AudioQualityEnhancer 0.14.0

Added

• Added an inactivity watchdog: FFmpeg/FFprobe processes that stop responding are terminated and reported instead of hanging the run.
• Added a free disk space check before processing starts, with needed/free numbers in the error message.
• Added folder support for drag and drop; dropped folders contribute their supported audio files recursively.
• Added window size and maximized-state persistence across sessions.
• Added a selectable loudness target (-14 LUFS streaming, -16 LUFS podcast, -23 LUFS EBU R128 broadcast) with the preset default as the automatic fallback; result validation checks against the selected target.
• Added an optional adaptive noise tracking mode for the noise reduction presets for material with varying background noise.
• Added safe cleanup for stale processed-preview WAV files.
• Added tests covering processed-preview cache invalidation for filter options and selected audio streams.
• Added settings persistence tests for corrupt, missing and non-writable settings paths.

Changed

• Mono sources are now normalized with a dual-mono EBU R128 measurement so they no longer end up about 3 LU too loud.
• Lossy export formats now get extra true-peak headroom (-2.0 dBTP instead of -1.5 dBTP) because lossy encoding can push peaks above the limited value.
• The preset preview now uses the same two-pass linear loudness normalization as the export and renders in 24-bit.
• The preset preview of longer sources now plays the loudest section instead of always the first 20 seconds.
• Adding multiple files now runs the analyses in parallel, and FFmpeg/FFprobe discovery results are cached for faster batch processing.
• The main view model is organized into feature-focused partial classes; log output is batched into fewer UI updates during processing.
• The main window XAML is organized into section UserControls with shared styles in an application-level resource dictionary.
• The in-memory log is now capped so very large batch runs keep the UI responsive; the oldest lines are replaced by a truncation marker.
• Missing output loudness/peak diagnostics now produce a warning for loudness-changing presets.

Fixed

• Fixed loudness-normalized exports being written at 192 kHz instead of the source sample rate (oversized WAV/FLAC files, 96 kHz AAC).