ss
Browse files- __pycache__/beat_analysis.cpython-310.pyc +0 -0
- app.py +21 -16
- beat_analysis.py +23 -7
- emotionanalysis.py +9 -21
__pycache__/beat_analysis.cpython-310.pyc
CHANGED
Binary files a/__pycache__/beat_analysis.cpython-310.pyc and b/__pycache__/beat_analysis.cpython-310.pyc differ
|
|
app.py
CHANGED
@@ -101,22 +101,6 @@ def process_audio(audio_file):
|
|
101 |
# Analyze music with MusicAnalyzer
|
102 |
music_analysis = music_analyzer.analyze_music(audio_file)
|
103 |
|
104 |
-
# Extract time signature from MusicAnalyzer result
|
105 |
-
time_signature = music_analysis["rhythm_analysis"]["estimated_time_signature"]
|
106 |
-
|
107 |
-
# Ensure time signature is one of the supported ones (4/4, 3/4, 2/4, 6/8)
|
108 |
-
if time_signature not in ["4/4", "3/4", "2/4", "6/8"]:
|
109 |
-
time_signature = "4/4" # Default to 4/4 if unsupported
|
110 |
-
music_analysis["rhythm_analysis"]["estimated_time_signature"] = time_signature
|
111 |
-
|
112 |
-
# Analyze beat patterns and create lyrics template using MusicAnalyzer's time signature
|
113 |
-
beat_analysis = beat_analyzer.analyze_beat_pattern(audio_file, time_signature=time_signature)
|
114 |
-
lyric_templates = beat_analyzer.create_lyric_template(beat_analysis)
|
115 |
-
|
116 |
-
# Store these in the music_analysis dict for use in lyrics generation
|
117 |
-
music_analysis["beat_analysis"] = beat_analysis
|
118 |
-
music_analysis["lyric_templates"] = lyric_templates
|
119 |
-
|
120 |
# Extract key information
|
121 |
tempo = music_analysis["rhythm_analysis"]["tempo"]
|
122 |
emotion = music_analysis["emotion_analysis"]["primary_emotion"]
|
@@ -151,6 +135,27 @@ def process_audio(audio_file):
|
|
151 |
genre_results_text = format_genre_results(top_genres)
|
152 |
primary_genre = top_genres[0][0]
|
153 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
154 |
# Prepare analysis summary
|
155 |
analysis_summary = f"""
|
156 |
### Music Analysis Results
|
|
|
101 |
# Analyze music with MusicAnalyzer
|
102 |
music_analysis = music_analyzer.analyze_music(audio_file)
|
103 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
104 |
# Extract key information
|
105 |
tempo = music_analysis["rhythm_analysis"]["tempo"]
|
106 |
emotion = music_analysis["emotion_analysis"]["primary_emotion"]
|
|
|
135 |
genre_results_text = format_genre_results(top_genres)
|
136 |
primary_genre = top_genres[0][0]
|
137 |
|
138 |
+
# Override time signature for pop and disco genres to always be 4/4
|
139 |
+
if any(genre.lower() in primary_genre.lower() for genre in ['pop', 'disco']):
|
140 |
+
music_analysis["rhythm_analysis"]["estimated_time_signature"] = "4/4"
|
141 |
+
time_signature = "4/4"
|
142 |
+
else:
|
143 |
+
# Use detected time signature for other genres
|
144 |
+
time_signature = music_analysis["rhythm_analysis"]["estimated_time_signature"]
|
145 |
+
|
146 |
+
# Ensure time signature is one of the supported ones (4/4, 3/4, 6/8)
|
147 |
+
if time_signature not in ["4/4", "3/4", "6/8"]:
|
148 |
+
time_signature = "4/4" # Default to 4/4 if unsupported
|
149 |
+
music_analysis["rhythm_analysis"]["estimated_time_signature"] = time_signature
|
150 |
+
|
151 |
+
# Analyze beat patterns and create lyrics template using the time signature
|
152 |
+
beat_analysis = beat_analyzer.analyze_beat_pattern(audio_file, time_signature=time_signature)
|
153 |
+
lyric_templates = beat_analyzer.create_lyric_template(beat_analysis)
|
154 |
+
|
155 |
+
# Store these in the music_analysis dict for use in lyrics generation
|
156 |
+
music_analysis["beat_analysis"] = beat_analysis
|
157 |
+
music_analysis["lyric_templates"] = lyric_templates
|
158 |
+
|
159 |
# Prepare analysis summary
|
160 |
analysis_summary = f"""
|
161 |
### Music Analysis Results
|
beat_analysis.py
CHANGED
@@ -15,12 +15,11 @@ except LookupError:
|
|
15 |
class BeatAnalyzer:
|
16 |
def __init__(self):
|
17 |
# Mapping for standard stress patterns by time signature
|
18 |
-
# Simplified to only include 4/4, 3/4,
|
19 |
self.stress_patterns = {
|
20 |
# Format: Strong (1.0), Medium (0.5), Weak (0.0)
|
21 |
"4/4": [1.0, 0.0, 0.5, 0.0], # Strong, weak, medium, weak
|
22 |
"3/4": [1.0, 0.0, 0.0], # Strong, weak, weak
|
23 |
-
"2/4": [1.0, 0.0], # Strong, weak
|
24 |
"6/8": [1.0, 0.0, 0.0, 0.5, 0.0, 0.0] # Strong, weak, weak, medium, weak, weak
|
25 |
}
|
26 |
|
@@ -58,6 +57,28 @@ class BeatAnalyzer:
|
|
58 |
# making them ideal for our beat-matching algorithm
|
59 |
self.supported_genres = ['pop', 'rock', 'country', 'disco', 'metal']
|
60 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
61 |
@lru_cache(maxsize=128)
|
62 |
def count_syllables(self, word):
|
63 |
"""Count syllables in a word using CMU dictionary if available, otherwise use rule-based method."""
|
@@ -157,11 +178,6 @@ class BeatAnalyzer:
|
|
157 |
stress = "M" # Medium
|
158 |
else: # Other beats (weak)
|
159 |
stress = "W" # Weak
|
160 |
-
elif time_signature == "2/4":
|
161 |
-
if metrical_position == 0: # First beat (strongest)
|
162 |
-
stress = "S" # Strong
|
163 |
-
else: # Second beat (weak)
|
164 |
-
stress = "W" # Weak
|
165 |
else:
|
166 |
# Default pattern for other time signatures
|
167 |
if metrical_position == 0:
|
|
|
15 |
class BeatAnalyzer:
|
16 |
def __init__(self):
|
17 |
# Mapping for standard stress patterns by time signature
|
18 |
+
# Simplified to only include 4/4, 3/4, and 6/8
|
19 |
self.stress_patterns = {
|
20 |
# Format: Strong (1.0), Medium (0.5), Weak (0.0)
|
21 |
"4/4": [1.0, 0.0, 0.5, 0.0], # Strong, weak, medium, weak
|
22 |
"3/4": [1.0, 0.0, 0.0], # Strong, weak, weak
|
|
|
23 |
"6/8": [1.0, 0.0, 0.0, 0.5, 0.0, 0.0] # Strong, weak, weak, medium, weak, weak
|
24 |
}
|
25 |
|
|
|
57 |
# making them ideal for our beat-matching algorithm
|
58 |
self.supported_genres = ['pop', 'rock', 'country', 'disco', 'metal']
|
59 |
|
60 |
+
# Common time signatures and their beat patterns with weights for prior probability
|
61 |
+
# Simplified to only include 4/4, 3/4, and 6/8
|
62 |
+
self.common_time_signatures = {
|
63 |
+
"4/4": {"beats_per_bar": 4, "beat_pattern": [1.0, 0.2, 0.5, 0.2], "weight": 0.55},
|
64 |
+
"3/4": {"beats_per_bar": 3, "beat_pattern": [1.0, 0.2, 0.3], "weight": 0.30},
|
65 |
+
"6/8": {"beats_per_bar": 6, "beat_pattern": [1.0, 0.2, 0.3, 0.8, 0.2, 0.3], "weight": 0.15}
|
66 |
+
}
|
67 |
+
|
68 |
+
# Add common accent patterns for different time signatures
|
69 |
+
self.accent_patterns = {
|
70 |
+
"4/4": [[1, 0, 0, 0], [1, 0, 2, 0], [1, 0, 2, 0, 3, 0, 2, 0]],
|
71 |
+
"3/4": [[1, 0, 0], [1, 0, 2]],
|
72 |
+
"6/8": [[1, 0, 0, 2, 0, 0], [1, 0, 0, 2, 0, 3]]
|
73 |
+
}
|
74 |
+
|
75 |
+
# Expected rhythm density (relative note density per beat) for different time signatures
|
76 |
+
self.rhythm_density = {
|
77 |
+
"4/4": [1.0, 0.7, 0.8, 0.6],
|
78 |
+
"3/4": [1.0, 0.6, 0.7],
|
79 |
+
"6/8": [1.0, 0.5, 0.4, 0.8, 0.5, 0.4]
|
80 |
+
}
|
81 |
+
|
82 |
@lru_cache(maxsize=128)
|
83 |
def count_syllables(self, word):
|
84 |
"""Count syllables in a word using CMU dictionary if available, otherwise use rule-based method."""
|
|
|
178 |
stress = "M" # Medium
|
179 |
else: # Other beats (weak)
|
180 |
stress = "W" # Weak
|
|
|
|
|
|
|
|
|
|
|
181 |
else:
|
182 |
# Default pattern for other time signatures
|
183 |
if metrical_position == 0:
|
emotionanalysis.py
CHANGED
@@ -36,19 +36,17 @@ class MusicAnalyzer:
|
|
36 |
self.key_names = ['C', 'C#', 'D', 'D#', 'E', 'F', 'F#', 'G', 'G#', 'A', 'A#', 'B']
|
37 |
|
38 |
# Common time signatures and their beat patterns with weights for prior probability
|
39 |
-
# Simplified to only include 4/4, 3/4,
|
40 |
self.common_time_signatures = {
|
41 |
"4/4": {"beats_per_bar": 4, "beat_pattern": [1.0, 0.2, 0.5, 0.2], "weight": 0.45},
|
42 |
"3/4": {"beats_per_bar": 3, "beat_pattern": [1.0, 0.2, 0.3], "weight": 0.25},
|
43 |
-
"
|
44 |
-
"6/8": {"beats_per_bar": 6, "beat_pattern": [1.0, 0.2, 0.3, 0.8, 0.2, 0.3], "weight": 0.15}
|
45 |
}
|
46 |
|
47 |
# Add common accent patterns for different time signatures
|
48 |
self.accent_patterns = {
|
49 |
"4/4": [[1, 0, 0, 0], [1, 0, 2, 0], [1, 0, 2, 0, 3, 0, 2, 0]],
|
50 |
"3/4": [[1, 0, 0], [1, 0, 2]],
|
51 |
-
"2/4": [[1, 0], [1, 2]],
|
52 |
"6/8": [[1, 0, 0, 2, 0, 0], [1, 0, 0, 2, 0, 3]]
|
53 |
}
|
54 |
|
@@ -56,8 +54,7 @@ class MusicAnalyzer:
|
|
56 |
self.rhythm_density = {
|
57 |
"4/4": [1.0, 0.7, 0.8, 0.6],
|
58 |
"3/4": [1.0, 0.6, 0.7],
|
59 |
-
"6/8": [1.0, 0.5, 0.4, 0.8, 0.5, 0.4]
|
60 |
-
"2/4": [1.0, 0.6]
|
61 |
}
|
62 |
|
63 |
def load_audio(self, file_path, sr=22050, duration=None):
|
@@ -343,16 +340,10 @@ class MusicAnalyzer:
|
|
343 |
for ts in self.common_time_signatures:
|
344 |
score = 0
|
345 |
|
346 |
-
if ts == "4/4" or ts == "
|
347 |
-
# Look for ratios close to
|
348 |
for ratio in tempo_ratios:
|
349 |
-
if abs(ratio -
|
350 |
-
score += 1
|
351 |
-
|
352 |
-
elif ts == "3/4" or ts == "6/8":
|
353 |
-
# Look for ratios close to 3 or 6
|
354 |
-
for ratio in tempo_ratios:
|
355 |
-
if abs(ratio - 3) < 0.2 or abs(ratio - 6) < 0.3:
|
356 |
score += 1
|
357 |
|
358 |
# Normalize score
|
@@ -440,7 +431,7 @@ class MusicAnalyzer:
|
|
440 |
def _estimate_from_tempo(self, tempo):
|
441 |
"""Use tempo to help estimate likely time signature"""
|
442 |
# Statistical tendencies: slower tempos often in compound meters (6/8)
|
443 |
-
# Fast tempos
|
444 |
|
445 |
scores = {}
|
446 |
|
@@ -449,7 +440,6 @@ class MusicAnalyzer:
|
|
449 |
scores = {
|
450 |
"4/4": 0.5,
|
451 |
"3/4": 0.4,
|
452 |
-
"2/4": 0.3,
|
453 |
"6/8": 0.7
|
454 |
}
|
455 |
elif 70 <= tempo <= 120:
|
@@ -457,14 +447,12 @@ class MusicAnalyzer:
|
|
457 |
scores = {
|
458 |
"4/4": 0.7,
|
459 |
"3/4": 0.6,
|
460 |
-
"2/4": 0.4,
|
461 |
"6/8": 0.3
|
462 |
}
|
463 |
else:
|
464 |
-
# Fast tempos favor
|
465 |
scores = {
|
466 |
-
"4/4": 0.
|
467 |
-
"2/4": 0.7,
|
468 |
"3/4": 0.4,
|
469 |
"6/8": 0.2
|
470 |
}
|
|
|
36 |
self.key_names = ['C', 'C#', 'D', 'D#', 'E', 'F', 'F#', 'G', 'G#', 'A', 'A#', 'B']
|
37 |
|
38 |
# Common time signatures and their beat patterns with weights for prior probability
|
39 |
+
# Simplified to only include 4/4, 3/4, and 6/8
|
40 |
self.common_time_signatures = {
|
41 |
"4/4": {"beats_per_bar": 4, "beat_pattern": [1.0, 0.2, 0.5, 0.2], "weight": 0.45},
|
42 |
"3/4": {"beats_per_bar": 3, "beat_pattern": [1.0, 0.2, 0.3], "weight": 0.25},
|
43 |
+
"6/8": {"beats_per_bar": 6, "beat_pattern": [1.0, 0.2, 0.3, 0.8, 0.2, 0.3], "weight": 0.30}
|
|
|
44 |
}
|
45 |
|
46 |
# Add common accent patterns for different time signatures
|
47 |
self.accent_patterns = {
|
48 |
"4/4": [[1, 0, 0, 0], [1, 0, 2, 0], [1, 0, 2, 0, 3, 0, 2, 0]],
|
49 |
"3/4": [[1, 0, 0], [1, 0, 2]],
|
|
|
50 |
"6/8": [[1, 0, 0, 2, 0, 0], [1, 0, 0, 2, 0, 3]]
|
51 |
}
|
52 |
|
|
|
54 |
self.rhythm_density = {
|
55 |
"4/4": [1.0, 0.7, 0.8, 0.6],
|
56 |
"3/4": [1.0, 0.6, 0.7],
|
57 |
+
"6/8": [1.0, 0.5, 0.4, 0.8, 0.5, 0.4]
|
|
|
58 |
}
|
59 |
|
60 |
def load_audio(self, file_path, sr=22050, duration=None):
|
|
|
340 |
for ts in self.common_time_signatures:
|
341 |
score = 0
|
342 |
|
343 |
+
if ts == "4/4" or ts == "6/8":
|
344 |
+
# Look for ratios close to 4 or 6
|
345 |
for ratio in tempo_ratios:
|
346 |
+
if abs(ratio - 4) < 0.2 or abs(ratio - 6) < 0.3:
|
|
|
|
|
|
|
|
|
|
|
|
|
347 |
score += 1
|
348 |
|
349 |
# Normalize score
|
|
|
431 |
def _estimate_from_tempo(self, tempo):
|
432 |
"""Use tempo to help estimate likely time signature"""
|
433 |
# Statistical tendencies: slower tempos often in compound meters (6/8)
|
434 |
+
# Fast tempos favor 4/4
|
435 |
|
436 |
scores = {}
|
437 |
|
|
|
440 |
scores = {
|
441 |
"4/4": 0.5,
|
442 |
"3/4": 0.4,
|
|
|
443 |
"6/8": 0.7
|
444 |
}
|
445 |
elif 70 <= tempo <= 120:
|
|
|
447 |
scores = {
|
448 |
"4/4": 0.7,
|
449 |
"3/4": 0.6,
|
|
|
450 |
"6/8": 0.3
|
451 |
}
|
452 |
else:
|
453 |
+
# Fast tempos favor 4/4
|
454 |
scores = {
|
455 |
+
"4/4": 0.8,
|
|
|
456 |
"3/4": 0.4,
|
457 |
"6/8": 0.2
|
458 |
}
|