Mirror/lidify
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

chore: Release v1.3.3 - Critical bug fixes and QoL improvements

Browse Source 
Critical Fixes:
- Docker permissions for PostgreSQL/Redis bind mounts
  Fixes #59, fixes #62
- Audio analyzer memory consumption and OOM crashes
  Fixes #21, fixes #26, fixes #53
- LastFM array normalization preventing .map crashes
  Fixes #37, fixes #39
- Wikidata 403 errors from missing User-Agent
  Fixes #57
- Singles directory creation race conditions
  Fixes #58
- Firefox FLAC playback stopping at ~4:34 mark
  Fixes #42, fixes #17

Quality of Life:
- Add Releases link to desktop sidebar navigation
  Fixes #41
- iPhone safe area insets for Dynamic Island/notch
  Fixes #54

Contributors: @arsaboo, @rustyricky, @RustyJonez, @tombatossals

No regressions detected, backward compatible, production ready.
This commit is contained in:
Your Name
2026-01-09 18:31:45 -06:00
parent ce597a318e
commit 0ac805b6fc
17 changed files with 857 additions and 94 deletions
Download Patch File Download Diff File Expand all files Collapse all files
services/audio-analyzer/analyzer.py
+39 -17
View File
@@ -65,6 +65,7 @@ import traceback
import numpy as np
from concurrent.futures import ProcessPoolExecutor, as_completed
import multiprocessing
import gc
# BrokenProcessPool was added in Python 3.9, provide compatibility for Python 3.8
try:
@@ -112,6 +113,14 @@ except ImportError as e:
TF_MODELS_AVAILABLE = False
TensorflowPredictMusiCNN = None
try:
import tensorflow as tf
# Limit TensorFlow memory usage (CPU & GPU)
try:
gpus = tf.config.experimental.list_physical_devices('GPU')
for gpu in gpus:
tf.config.experimental.set_memory_growth(gpu, True)
except Exception:
pass
from essentia.standard import TensorflowPredictMusiCNN
TF_MODELS_AVAILABLE = True
logger.info("TensorflowPredictMusiCNN available - Enhanced mode enabled")
@@ -376,14 +385,18 @@ class AudioAnalyzer:
traceback.print_exc()
self.enhanced_mode = False
def load_audio(self, file_path: str, sample_rate: int = 16000) -> Optional[Any]:
"""Load audio file as mono signal"""
def load_audio(self, file_path: str, sample_rate: int = 16000, max_duration: int = 90) -> Optional[Any]:
"""Load up to max_duration seconds of audio as mono signal (to limit memory usage)"""
if not ESSENTIA_AVAILABLE:
return None
try:
loader = es.MonoLoader(filename=file_path, sampleRate=sample_rate)
audio = loader()
# Limit to max_duration seconds
max_samples = int(sample_rate * max_duration)
if len(audio) > max_samples:
audio = audio[:max_samples]
return audio
except Exception as e:
logger.error(f"Failed to load audio {file_path}: {e}")
@@ -514,12 +527,17 @@ class AudioAnalyzer:
result['_error'] = 'Essentia library not installed'
return result
# Load audio at different sample rates for different algorithms
audio_44k = self.load_audio(file_path, 44100)
audio_16k = self.load_audio(file_path, 16000)
# Limit memory: only analyze up to MAX_ANALYZE_SECONDS (default 90s)
MAX_ANALYZE_SECONDS = int(os.getenv('MAX_ANALYZE_SECONDS', '90'))
try:
# Load audio at different sample rates for different algorithms, limit duration
audio_44k = self.load_audio(file_path, 44100, max_duration=MAX_ANALYZE_SECONDS)
audio_16k = self.load_audio(file_path, 16000, max_duration=MAX_ANALYZE_SECONDS)
except MemoryError:
logger.error(f"MemoryError: Could not load audio for {file_path}")
result['_error'] = 'MemoryError: audio file too large'
return result
if audio_44k is None or audio_16k is None:
result['_error'] = 'Failed to load audio file'
return result
# Validate audio before analysis (Phase 2 defensive improvement)
@@ -586,7 +604,10 @@ class AudioAnalyzer:
# Process audio in frames for detailed analysis
frame_size = 2048
hop_size = 1024
for i in range(0, len(audio_44k) - frame_size, hop_size):
max_frames = int((44100 * MAX_ANALYZE_SECONDS - frame_size) / hop_size)
for idx, i in enumerate(range(0, len(audio_44k) - frame_size, hop_size)):
if idx > max_frames:
break
frame = audio_44k[i:i + frame_size]
windowed = self.windowing(frame)
spectrum = self.spectrum(windowed)
@@ -599,7 +620,6 @@ class AudioAnalyzer:
# RMS-based energy (properly normalized to 0-1)
if rms_values:
avg_rms = np.mean(rms_values)
# RMS is typically 0.0-0.5 for normalized audio, scale to 0-1
result['energy'] = round(min(1.0, float(avg_rms) * 3), 3)
else:
result['energy'] = 0.5
@@ -616,7 +636,6 @@ class AudioAnalyzer:
result['_zcr'] = np.mean(zcr_values) if zcr_values else 0.1
# Basic Danceability (non-ML)
# Note: es.Danceability() can return values > 1.0, so we clamp
danceability, _ = self.danceability_extractor(audio_44k)
result['danceability'] = round(max(0.0, min(1.0, float(danceability))), 3)
@@ -632,22 +651,25 @@ class AudioAnalyzer:
traceback.print_exc()
self._apply_standard_estimates(result, scale, bpm)
else:
# === STANDARD MODE: Use heuristics ===
self._apply_standard_estimates(result, scale, bpm)
# Generate mood tags based on all features
result['moodTags'] = self._generate_mood_tags(result)
logger.info(f"Analysis complete [{result['analysisMode']}]: BPM={result['bpm']}, Key={result['key']} {result['keyScale']}, Valence={result['valence']}, Arousal={result['arousal']}")
except MemoryError:
logger.error(f"MemoryError during analysis of {file_path}")
result['_error'] = 'MemoryError: analysis exceeded memory limits'
except Exception as e:
logger.error(f"Analysis error: {e}")
traceback.print_exc()
# Clean up internal fields before returning
for key in ['_spectral_centroid', '_spectral_flatness', '_zcr']:
result.pop(key, None)
finally:
# Clean up internal fields before returning
for key in ['_spectral_centroid', '_spectral_flatness', '_zcr']:
result.pop(key, None)
# Explicitly free memory
del audio_44k, audio_16k
gc.collect()
return result
def _extract_ml_features(self, audio_16k) -> Dict[str, Any]: