Features
Audio Analysis
DR Score (Dynamic Range)
- Block-based analysis (3-second blocks) using the DR Database methodology
- Top 20% loudest blocks used for accurate DR calculation
- Rating: DR14+ Excellent, DR10-13 Good, DR6-9 Compressed, DR<6 Crushed
Integrated Loudness (LUFS)
- EBU R128 gated measurement
- Absolute and relative gating thresholds
- Loudness Range (LRA) calculation
Clipping Detection
- Scans for samples at 0dBFS threshold
- Consecutive clip event counting
- True Peak via 4x oversampling (inter-sample peak detection)
Spectral Analysis & Lossy Detection
- 8192-point FFT with Hann windowing
- Average spectrum across overlapping windows
- Gradient analysis detects sharp spectral cutoffs:
- ~16kHz = MP3 128kbps
- ~19kHz = MP3 256kbps
- ~20kHz = MP3 320kbps / AAC 256kbps
- Confidence scoring to minimize false positives
DSD Support
- Custom DSF binary parser (reads header + data chunks)
- Two-stage DSD-to-PCM decimation:
- Byte-level popcount at intermediate rate (352.8kHz for DSD64)
- 64-tap FIR low-pass filter (windowed sinc, 20kHz cutoff) + decimation via vDSP
- Efficient network reading: only 2 I/O operations (512-byte header + ~1.2MB audio data)
Visualization
- Min/max envelope with RMS overlay
- Color-coded by level (green / yellow / red)
- 2000-point resolution
Frequency Spectrum
- Logarithmic frequency scale (20Hz to Nyquist)
- Fixed 0 to -90dB range
- Filled gradient display
- Cutoff marker for suspicious files
Integration
- Finder: Right-click any audio file, Open With, Dr. Doctor
- Drag & drop: Drop audio files directly onto the window
- Path input: Paste file paths for network drives
- File picker: Standard macOS open dialog
Technical
- Built with SwiftUI (macOS 14+ Sonoma)
- Accelerate/vDSP for hardware-accelerated signal processing
- Signed and notarized with Apple Developer ID
- Universal binary (Intel + Apple Silicon)