How to Use Spectrograms for Audio Comparison and Analysis
While waveforms show amplitude over time, spectrograms reveal the frequency content of audio. They're essential for comparing codecs, analyzing EQ changes, detecting artifacts, and understanding what's really happening in your audio files. This guide explains how to read spectrograms and use them effectively for audio comparison in DualView.
What Is a Spectrogram?
A spectrogram is a visual representation of the frequency spectrum over time. It's created using Fast Fourier Transform (FFT) to break down the audio signal into its component frequencies. The three dimensions of a spectrogram are:
- X-axis (horizontal) — Time progression
- Y-axis (vertical) — Frequency (low at bottom, high at top)
- Color/brightness — Amplitude (louder = brighter)
Frequency Range Reference
Frequency Ranges Explained
| Range | Frequencies | What You'll See |
|---|---|---|
| Sub-bass | 20-60 Hz | Kick drum fundamental, bass drops, rumble |
| Bass | 60-250 Hz | Bass guitar, low synths, vocal chest tone |
| Low Mids | 250-500 Hz | Warmth, body of instruments, muddiness zone |
| Midrange | 500 Hz-2 kHz | Vocals, guitars, snare body, most instruments |
| Upper Mids | 2-4 kHz | Vocal presence, guitar attack, clarity |
| Presence | 4-6 kHz | Definition, consonants, snare crack |
| Brilliance | 6-20 kHz | Air, cymbals, sibilance, high harmonics |
Using Spectrograms for Comparison
Comparing Codecs (MP3, AAC, FLAC)
One of the most common uses for spectrogram comparison is evaluating lossy compression. Load the original lossless file and the compressed version into DualView and compare their spectrograms side-by-side. Here's what to look for:
- High-frequency cutoff — MP3 at 128kbps cuts off around 16kHz
- Missing content above cutoff — Completely black area at top
- Smearing — Less defined transients in the compressed version
- Pre-echo — Artifacts appearing before transients
MP3 128kbps: Cuts ~16kHz
MP3 256kbps: Cuts ~18-19kHz
MP3 320kbps: Cuts ~20kHz
AAC 256kbps: Often preserves to 20kHz with better efficiency
Comparing EQ Changes
When comparing different mix versions or EQ settings, the spectrogram shows exactly what frequencies changed:
- Brighter areas = boosted frequencies
- Darker areas = cut frequencies
- Look for consistent changes across the frequency band
Identifying Noise and Artifacts
Spectrograms reveal problems that may not be obvious when listening:
- Hum (50/60Hz) — Constant horizontal line at 50 or 60Hz
- Hiss — Consistent energy in high frequencies
- Clicks/pops — Vertical lines spanning all frequencies
- Digital artifacts — Unusual patterns, especially in quiet sections
Spectrogram Settings
DualView's spectrogram uses WebGL for real-time rendering. The FFT size determines the trade-off between time resolution and frequency resolution:
- Smaller FFT — Better time resolution, see transients clearly
- Larger FFT — Better frequency resolution, see harmonics clearly
Practical Comparison Workflow
Comparing Audio Files
- Load both audio files into DualView (drag and drop)
- Enable spectrogram view in the analysis panel
- Play both files (they sync automatically)
- Look for differences in frequency content
- Use the slider or side-by-side mode to compare directly
What to Look For
- Missing frequencies — Indicates compression or filtering
- Added noise — Brighter areas that shouldn't be there
- Harmonic changes — Different overtone structure
- Dynamic range differences — Overall brightness variation
Beyond Basic Comparison
For comprehensive audio analysis, combine spectrograms with DualView's other audio tools:
- Waveform — Overall level and dynamics
- LUFS Meter — Loudness measurement
- Goniometer — Stereo field analysis
- Correlation Meter — Phase coherence