Academia.eduAcademia.edu

Figure 6 – uploaded by Kaustuv Ganguli

See full PDFdownloadDownload figure
igure 6: Svara count histograms (octave folded) of 6 concerts in raga Deshkar (left) and 11 concerts in raga Bhupali (right).

Figure 6 igure 6: Svara count histograms (octave folded) of 6 concerts in raga Deshkar (left) and 11 concerts in raga Bhupali (right).

Related Figures (26)

Figure 1: The solfege of Hindustani music shown with an arbitrarily chosen tonic (S) location.
Figure 1: The solfege of Hindustani music shown with an arbitrarily chosen tonic (S) location.
Table 1: Specification of raga grammar for three allied raga-pairs (ITC-SRA, 2017; Rao et al., 1999; Autrim-NCPA, 2017; Oak, 2017; Kulkarni, 2011; SwarGanga, 2017; Mahajan, 2010).
Table 1: Specification of raga grammar for three allied raga-pairs (ITC-SRA, 2017; Rao et al., 1999; Autrim-NCPA, 2017; Oak, 2017; Kulkarni, 2011; SwarGanga, 2017; Mahajan, 2010).
gure 2: Taxonomy of the previous endeavors in raga recognition from first order pitch distributions
gure 2: Taxonomy of the previous endeavors in raga recognition from first order pitch distributions
Figure 3: Block diagram of the signal processing chain from audio signal to pitch distributions.
Figure 3: Block diagram of the signal processing chain from audio signal to pitch distributions.
information is lost in the latter. ‘igure 5: Svara salience histograms (octave folded) of 6 concerts in raga Deshkar (left) and and 11 concerts in raga Bhupali (right).
information is lost in the latter. ‘igure 5: Svara salience histograms (octave folded) of 6 concerts in raga Deshkar (left) and and 11 concerts in raga Bhupali (right).
Figure 4: Pitch salience histograms (octave folded, 1 cent bin resolution) of 6 concerts in raga Deshkar (left) and and 11 concerts in raga Bhupali (right).
Figure 4: Pitch salience histograms (octave folded, 1 cent bin resolution) of 6 concerts in raga Deshkar (left) and and 11 concerts in raga Bhupali (right).
continuous-pitch distributions. Note that the svara his- togram performances are considerably higher than the performance of the same dimension pitch salience dis- tribution of 100-cents bin width. We will see that the similarity in performance between the low-dimensional svara-based histograms and the high-dimensional fine bin-width pitch-salience histograms is reflected also in the outcomes of Experiment 2. This serves to re- inforce the current conclusion about the segmented pitch contour corresponding to the svara regions alone containing sufficient discriminatory information in the context of distributions derived from the melodic con- tour across the concert. That the observed clustering in all configurations actually captures raga characteristics was confirmed by noting that each discovered cluster was heterogeneous in performer and performance meta- data. Figure 7: Cluster purity (CP) values obtained for dif- ferent values of bin resolution for the pitch salience histograms in each of the 3 raga-pairs. 6. Discussion
continuous-pitch distributions. Note that the svara his- togram performances are considerably higher than the performance of the same dimension pitch salience dis- tribution of 100-cents bin width. We will see that the similarity in performance between the low-dimensional svara-based histograms and the high-dimensional fine bin-width pitch-salience histograms is reflected also in the outcomes of Experiment 2. This serves to re- inforce the current conclusion about the segmented pitch contour corresponding to the svara regions alone containing sufficient discriminatory information in the context of distributions derived from the melodic con- tour across the concert. That the observed clustering in all configurations actually captures raga characteristics was confirmed by noting that each discovered cluster was heterogeneous in performer and performance meta- data. Figure 7: Cluster purity (CP) values obtained for dif- ferent values of bin resolution for the pitch salience histograms in each of the 3 raga-pairs. 6. Discussion
Figure 8: Raga-pair Deshkar-Bhupali (full concerts, octave-folded): ROCs obtained for four different distance measures from pitch salience (left), svara salience (middle), and svara count (right) histograms.
Figure 8: Raga-pair Deshkar-Bhupali (full concerts, octave-folded): ROCs obtained for four different distance measures from pitch salience (left), svara salience (middle), and svara count (right) histograms.
Figure 9: Raga-pair Deshkar-Bhupali (full concerts, not octave-folded): ROCs obtained for four different distance measures from pitch salience (left), svara salience (middle), and svara count (right) histograms.
Figure 9: Raga-pair Deshkar-Bhupali (full concerts, not octave-folded): ROCs obtained for four different distance measures from pitch salience (left), svara salience (middle), and svara count (right) histograms.
‘igure 10: Combination of all three raga-pairs (full concerts, octave-folded): ROCs obtained for four different distance measures from pitch salience (left), svara salience (middle), and svara count (right) histograms from the combined distance vectors for all three raga-pairs.
‘igure 10: Combination of all three raga-pairs (full concerts, octave-folded): ROCs obtained for four different distance measures from pitch salience (left), svara salience (middle), and svara count (right) histograms from the combined distance vectors for all three raga-pairs.
‘igure 11: Raga-pair Deshkar-Bhupali (partial concerts): Comparison of ROCs obtained with Bhattacharyya distance for different time-scales (n=1,2,3) of the concerts.
‘igure 11: Raga-pair Deshkar-Bhupali (partial concerts): Comparison of ROCs obtained with Bhattacharyya distance for different time-scales (n=1,2,3) of the concerts.
‘igure 12: Raga-pair Deshkar-Bhupali (initial portion of concerts): Comparison of ROCs obtained with Bhat- tacharyya distance for annotated initial portion (alap+vistar) of the concerts. We plot the ROCs for full (n=1) and half (n=2) concerts for performance comparison.
‘igure 12: Raga-pair Deshkar-Bhupali (initial portion of concerts): Comparison of ROCs obtained with Bhat- tacharyya distance for annotated initial portion (alap+vistar) of the concerts. We plot the ROCs for full (n=1) and half (n=2) concerts for performance comparison.
Figure 13: Svara salience histograms across tala cycles for DK_AC-1 (left), DK_AC-2 (middle), and DK_KA (right).
Figure 13: Svara salience histograms across tala cycles for DK_AC-1 (left), DK_AC-2 (middle), and DK_KA (right).
Figure 14: Svara salience histograms across tala cycles for BP_AC-1 (left), BP_AC-2 (middle), and BP_RK (right).
Figure 14: Svara salience histograms across tala cycles for BP_AC-1 (left), BP_AC-2 (middle), and BP_RK (right).
each concert. This contour is smoothed using a 7-point median filter. Next the individual concert summaries are each time-normalized and superposed in Figure 15. We clearly observe the upward shifting with time of the “focal” note through the allowed svaras. The relative duration spent on each svara is concert dependent. We note the omission of the R svara in Deshkar curves as expected from its grammar. The step-wise movement of the salient svara bears a high resemblance to the time evolution of melody in course of the vistar, as shown in Figure 11 in (Widdess, 2011) including the sharp fall in pitch of the salient svara at the end of many concerts. The latter depicts the descent to a lower svara marking the end of the vistar. the end of the vistar. sure 15: Time-normalized summary of cycle-level salient svara curves over the chosen 6 concerts in raga Deshkar (left) and raga Bhupali (right).
each concert. This contour is smoothed using a 7-point median filter. Next the individual concert summaries are each time-normalized and superposed in Figure 15. We clearly observe the upward shifting with time of the “focal” note through the allowed svaras. The relative duration spent on each svara is concert dependent. We note the omission of the R svara in Deshkar curves as expected from its grammar. The step-wise movement of the salient svara bears a high resemblance to the time evolution of melody in course of the vistar, as shown in Figure 11 in (Widdess, 2011) including the sharp fall in pitch of the salient svara at the end of many concerts. The latter depicts the descent to a lower svara marking the end of the vistar. the end of the vistar. sure 15: Time-normalized summary of cycle-level salient svara curves over the chosen 6 concerts in raga Deshkar (left) and raga Bhupali (right).
Table 4: Metadata associated with each concert in our dataset. We list the artist, tonic, tala, laya, and duration each concert. The acronym correspondences are: Vil = vilambit, Madh = madhyalaya.
Table 4: Metadata associated with each concert in our dataset. We list the artist, tonic, tala, laya, and duration each concert. The acronym correspondences are: Vil = vilambit, Madh = madhyalaya.
samvadi D svara. Figure 16: Pitch salience histograms (octave folded, 1 cent bin resolution) of 7 concerts in raga Puriya (left) and 12 concerts in raga Marwa (right). _ ail ~~ _ mae _ _ From the pitch salience histogram (Figure 16) of raga Puriya (left), we observe the svaras G and N as most prominent, which corresponds to the vadi and samvadi svaras. However, the peak height for the vadi G being less than that for the samvadi N is a result of octave-folding. N has more salience (accumulated duration) because of its presence in both lower and middle octaves, whereas G had most of its occurrences in the middle octave only. The r and D svaras have very low salience, which reinforces the grammaticality in the eminent artists’ performance. The r peak is almost merged with the S peak, this is indicative of the fact that r is taken only within the glide (e.g. N(r)S, N(x)G, G(r)S) and is never sustained as an individual svara. In contrast, the r svara has the highest salience for raga Marwa (right), which also happens to be its vadi svara. The second most salience peak corresponds to the samvadi D svara.
samvadi D svara. Figure 16: Pitch salience histograms (octave folded, 1 cent bin resolution) of 7 concerts in raga Puriya (left) and 12 concerts in raga Marwa (right). _ ail ~~ _ mae _ _ From the pitch salience histogram (Figure 16) of raga Puriya (left), we observe the svaras G and N as most prominent, which corresponds to the vadi and samvadi svaras. However, the peak height for the vadi G being less than that for the samvadi N is a result of octave-folding. N has more salience (accumulated duration) because of its presence in both lower and middle octaves, whereas G had most of its occurrences in the middle octave only. The r and D svaras have very low salience, which reinforces the grammaticality in the eminent artists’ performance. The r peak is almost merged with the S peak, this is indicative of the fact that r is taken only within the glide (e.g. N(r)S, N(x)G, G(r)S) and is never sustained as an individual svara. In contrast, the r svara has the highest salience for raga Marwa (right), which also happens to be its vadi svara. The second most salience peak corresponds to the samvadi D svara.
Figure 17: Svara salience histograms (octave folded) of 7 concerts in raga Puriya (left) and 12 concerts in raga Marwa (right).
Figure 17: Svara salience histograms (octave folded) of 7 concerts in raga Puriya (left) and 12 concerts in raga Marwa (right).
Figure 19: Raga-pair Puriya-Marwa (full concerts, octave-folded): ROCs obtained for four different distance measures from pitch salience (left), svara salience (middle), and svara count (right) histograms. The interpretations from the svara salience histograms (Figure 17) are similar. There is a high visual similarity yetween the pitch salience and svara salience histograms for the corresponding ragas (ignoring the bin mapping lue to difference in resolution), this supports the musicological knowledge that major portion of a phrase in course of the raga development is covered by stable svaras (Kulkarni, 2011). However, in comparison, we find in interesting observation on the relative salience of the r svara in the svara count histograms (Figure 18). In aga Puriya (left), the svara count is relatively high though the same has a very low salience in the svara salience listograms (Figure 17). This indicates a large number of detected r svaras that were of short durations, thereby iccumulated to a low peak in the svara salience histogram. In contrast, the average duration of the r svaras in aga Marwa (right) are high, thereby accumulating to a high peak in the svara salience histogram for a moderate yeak in the svara count. Raga pair Multani-Todi
Figure 19: Raga-pair Puriya-Marwa (full concerts, octave-folded): ROCs obtained for four different distance measures from pitch salience (left), svara salience (middle), and svara count (right) histograms. The interpretations from the svara salience histograms (Figure 17) are similar. There is a high visual similarity yetween the pitch salience and svara salience histograms for the corresponding ragas (ignoring the bin mapping lue to difference in resolution), this supports the musicological knowledge that major portion of a phrase in course of the raga development is covered by stable svaras (Kulkarni, 2011). However, in comparison, we find in interesting observation on the relative salience of the r svara in the svara count histograms (Figure 18). In aga Puriya (left), the svara count is relatively high though the same has a very low salience in the svara salience listograms (Figure 17). This indicates a large number of detected r svaras that were of short durations, thereby iccumulated to a low peak in the svara salience histogram. In contrast, the average duration of the r svaras in aga Marwa (right) are high, thereby accumulating to a high peak in the svara salience histogram for a moderate yeak in the svara count. Raga pair Multani-Todi
Figure 22: Svara count histograms (octave folded) of 7 concerts in raga Multani (left) and 12 concerts in raga Todi (right).
Figure 22: Svara count histograms (octave folded) of 7 concerts in raga Multani (left) and 12 concerts in raga Todi (right).
Figure 23: Raga-pair Multani-Todi (full concerts, octave-folded): ROCs obtained for four different distance measures from pitch salience (left), svara salience (middle), and svara count (right) histograms.
Figure 23: Raga-pair Multani-Todi (full concerts, octave-folded): ROCs obtained for four different distance measures from pitch salience (left), svara salience (middle), and svara count (right) histograms.
Related topics:
Computer ScienceRepresentation Politics
580 California St., Suite 400
San Francisco, CA, 94104
© 2025 Academia. All rights reserved