Academia.eduAcademia.edu

Figure 5 – uploaded by Dimitrios Margounakis

See full PDFdownloadDownload figure
Figure 5. 7 transition during the melodic progression of the Ist violin.

Figure 5 7 transition during the melodic progression of the Ist violin.

Related Figures (14)

Figure 1. The factors that affect music chromaticism in a framework. sic resulted in the modeling of a theoretical framework that calculates the factors of music chromaticism by the authors and is shown in Figure 1 [12].
Figure 1. The factors that affect music chromaticism in a framework. sic resulted in the modeling of a theoretical framework that calculates the factors of music chromaticism by the authors and is shown in Figure 1 [12].
The approach of computational chromatic analysis consists of five stages: (a) extraction of melodic salient pitches (for audio files) or notation for MIDI files, ( matching a scale to the music piece, (c) segmentation, (d) calculation of the chromatic elements, and (e) visu- alization of Chromaticism using colorful strips. T fundamental concepts and algorithms of chromatic analysis are presented in [11] and [12]. Also several b) ne discussions can be found in [10], [14] and [15]. T ne analysis has been conducted with the use of our tool MEL-IRIS client V.3.0. The application’s core function is the chromatic analysis of music, as described here. MEL-IRIS was mainly developed in Borland C++ Builder 6 and uses the MS-SQL database (Figure 2). Further manual mathematic processing has taken place in some cases, using the raw data of MEL-IRIS databa se after the analysis of the quartet. The strategy of our ap- proach is to construct a semantic chromatic graph and identify significant chromatic segments and repetitive patterns during the piece. Figure 2. Snapshots from the tool MEL-IRIS v.2 (top) and MEL-IRIS v.3 (bottom).
The approach of computational chromatic analysis consists of five stages: (a) extraction of melodic salient pitches (for audio files) or notation for MIDI files, ( matching a scale to the music piece, (c) segmentation, (d) calculation of the chromatic elements, and (e) visu- alization of Chromaticism using colorful strips. T fundamental concepts and algorithms of chromatic analysis are presented in [11] and [12]. Also several b) ne discussions can be found in [10], [14] and [15]. T ne analysis has been conducted with the use of our tool MEL-IRIS client V.3.0. The application’s core function is the chromatic analysis of music, as described here. MEL-IRIS was mainly developed in Borland C++ Builder 6 and uses the MS-SQL database (Figure 2). Further manual mathematic processing has taken place in some cases, using the raw data of MEL-IRIS databa se after the analysis of the quartet. The strategy of our ap- proach is to construct a semantic chromatic graph and identify significant chromatic segments and repetitive patterns during the piece. Figure 2. Snapshots from the tool MEL-IRIS v.2 (top) and MEL-IRIS v.3 (bottom).
Moreover, the idiomatic pattern in meter 39, which Forte [4] includes in his table of motives (see Figure 3) adds chroma to the piece (*1.0198), since it contains 2 tones that decrease y by 1.005, but also a chromatic 3/2- tone (Gb-Eb) that increases y by 1.03. Figure 3. A chromatic pattern in Brahms’s quartet (m.39).
Moreover, the idiomatic pattern in meter 39, which Forte [4] includes in his table of motives (see Figure 3) adds chroma to the piece (*1.0198), since it contains 2 tones that decrease y by 1.005, but also a chromatic 3/2- tone (Gb-Eb) that increases y by 1.03. Figure 3. A chromatic pattern in Brahms’s quartet (m.39).
Figure 4. (a)The chromatic graph of the 1st Violin melodic progression.(b) The chromatic graph of the audio recording.
Figure 4. (a)The chromatic graph of the 1st Violin melodic progression.(b) The chromatic graph of the audio recording.
Table 3. Discrete chromatic patterns in the Ist voice
Table 3. Discrete chromatic patterns in the Ist voice
the greatest deviation (0,6). Also, two great peaks (de- viation: 0,4) are observed between measures 200 and 220 (corresponding to pattern A8 of the Ist violin), which qualify Coda.
the greatest deviation (0,6). Also, two great peaks (de- viation: 0,4) are observed between measures 200 and 220 (corresponding to pattern A8 of the Ist violin), which qualify Coda.
Figure 9. The chromatic motive V4 in the Viola (m. 132-136). Figure 8. The chromatic motive V1 in the Viola (m. 33-35).
Figure 9. The chromatic motive V4 in the Viola (m. 132-136). Figure 8. The chromatic motive V1 in the Viola (m. 33-35).
quoted intervals ([105880 ms — 113113 ms]). V4 creates Chromaticism by changing rapidly tonalities, scales and by sharpening and flattening notes. Figure 10. The chromatic motive C3 in the 4th Voice (m. 63-66).
quoted intervals ([105880 ms — 113113 ms]). V4 creates Chromaticism by changing rapidly tonalities, scales and by sharpening and flattening notes. Figure 10. The chromatic motive C3 in the 4th Voice (m. 63-66).
After the separate analysis of the four voices, a cumu- lative result will be presented. By summing up the chro- matic deviations during the time axis, we can get a cleaner image of total chroma through the music con- text. For readability purposes we split the quartet in segments of 10,000 ms (for a total of 50 segments) and we sum the average deviations from the base chroma for all voices in each segment. This way we can gain a view of the harmonic relations of the melodic lines for our MIDI file. Figure 11 shows the result of summing up all voices. The parts of the quartet have approximately the following boundaries: Exposition [segments 1-23] (repetition included), Development [segments 24-30], Recapitulation [segments 31-50] and Coda [segments 44-50]. Figure 11. Adding the chromatic effect of all voices in the 1st Movement.
After the separate analysis of the four voices, a cumu- lative result will be presented. By summing up the chro- matic deviations during the time axis, we can get a cleaner image of total chroma through the music con- text. For readability purposes we split the quartet in segments of 10,000 ms (for a total of 50 segments) and we sum the average deviations from the base chroma for all voices in each segment. This way we can gain a view of the harmonic relations of the melodic lines for our MIDI file. Figure 11 shows the result of summing up all voices. The parts of the quartet have approximately the following boundaries: Exposition [segments 1-23] (repetition included), Development [segments 24-30], Recapitulation [segments 31-50] and Coda [segments 44-50]. Figure 11. Adding the chromatic effect of all voices in the 1st Movement.
Figure 12. y transition of the melodic salient pitches of a live performance (MP3 format). inis 1S because Of the nature ofr Salient pitches that are detected by the system. MIDI files follow the rule of the ‘well-tempered clavier’ while reproducing exactly the music score. As a result, all the intervals, created in MIDI executions, are integer multipliers of the semi- one. This creates limited chromatic perception com- pared to a live recording, where pitches from the whole frequency spectrum are indicated. This may even lead to ‘odd’ intervals, e.g. 35 cents or 118 cents, which en- compasses added chroma. Therefore, proper playing * by the instrument players can create their intended chromatic perception to the audience. It should be pointed out here that strings (in our case the violins, the viola, and the cello) are able to reproduce any pitch, in contrast to the well-tempered piano. However, also a certain amount of ‘noise’ in calculating chroma arises from overtones from each instrument in the signal; these are quite prominent for bowed strings.
Figure 12. y transition of the melodic salient pitches of a live performance (MP3 format). inis 1S because Of the nature ofr Salient pitches that are detected by the system. MIDI files follow the rule of the ‘well-tempered clavier’ while reproducing exactly the music score. As a result, all the intervals, created in MIDI executions, are integer multipliers of the semi- one. This creates limited chromatic perception com- pared to a live recording, where pitches from the whole frequency spectrum are indicated. This may even lead to ‘odd’ intervals, e.g. 35 cents or 118 cents, which en- compasses added chroma. Therefore, proper playing * by the instrument players can create their intended chromatic perception to the audience. It should be pointed out here that strings (in our case the violins, the viola, and the cello) are able to reproduce any pitch, in contrast to the well-tempered piano. However, also a certain amount of ‘noise’ in calculating chroma arises from overtones from each instrument in the signal; these are quite prominent for bowed strings.
By using the method of averaging the standard devia- tions as in the case of MIDI, we tried to compare the two graphs for similarities. Obviously, the comparison cannot be direct for two reasons. Firstly, the MIDI exe- cution bears two repetitions of the Exposition, while the audio recording bears only one. Secondly, the times of execution in live performance are not analogical, since (as previously discussed) the musicians do not toe the mark of the rhythmic motive as in the MIDI execution. Consequently, since no total time concurrency exists, we examined each part separately by stretching the timers where needed. Similarities in chromatic patterns were detected in the Development and Coda (Figure 14). Figure 14. Similarities in chromatic patterns between MIDI and audio: (a)Development, (b) Coda.
By using the method of averaging the standard devia- tions as in the case of MIDI, we tried to compare the two graphs for similarities. Obviously, the comparison cannot be direct for two reasons. Firstly, the MIDI exe- cution bears two repetitions of the Exposition, while the audio recording bears only one. Secondly, the times of execution in live performance are not analogical, since (as previously discussed) the musicians do not toe the mark of the rhythmic motive as in the MIDI execution. Consequently, since no total time concurrency exists, we examined each part separately by stretching the timers where needed. Similarities in chromatic patterns were detected in the Development and Coda (Figure 14). Figure 14. Similarities in chromatic patterns between MIDI and audio: (a)Development, (b) Coda.
increase by U.Z 1s observed in a percentage Of 2.63'/, which implies the salience of the phenomenon. An in- crease of chroma greater than 0.015 is statistically sig- nificant and it is not a very common phenomenon in many cases of western music. Such an increase is a re- sult of continuous chromatic phenomena, which draw the attention for further analysis. That’s why 0.015 is set as the y distance between different colors in our twelve- grade color scale. An increase of chroma greater than 2 usually betrays the temporary change of the underlying key or the transition to another music scale. Although the other chromatic deviations do not occur often in the context (<1% frequency), the greatest deviation that occurs at the end of the exposition should be mentioned (vy ~ 3). The incident represents the colorful melancholic impression of measures 75-80, as they are performed by Delirium Tremens.
increase by U.Z 1s observed in a percentage Of 2.63'/, which implies the salience of the phenomenon. An in- crease of chroma greater than 0.015 is statistically sig- nificant and it is not a very common phenomenon in many cases of western music. Such an increase is a re- sult of continuous chromatic phenomena, which draw the attention for further analysis. That’s why 0.015 is set as the y distance between different colors in our twelve- grade color scale. An increase of chroma greater than 2 usually betrays the temporary change of the underlying key or the transition to another music scale. Although the other chromatic deviations do not occur often in the context (<1% frequency), the greatest deviation that occurs at the end of the exposition should be mentioned (vy ~ 3). The incident represents the colorful melancholic impression of measures 75-80, as they are performed by Delirium Tremens.
Related topics:
Computer MusicVisual Music
580 California St., Suite 400
San Francisco, CA, 94104
© 2025 Academia. All rights reserved