image the length of a drum pattern from a longer note image. This allows semantically meaningful information about the entire piece of music to be extracted from drum patterns, such as genre, tempo, time signature, similarity to other pieces of music, etc.
  • Figure 1 is a block diagram of an inventive device for characterizing a sound signal.
  • FIG. 3 shows a schematic diagram to illustrate a quantization grid and a quantization of the notes on the basis of the grid
  • 5 shows an exemplary pattern histogram as an example of combined sub-sequences for the individual sound sources (instruments); and 6 shows a post-processed pattern histogram as an example of an alternative characteristic of the audio signal.
  • FIG. 1 shows a device according to the invention for characterizing a sound signal.
  • FIG. 1 comprises a device 10 for providing a sequence of operating times for each sound source from at least two sound sources over time.
  • the times of use are preferably already quantized times of use which are present in a quantization grid.
  • FIG. 2 shows a sequence of times of use of notes from different sound sources, that is to say instruments 1, 2,..., N, which are designated by “x” in FIG. 2
  • FIG. 3 shows one in a grid that is in 3 shows a quantized sequence of quantized times of use for each sound source, that is to say for each instrument 1, 2,..., N.
  • FIG. 3 simultaneously represents a matrix or list of times of deployment, with a column in FIG. 3 corresponding to a distance between two grid points or grid lines and thus representing a time interval in which, depending on the sequence of times of use, a note is present or not.
  • a column in FIG. 3 corresponding to a distance between two grid points or grid lines and thus representing a time interval in which, depending on the sequence of times of use, a note is present or not.
  • the instrument n has no point in time in the time interval shown by reference numeral 30.
  • the multiple sequences of preferably quantized times of use are fed from the device 10 to a device 12 for determining a common period length.
  • the device 12 for determining a common period length is designed so that it does not itself determine its own period length for each sequence of times of use, but rather to find a common period length that most closely underlies the at least two sound sources. This is based on the fact that even if e.g. B. play several percussive instruments in one piece, all play more or less the same rhythm, so that there must be a common period length, to which practically all instruments that contribute to the sound signal, ie all sound sources, will adhere.
  • the common tone period length is then fed to a device 14 for dividing each sequence of times of use in order to obtain a set of sub-sequences for each sound source on the output side.
  • FIG. 4 it can be seen that a common period length 40 has been found, namely for all instruments 1, 2,..., N, the device 14 being designed for dividing into successions by all To divide sequences of times of use into sub-sequences of the length of the common period length 40.
  • the sequence of times of use for the instrument would then, as shown in FIG. 4, be divided into a first sub-sequence 41, a subsequent second sub-sequence 42 and a subsequent sub-sequence 43, in order for the example shown in FIG to get the sequence for instrument 1 three sub-sequences.
  • the other consequences for instruments 2, ..., n e- if necessary, divided into corresponding adjacent sub-sequences as has been shown on the basis of the sequence of times for instrument 1.
  • the sets of sub-sequences for the sound sources are then fed to a combining device 16 for each sound source in order to obtain a combined sub-sequence for the first sound source and a combined sub-sequence for the second sound source as a characteristic of the sound signal.
  • the summary preferably takes the form of a pattern histogram.
  • the sub-sequences for the first instrument are aligned one above the other in such a way that the first interval of each sub-sequence is to a certain extent “above” the first interval of each other sub-sequence. Then, as shown with reference to FIG.
  • the entries in The combined sub-sequence for the first sound source would therefore be a first line 50 of the pattern histogram in the example shown in Fig. 5.
  • the instrument 2 For the second sound source, that is to say, for each slot of a combined sequence or in each histogram bin of the pattern histogram
  • the instrument 2 would be the combined sub-sequence the second line 52 of the pattern histogram etc.
  • the pattern histogram in FIG. 5 thus represents the characteristic for the sound signal, which can then be used for various other purposes.
  • the pattern length can be found in various ways, namely, for example, from an a priori criterion, which directly estimates the periodicity / pattern length based on the previously existing note information provides, or alternatively z. B. by a preferably iterative search algorithm, which accepts a number of hypotheses for the pattern length and checks their plausibility on the basis of the results obtained. This can also be done, for example, again by evaluating a pattern histogram, as is preferably implemented by the device 16 for summarizing, or by using other self-similarity measures.
  • the pattern histogram as shown in FIG. 5 can be generated by the means 16 for summarizing.
  • the pattern histogram can also take the intensities of the individual notes into account in order to achieve a weighting of the notes according to their relevance.
  • the histogram may only contain information as to whether or not a tone is present in a sub-sequence or in a bin or time slot of a sub-sequence.
  • a weighting of the individual notes with regard to their relevance would not be included in the histogram.
  • the characteristic shown in FIG. 5, which here is preferably a pattern histogram is processed further.
  • a grade selection can be made on the basis of a criterion, for example by comparing the frequency or the combined intensity values with a threshold value.
  • This threshold can also depend, among other things, on the type of instrument or the flatness of the histogram.
  • the entries in drum patterns can be Boolean sizes, with a "1" for the fact before it would stand for a grade, while a "0" would stand for the fact that no grade would occur.
  • an entry in the histogram can also be a measure of how high the intensity (loudness) or relevance of the timeslot If note 6 is viewed, it can be seen that the threshold value was chosen such that all time slots or bins in the pattern histogram are marked with an “x” for each instrument where the number of entries is greater than or equal to 3. On the other hand, all bins in which the number of entries is less than 3, for example 2 or 1, are deleted.
  • a musical “result” or score is generated from percussive instruments that are not or not significantly characterized by a pitch.
  • a musical event is defined as the occurrence of a tone of a musical instrument.
  • the musical score or the characteristic preferably comprises the rhythmic information, such as start time and duration.
  • this metric information namely a time signature
  • An automatic transcription process can therefore be divided into two tasks, namely the recording and classification of the musical events, ie notes, and the generation of a musical score from the recorded notes, ie the drum pattern, as has already been explained above.
  • the metric structure of the music is preferably estimated, it also being possible to quantize the temporal positions of the recorded notes and to identify the starts and determine the position of the bar lines.
  • the recording and classification of the events is preferably carried out using the independent subspace analysis method.
  • ICA Independent Subspace Analysis
  • ISA Independent Subspace Analysis
  • the components are divided into independent subspaces or subspaces, the components of which do not have to be statistically independent.
  • Position of the mixed signal determined and the last assumption for the ICA complied.
  • Various methods for calculating the independent components have been developed in recent years. Relevant references, some of which deal with the analysis of audio signals, are as follows:
  • the recorded events are preferably aligned with the estimated tatum grid. This process corresponds approximately to the known quantization function in common MIDI sequencer software programs for music production.
  • the measure length is estimated from the quantized event list and recurring rhythmic structures are identified. Knowledge of the rhythmic structures is used to correct the estimated tempo and to identify the position of the bar lines using musical background knowledge.
  • the device 10 for providing sequences of times of use for a plurality of sound sources preferably carries out a quantization.
  • the recorded events are preferably quantized in the tatum grid.
  • the tatum grid is estimated using the note entry times of the recorded events together with note entry times that operate using conventional note entry methods.
  • the Generation of the tatum grid based on the percussive events recorded works reliably and robustly. It should be noted that the distance between two halftone dots in a piece of music is usually the fastest note played. If a piece of music therefore contains at most sixteenth notes and no faster than the sixteenth notes, the distance between two grid points of the Tatum Grid is equal to the length of time of a sixteenth note of the audio signal.
  • the distance between two grid points corresponds to the largest note value that is required to represent all occurring note values or time periods by forming integer multiples of this note value.
  • the grid spacing is the largest common divisor of all occurring note durations / period lengths etc.
  • the tatum grid is represented using a 2-way mismatch procedure (TWM).
  • TWM 2-way mismatch procedure
  • a series of test values for the tatum period i.e. for the distance between two grid points, is derived from a histogram for an inter-onset interval (IOI).
  • IOI inter-onset interval
  • the calculation of the IOI is not limited to successive onsets, but to practically all pairs of onsets in a time frame.
  • Tatum candidates are calculated as integer fractions of the most common IOI. The candidate is selected who best predicts the harmonic structure of the IOI according to the 2-way mismatch error function.
  • the estimated tatum period is subsequently calculated by calculating the error function between the comb-grid tu period is derived and the onset times of the signal are calculated.
  • the histogram of the IOI is thus generated and smoothed using an FIR low-pass filter. Tatum candidates are thus divided by dividing the IOI according to the peaks in the IOI histogram by a set of values between e.g. B. 1 and 4 received.
  • a raw estimate of the tatum period is derived from the IOI histogram after applying the TWM. Then the phase of the tatum grid and an exact estimate of the tatum period are calculated by means of the TWM between the note insertion times and several tatum grids with periods close to the previously estimated tatum period.
  • the second method refined and provides the Tatum grid by calculating the best match between the No ⁇ tencommunvektor and the Tatum grid represents, by using a coefficient of correlation R xy between the note entry vector x and the y Tatu.
  • the Tatum Grid is used for neighboring frames with e.g. B. estimated a length of 2.5 seconds.
  • the transitions between the tatum grids of neighboring frames are smoothed by low-pass filtering the IOI vector of the tatum grid points, and the tatum grid is restored from the smoothed IOI vector.
  • Each event is then assigned to its closest grid position. In a way, a quantization is carried out.
  • the intensity of the detected events can either be removed or used, which results in a Boolean matrix or in a matrix with intensity values.
  • the quantized representation of the percussive events provides valuable information for the assessment of the musical measure or a periodicity, which underlies the playing of the sound sources.
  • the periodicity at the clock level for example, is determined in two stages. First a periodicity is calculated in order to then estimate the cycle length.
  • ACF autocorrelation function
  • AMDF mean difference-of-magnitude function
  • the AMDF is also used to estimate the fundamental frequency for music and speech signals and to estimate the musical measure.
  • a periodicity function measures the similarity or dissimilarity between the signal and its temporally different version.
  • Various measures of similarity are known. For example, there is the Hamming distance (HD), which calculates a dissimilarity between two Boolean vectors Bi and B 2 according to the following equation.
  • the similarity measure M is obtained by summing the elements of B, as set out below.
  • MHD modified Hamming distance
  • the similarity measures for Boolean matrices can be expanded by weighting B with the mean of Ti and T 2 to take intensity values into account. Distances or dissimilarities are interpreted as negative similarities.
  • the time signature is determined by comparing P with a number of metric models.
  • the implemented metric models Q consist of a train of spikes with typical accent positions for different time signatures and micro times.
  • a micro time is the integer ratio between the duration of a musical beat, ie the note value that determines the musical tempo (e.g. quarter note), and the duration of a tatum period.
  • the best match between P and Q is obtained when the correlation coefficient reaches its maximum.
  • 13 metric models are implemented for seven different time signatures.
  • a score T from the length of a clock b is obtained by summing the matrix elements T with a similar metric position according to the following equation:
  • b denotes an estimated cycle length and p the number of cycles in T.
  • T ' is referred to as a score histogram or pattern histogram.
  • Drum patterns are obtained from the score histogram T 'by searching for score elements T' ⁇ , j with large histogram values. Patterns longer than one measure are retrieved by repeating the procedure described above for integer values of the measured length. The pattern length with the most hits, in relation to the pattern length itself, is selected in order to obtain a maximally representative pattern as a further or alternative characteristic for the sound signal.
  • the identified rhythmic patterns are interpreted using a set of rules derived from musical knowledge.
  • One example is the very common use of the snare drum or tambourines or "hand claps" in the second and fourth beats in a four-quarter time.
  • This concept serves as an indicator of the position of the bar lines. If a backbeat pattern is present, a bar starts between two stops of the small drum.
  • Another indication of the positioning of the tact lines is the occurrence of kick drum events, i.e. events of a large drum that is typically operated by a foot.
  • a preferred application of the characteristic as obtained by the means 16 for summarizing for each sound source, as shown and described in FIG. 1, as e.g. 5 or 6, is the genre classification of popular music.
  • Various high-level features can be derived from the drum patterns obtained to identify typical playing styles.
  • a classification procedure evaluates these features in connection with information about the musical measure, ie the speed, in z. B. beats per minute or beats per minute and using the percussive instruments used te.
  • the concept is based on the fact that all percussive instruments carry rhythm information and are often played repetitively.
  • Drum patterns have genre-specific characteristics. Therefore, these drum patterns can be used to classify the music genre.
  • a classification of different playing styles is carried out, each of which is assigned to individual instruments.
  • playing style is that events only occur on every quarter note.
  • An associated instrument for this style of play is the kick drum, i.e. the large drum of the drum that is operated with the foot.
  • This style of play is abbreviated to FS.
  • An alternative style of play is, for example, that events occur every second and fourth quarter notes of a four-quarter time. This is mainly played by the snare drum and tambourines, i.e. the hand claps.
  • This style of play is abbreviated as BS.
  • Exemplary other playing styles are that notes often appear on the first and third notes of a triplet. This is abbreviated to SP and is often seen in a hi-hat or cymbal.
  • the first characteristic FS is a boolean and true if kick drum events occur only on every quarter note. Only for certain values are no Boolean variables calculated, rather certain numbers are determined, such as for the relationship between the number of off-beat events and the number of on-beat events, such as those of a hi-hat, a shaker or a tambourine.
  • Typical combinations of drum instruments are classified in one of the various drum set types, such as rock, jazz, Latin, disco and techno, in order to obtain another characteristic for the genre classification.
  • the classification of the drum set is not derived using the instrument tones, but by generally examining the occurrence of drum instruments in various pieces belonging to the individual genres.
  • the rock drum set type is characterized, for example, by the fact that there is a kick drum, a snare drum, a hi-hat and a cymbal.
  • the "Latin" type there is a bongo, a conga, claves and shakers.
  • rhythmic features of the drum score or drum pattern are derived from the rhythmic features of the drum score or drum pattern. These features include musical tempo, time signature, micro time, etc.
  • a measure of the variation in the occurrence of kick drum notes is obtained by counting the number of different IOIs that occur in the drum pattern.
  • Classification of the musical genre using the drum pattern is carried out using a rule-based decision network. Potential genre candidates are rewarded if they meet a hypothesis currently under investigation and are "punished” if they fail to meet aspects of a currently under-hypothesis. This process results in the selection of favorable combinations of features for each genre.
  • the genre Disco is recognized when a drum set type is Disco, when the tempo is in the range between 115 and 132 bpm, when a time signature is 4/4 bits and the micro time is 2.
  • Another feature of the genre disco is that a game style FS z. B. is present, and that z. B. there is yet another style of play, namely that events occur at every off-beat position. Similar criteria can be set for other genres, such as hip-hop, soul / funk, drum and bass, jazz / swing, rock / pop, heavy metal, Latin, waltz, polka / punk or techno.
  • the method according to the invention for characterizing an audio signal can be implemented in hardware or in software.
  • the implementation can take place on a digital storage medium, in particular a floppy disk or CD with electronically readable control signals, which can cooperate with a programmable computer system in such a way that the method is carried out.
  • the invention thus also consists in a computer program product with a program code stored on a machine-readable carrier for carrying out the method when the computer program product runs on a computer.
  • the invention can thus be implemented as a computer program with a program code for carrying out the method if the computer program runs on a computer.

Landscapes

  • Engineering & Computer Science (AREA)
  • Computational Linguistics (AREA)
  • Signal Processing (AREA)
  • Health & Medical Sciences (AREA)
  • Audiology, Speech & Language Pathology (AREA)
  • Human Computer Interaction (AREA)
  • Physics & Mathematics (AREA)
  • Acoustics & Sound (AREA)
  • Multimedia (AREA)
  • Auxiliary Devices For Music (AREA)

Abstract

Afin de caractériser un signal audio, on prépare une séquence de points d'application quantifiés dans le temps pour chaque source audio parmi au moins deux sources audio, sur la base d'une grille de quantification (1). On détermine ensuite une longueur périodique commune aux deux sources audio, en utilisant les séquences de points d'application dans le temps (12). On subdivise ensuite la séquence de points d'application dans le temps en sous-séquences correspondantes (14), la longueur d'une sous-séquence étant égale à la longueur périodique commune. Finalement, on combine les sous-séquences d'une première source audio en une première sous-séquence combinée et on combine les sous-séquences de la deuxième source audio en une deuxième sous-séquence combinée (16), en utilisant, par exemple, un histogramme à motifs, afin de caractériser le signal audio, par exemple son rythme, sa vitesse ou son genre, sur la base de la première sous-séquence combinée et de la deuxième sous-séquence combinée.
EP05735854A 2004-05-07 2005-04-27 Procede et dispositif pour caracteriser un signal audio Ceased EP1671315B1 (fr)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
DE200410022659 DE102004022659B3 (de) 2004-05-07 2004-05-07 Vorrichtung zum Charakterisieren eines Tonsignals
PCT/EP2005/004517 WO2005114650A1 (fr) 2004-05-07 2005-04-27 Procede et dispositif pour caracteriser un signal audio

Publications (2)

Publication Number Publication Date
EP1671315A1 true EP1671315A1 (fr) 2006-06-21
EP1671315B1 EP1671315B1 (fr) 2007-05-02

Family

ID=34965834

Family Applications (1)

Application Number Title Priority Date Filing Date
EP05735854A Ceased EP1671315B1 (fr) 2004-05-07 2005-04-27 Procede et dispositif pour caracteriser un signal audio

Country Status (4)

Country Link
EP (1) EP1671315B1 (fr)
JP (1) JP4926044B2 (fr)
DE (2) DE102004022659B3 (fr)
WO (1) WO2005114650A1 (fr)

Families Citing this family (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2019026236A1 (fr) * 2017-08-03 2019-02-07 Ｐｉｏｎｅｅｒ ＤＪ株式会社 Dispositif d'analyse de composition musicale et programme d'analyse de composition musicale
JP6920445B2 (ja) * 2017-08-29 2021-08-18 ＡｌｐｈａＴｈｅｔａ株式会社 楽曲解析装置および楽曲解析プログラム
CN108257588B (zh) * 2018-01-22 2022-03-01 姜峰 一种谱曲方法及装置

Family Cites Families (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US6201176B1 (en) * 1998-05-07 2001-03-13 Canon Kabushiki Kaisha System and method for querying a music database
US6990453B2 (en) * 2000-07-31 2006-01-24 Landmark Digital Services Llc System and methods for recognizing sound and music signals in high noise and distortion
DE10157454B4 (de) * 2001-11-23 2005-07-07 Fraunhofer-Gesellschaft zur Förderung der angewandten Forschung e.V. Verfahren und Vorrichtung zum Erzeugen einer Kennung für ein Audiosignal, Verfahren und Vorrichtung zum Aufbauen einer Instrumentendatenbank und Verfahren und Vorrichtung zum Bestimmen der Art eines Instruments
JP2004029274A (ja) * 2002-06-25 2004-01-29 Fuji Xerox Co Ltd 信号パターン評価装置、信号パターン評価方法及び信号パターン評価プログラム

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
See references of WO2005114650A1 *

Also Published As

Publication number Publication date
EP1671315B1 (fr) 2007-05-02
JP4926044B2 (ja) 2012-05-09
JP2007536586A (ja) 2007-12-13
WO2005114650A1 (fr) 2005-12-01
DE102004022659B3 (de) 2005-10-13
DE502005000658D1 (de) 2007-06-14

Similar Documents

Publication Publication Date Title
US7273978B2 (en) Device and method for characterizing a tone signal
Mitrović et al. Features for content-based audio retrieval
EP1371055B1 (fr) Dispositif pour l'analyse d'un signal audio concernant des informations de rythme de ce signal a l'aide d'une fonction d'auto-correlation
EP2351017B1 (fr) Procédé permettant de détecter des motifs de notes dans des pièces musicales
EP1523719A2 (fr) Systeme et procede pour caracteriser un signal d'information
DE10123366C1 (de) Vorrichtung zum Analysieren eines Audiosignals hinsichtlich von Rhythmusinformationen
DE60315880T2 (de) Datenerzeugungseinrichtung und verfahren für musikkompositionen
WO2002084641A1 (fr) Procede pour convertir un signal musical en une description fondee sur des notes et pour referencer un signal musical dans une base de donnees
DE102004028693B4 (de) Vorrichtung und Verfahren zum Bestimmen eines Akkordtyps, der einem Testsignal zugrunde liegt
EP1671315B1 (fr) Procede et dispositif pour caracteriser un signal audio
EP1377924B1 (fr) Procede et dispositif permettant d'extraire une identification de signaux, procede et dispositif permettant de creer une banque de donnees a partir d'identifications de signaux, et procede et dispositif permettant de se referencer a un signal temps de recherche
Tzanetakis et al. Subband-based drum transcription for audio signals
DE112020002116T5 (de) Informationsverarbeitungsvorrichtung und Verfahren und Programm
Bader Neural coincidence detection strategies during perception of multi-pitch musical tones
EP1743324B1 (fr) Dispositif et procede pour analyser un signal d'information
Tjahyanto et al. Gamelan instrument sound recognition using spectral and facial features of the first harmonic frequency
Wang et al. The analysis and comparison of vital acoustic features in content-based classification of music genre
Morman et al. A system for the automatic segmentation and classification of chord sequences
Pérez Fernández et al. A comparison of pitch chroma extraction algorithms
Raposo Improving Acoustic Features for Structural Segmentation of Music
Krusche Visualization and auralization of features learned by neural networks for musical instrument recognition
DE102006014507B4 (de) Verfahren und Vorrichtung zur Klassifikation und Beurteilung von Musikinstrumenten gleicher Instrumentengruppen
Vogl et al. NEURAL NETWORKS FOR DRUM TRANSCRIPTION
Nam et al. Automatic Music Style Classification: towards the detection of perceptually similar music
Osmanovic Classification of musical genres using audio waveform descriptors in MPEG-7

Legal Events

Date Code Title Description
PUAI Public reference made under article 153(3) epc to a published international application that has entered the european phase

Free format text: ORIGINAL CODE: 0009012

17P Request for examination filed

Effective date: 20060420

AK Designated contracting states

Kind code of ref document: A1

Designated state(s): AT BE BG CH CY CZ DE DK EE ES FI FR GB GR HU IE IS IT LI LT LU MC NL PL PT RO SE SI SK TR

AX Request for extension of the european patent

Extension state: AL BA HR LV MK YU

RIC1 Information provided on ipc code assigned before grant

Ipc: G10L 11/00 20060101AFI20060811BHEP

GRAP Despatch of communication of intention to grant a patent

Free format text: ORIGINAL CODE: EPIDOSNIGR1

RAP1 Party data changed (applicant data changed or rights of an application transferred)

Owner name: FRAUNHOFER-GESELLSCHAFT ZUR FOERDERUNG DER ANGEWAN

RIN1 Information on inventor provided before grant (corrected)

Inventor name: UHLE, CHRISTIAN

Inventor name: CREMER, MARKUS

GRAS Grant fee paid

Free format text: ORIGINAL CODE: EPIDOSNIGR3

GRAA (expected) grant

Free format text: ORIGINAL CODE: 0009210

AK Designated contracting states

Kind code of ref document: B1

Designated state(s): DE FR GB

DAX Request for extension of the european patent (deleted)
RBV Designated contracting states (corrected)

Designated state(s): DE FR GB

REG Reference to a national code

Ref country code: GB

Ref legal event code: FG4D

Free format text: NOT ENGLISH

GBT Gb: translation of ep patent filed (gb section 77(6)(a)/1977)

Effective date: 20070510

REF Corresponds to:

Ref document number: 502005000658

Country of ref document: DE

Date of ref document: 20070614

Kind code of ref document: P

ET Fr: translation filed
PLBE No opposition filed within time limit

Free format text: ORIGINAL CODE: 0009261

STAA Information on the status of an ep patent application or granted ep patent

Free format text: STATUS: NO OPPOSITION FILED WITHIN TIME LIMIT

26N No opposition filed

Effective date: 20080205

REG Reference to a national code

Ref country code: FR

Ref legal event code: PLFP

Year of fee payment: 12

REG Reference to a national code

Ref country code: FR

Ref legal event code: PLFP

Year of fee payment: 13

REG Reference to a national code

Ref country code: FR

Ref legal event code: PLFP

Year of fee payment: 14

PGFP Annual fee paid to national office [announced via postgrant information from national office to epo]

Ref country code: DE

Payment date: 20180423

Year of fee payment: 14

PGFP Annual fee paid to national office [announced via postgrant information from national office to epo]

Ref country code: FR

Payment date: 20180424

Year of fee payment: 14

PGFP Annual fee paid to national office [announced via postgrant information from national office to epo]

Ref country code: GB

Payment date: 20180403

Year of fee payment: 14

REG Reference to a national code

Ref country code: DE

Ref legal event code: R119

Ref document number: 502005000658

Country of ref document: DE

GBPC Gb: european patent ceased through non-payment of renewal fee

Effective date: 20190427

PG25 Lapsed in a contracting state [announced via postgrant information from national office to epo]

Ref country code: DE

Free format text: LAPSE BECAUSE OF NON-PAYMENT OF DUE FEES

Effective date: 20191101

Ref country code: GB

Free format text: LAPSE BECAUSE OF NON-PAYMENT OF DUE FEES

Effective date: 20190427

PG25 Lapsed in a contracting state [announced via postgrant information from national office to epo]

Ref country code: FR

Free format text: LAPSE BECAUSE OF NON-PAYMENT OF DUE FEES

Effective date: 20190430