|
wolffd@0
|
1 function demo7tonality
|
|
wolffd@0
|
2 %To get familiar with some approaches of tonal analysis using MIRtoolbox,
|
|
wolffd@0
|
3 % and to assess their performances.
|
|
wolffd@0
|
4
|
|
wolffd@0
|
5 % Part 1. We will first investigate the performance of the chromagram
|
|
wolffd@0
|
6 % analysis, using very simple musical samples.
|
|
wolffd@0
|
7
|
|
wolffd@0
|
8 % 1.3. In the audio file ÔtrumpetÕ, the same pitch is played by a trumpet.
|
|
wolffd@0
|
9 % Compute its chromagram. What are the chromas detected by the function?
|
|
wolffd@0
|
10 % Can you explain the result?
|
|
wolffd@0
|
11 mirchromagram('trumpet')
|
|
wolffd@0
|
12
|
|
wolffd@0
|
13 % 1.4. A more detailed representation of the chromagram can be obtained by
|
|
wolffd@0
|
14 % decomposing each pitch class into its different possible absolute values.
|
|
wolffd@0
|
15 % For that purpose, just add the parameter:
|
|
wolffd@0
|
16 c = mirchromagram('trumpet','Wrap',0)
|
|
wolffd@0
|
17
|
|
wolffd@0
|
18 % 1.5. Compute also the key strength related to the chromagram.
|
|
wolffd@0
|
19 mirkeystrength(c)
|
|
wolffd@0
|
20
|
|
wolffd@0
|
21 pause, close all
|
|
wolffd@0
|
22
|
|
wolffd@0
|
23 % 1.6. In the audio file ÔpianoF4Õ, the same pitch is played by a piano.
|
|
wolffd@0
|
24 [ks c] = mirkeystrength('pianoF4')
|
|
wolffd@0
|
25
|
|
wolffd@0
|
26 a = miraudio('pianoF4','excerpt',.7,2);
|
|
wolffd@0
|
27 [ks c] = mirkeystrength(a)
|
|
wolffd@0
|
28
|
|
wolffd@0
|
29 pause, close all
|
|
wolffd@0
|
30
|
|
wolffd@0
|
31 % 1.7. Investigate the chromagram analysis of triad chords
|
|
wolffd@0
|
32 [ks c] = mirkeystrength('Amin3')
|
|
wolffd@0
|
33 p = mirpeaks(ks)
|
|
wolffd@0
|
34 mirkey(p)
|
|
wolffd@0
|
35
|
|
wolffd@0
|
36 pause, close all
|
|
wolffd@0
|
37
|
|
wolffd@0
|
38 [k kc ks] = mirkey('Amaj3')
|
|
wolffd@0
|
39 [k kc ks] = mirkey('Amin4')
|
|
wolffd@0
|
40 [k kc ks] = mirkey('Amaj4')
|
|
wolffd@0
|
41
|
|
wolffd@0
|
42 pause, close all
|
|
wolffd@0
|
43
|
|
wolffd@0
|
44 [k kc ks] = mirkey('Amin5')
|
|
wolffd@0
|
45 [k kc ks] = mirkey('Amaj5')
|
|
wolffd@0
|
46 [k kc ks] = mirkey('Cmaj')
|
|
wolffd@0
|
47
|
|
wolffd@0
|
48 pause, close all
|
|
wolffd@0
|
49
|
|
wolffd@0
|
50 [k kc ks] = mirkey(miraudio('Amin3','Excerpt',.2,1))
|
|
wolffd@0
|
51 [k kc ks] = mirkey(miraudio('Amin4','Excerpt',.2,1))
|
|
wolffd@0
|
52 [k kc ks] = mirkey(miraudio('Amaj3','Excerpt',.2,1))
|
|
wolffd@0
|
53 [k kc ks] = mirkey(miraudio('Amaj4','Excerpt',.2,1))
|
|
wolffd@0
|
54
|
|
wolffd@0
|
55 pause, close all
|
|
wolffd@0
|
56
|
|
wolffd@0
|
57 [k kc ks] = mirkey(miraudio('Amin5','Excerpt',.2,1))
|
|
wolffd@0
|
58 [k kc ks] = mirkey(miraudio('Amaj5','Excerpt',.2,1))
|
|
wolffd@0
|
59 [k kc ks] = mirkey(miraudio('Cmaj','Excerpt',.2,1))
|
|
wolffd@0
|
60
|
|
wolffd@0
|
61 pause, close all
|
|
wolffd@0
|
62
|
|
wolffd@0
|
63 %Part 2. Let's analyze several extracts from
|
|
wolffd@0
|
64 %real music. For each extract, try the
|
|
wolffd@0
|
65 %following:
|
|
wolffd@0
|
66
|
|
wolffd@0
|
67 %2.1. Listen to the piece:
|
|
wolffd@0
|
68 mirplay('vivaldi')
|
|
wolffd@0
|
69 soundsc(sin(2*pi*440*(0:1/8192:1)))
|
|
wolffd@0
|
70
|
|
wolffd@0
|
71 %2.2. Compute the chromagram of the
|
|
wolffd@0
|
72 %whole extract. What tonal center could be
|
|
wolffd@0
|
73 %inferred from the curve? Does it
|
|
wolffd@0
|
74 %correspond to your expectation?
|
|
wolffd@0
|
75 c = mirchromagram('vivaldi')
|
|
wolffd@0
|
76
|
|
wolffd@0
|
77 %2.3. Compute the key strength related to
|
|
wolffd@0
|
78 %the chromagram. Is the result congruent
|
|
wolffd@0
|
79 %with the tonality inferred in 2.1. and 2.2.?
|
|
wolffd@0
|
80 ks = mirkeystrength(c)
|
|
wolffd@0
|
81 [k kc ks] = mirkey(ks)
|
|
wolffd@0
|
82
|
|
wolffd@0
|
83 %2.4. A more detailed representation of the
|
|
wolffd@0
|
84 %key strengths can be obtained by
|
|
wolffd@0
|
85 %computing the self-organizing map:
|
|
wolffd@0
|
86 som = mirkeysom(c)
|
|
wolffd@0
|
87
|
|
wolffd@0
|
88 pause, close all
|
|
wolffd@0
|
89
|
|
wolffd@0
|
90 %2.5. Analyse in the same way other audio files
|
|
wolffd@0
|
91 [k kc ks] = mirkey('czardas')
|
|
wolffd@0
|
92 mirkeysom('czardas')
|
|
wolffd@0
|
93
|
|
wolffd@0
|
94 pause, close all
|
|
wolffd@0
|
95
|
|
wolffd@0
|
96 %Part 3. The temporal evolution of the
|
|
wolffd@0
|
97 %tonal dimension can be assessed by
|
|
wolffd@0
|
98 %decomposing first the audio into frames
|
|
wolffd@0
|
99 [k kc ks] = mirkey('czardas','frame')
|
|
wolffd@0
|
100 mirkeysom('czardas','frame') |
