Audio Engineering: Converting Frequencies, Decibels and Bitrates
Published April 24, 2026
Audio uses specialized unit systems: hertz for frequency, decibels for power/pressure (logarithmic), samples per second for digital audio, and bitrates for compression. The decibel scale is logarithmic—10 dB is 10× power, not 10× louder. Understanding these conversions enables audio engineers to design filters, set levels accurately, and diagnose problems in recording, mixing, and playback systems.
Table of Contents
Understanding the Basics
Audio presents unique conversion challenges because human hearing is logarithmic. A 10× increase in power (10 dB) sounds like roughly 2× louder to human ears. A 100× power increase (20 dB) sounds about 4× louder. This logarithmic relationship—the decibel scale—is fundamental to audio engineering. Additionally, audio frequencies range from 20 Hz (lowest human hearing) to 20 kHz (high frequencies); a mixing engineer might work with 100 Hz (bass), 1 kHz (midrange), and 10 kHz (presence) in the same mixing session.
Digital audio introduces additional unit systems: sample rate (44.1 kHz, 48 kHz), bit depth (16, 24 bits), and bitrate (128-320 kbps for MP3). These units relate through Nyquist theorem (sample rate determines frequency ceiling) and Shannon capacity (sample rate and bit depth determine bitrate). Understanding these conversions enables proper audio file selection, streaming optimization, and quality assessment.
Audio Units
Frequency and Acoustics
- Hertz (Hz): Cycles per second. 20 Hz-20 kHz = human hearing range. 440 Hz = musical A note.
- Kilohertz (kHz): 1 kHz = 1,000 Hz. Treble frequencies typically 2-20 kHz.
- Decibel (dB): Logarithmic ratio of power. 10 dB = 10× power; 20 dB = 100× power; 3 dB ≈ 2× power (doubling).
Digital Audio
- Sample Rate: 44.1 kHz (CD), 48 kHz (video/broadcast), 96 kHz (high-resolution). Nyquist: sample rate ÷ 2 = max frequency captured.
- Bit Depth: 16 bits (CD, 96 dB SNR), 24 bits (pro audio, 144 dB SNR). Bits determine quantization noise floor.
- Bitrate: 128-320 kbps (MP3), 128-256 kbps (AAC). Kbps = kilobits per second of compressed audio data.
Conversion Formulas
| Conversion | Formula |
|---|---|
| Power to dB | dB = 10 × log10(P1/P0) |
| Voltage to dB | dB = 20 × log10(V1/V0) |
| dB to Power Ratio | Ratio = 10^(dB/10) |
| Nyquist Frequency | Max Freq = Sample Rate ÷ 2 |
Worked Examples
Example 1: Power Ratio to Decibels
An amplifier has output power 10 times the input. What is the gain in dB?
dB = 10 × log10(10) = 10 × 1 = 10 dB gain. A 10× power increase = 10 dB, barely perceptible as louder.
Example 2: Sample Rate and Frequency
A 48 kHz sample rate can capture audio up to what frequency?
Max Freq = 48,000 ÷ 2 = 24 kHz. Nyquist theorem: 48 kHz sample rate captures up to 24 kHz frequency, beyond typical human hearing (20 kHz).
Practical Applications
Audio mixing uses dB-scale faders: a -6 dB reduction cuts power to 1/4 (20 × log10(0.5) = -6 dB). A -20 dB reduction is essentially silence (1/100 power). Understanding dB enables quick mixing adjustments: want an instrument 2× quieter? Reduce by 6 dB (3 dB per doubling/halving of power).
Streaming audio compression selects bitrate based on bandwidth and quality requirements. CD-quality MP3 is typically 192-256 kbps; lo-fi streaming is 64-128 kbps. Converting bitrate calculations: 256 kbps MP3 ÷ 44.1 kHz sample rate = ~5.8 bits per sample (lossy compression; CD uses 16 bits uncompressed).
Audio interfaces and microphones use dB SPL (sound pressure level) relative to threshold of hearing. 0 dB SPL = 20 micropascals (hearing threshold), 130 dB SPL = hearing damage threshold. A microphone specifying "maximum 130 dB SPL" cannot record louder sounds without distortion.
Best Practices
💡 Pro Tip: Reference Levels
Set your mixing reference level (typically -18 to -23 dBFS on digital meters) and monitor at constant volume (85 dB SPL in control room) for consistent mixing. Mix at moderate levels—prolonged exposure to loud audio fatigues hearing and impairs mixing decisions.
- Use headroom: Keep peak levels 6-12 dB below 0 dBFS to avoid digital clipping and allow mastering headroom.
- Understand reference levels: -3 dBFS on a mixer ≠ -3 dBFS after loudness compression in mastering.
- Check audio file format: 44.1 kHz is standard for CD/streaming; higher rates (96 kHz+) for archival only if source material justifies it.
- Optimize streaming bitrate: 128 kbps for speech, 192-256 kbps for music, 320 kbps for lossless approximation.
Common Mistakes
⚠️ Confusing dB Power vs. Voltage
dB formula depends on measurement type: power uses 10 × log10, voltage/amplitude uses 20 × log10 (due to P = V²/R). A 10× voltage increase is 20 dB (not 10 dB). Mixing faders are voltage-based; power amps are power-based. Confusing these introduces 6 dB errors.
Tools and Resources
- DAWs (Ableton, Pro Tools, Logic): Built-in metering shows dB levels; most use -3 dBFS mixing reference.
- Frequency Analyzers: Spectrum analyzers show audio content across frequency range; identify problematic resonances.
- Loudness Meters: LUFS (Loudness Units relative to Full Scale) measure perceived loudness; YouTube and Spotify specify target loudness.
Key Takeaways
- Decibel scale is logarithmic: 10 dB = 10× power change, 3 dB ≈ doubling/halving, 20 dB = 100× power
- dB power = 10 × log10(ratio); dB voltage = 20 × log10(ratio) (due to V² relationship)
- Nyquist theorem: sample rate ÷ 2 = maximum frequency captured (48 kHz → 24 kHz max)
- Bitrate depends on sample rate and bit depth; MP3 compression ratios 10-12:1 typical
- Mix at -18 to -23 dBFS reference level; leave 6-12 dB peak headroom for mastering
Ready to Convert?
Try our free converter for instant results.