Reference

Audio Engineering Glossary

21 terms covering mastering, loudness normalization, streaming standards, and DJ mix delivery.

Bit Depth#: The number of bits used to represent each audio sample. Higher bit depth (e.g., 24-bit vs. 16-bit) provides greater dynamic range and lower noise floor. Streaming delivery typically uses 16-bit, while production and processing are done at 24-bit or 32-bit float.

ClippingAudio Clipping#: Distortion that occurs when an audio signal exceeds the maximum level a system can represent (0 dBFS in digital audio). Clipping flattens the peaks of the waveform, introducing harsh harmonic distortion. It is irreversible once written to a file.
CompressionDynamic Range Compression#: A process that reduces the dynamic range of an audio signal by attenuating levels above a set threshold. A compressor makes loud parts quieter relative to quiet parts, resulting in a more consistent overall level. Compression affects the shape of the waveform over time but does not set a specific loudness target.

dBFSDecibels Full Scale#: The amplitude unit used in digital audio. 0 dBFS is the maximum level a digital system can represent without clipping. All levels in digital audio are expressed as negative numbers below 0 dBFS.
dBTPDecibels True Peak#: A measurement of the maximum true peak level of a digital audio signal, accounting for inter-sample peaks that occur during digital-to-analog conversion. Streaming platforms specify a maximum true peak (typically −1 dBTP) to prevent clipping during playback on consumer devices.
Dynamic Range#: The difference between the quietest and loudest parts of an audio signal, measured in decibels (dB). Higher dynamic range means greater contrast between soft and loud moments. Compression reduces dynamic range; loudness normalization preserves it while adjusting the overall level.

Integrated Loudness#: The average loudness of an audio file measured over its entire duration, expressed in LUFS. This is the primary loudness metric used by streaming platforms. It is defined in ITU-R BS.1770 and uses a gating algorithm that excludes near-silence passages from the average.
Inter-Sample Peaks#: Peaks in the reconstructed analog waveform that exceed the level of the digital samples surrounding them. Because digital audio is reconstructed by interpolating between discrete samples, the actual analog output can be louder than the highest digital sample value. This is why true peak metering (dBTP) gives a more accurate representation of real-world peak levels than simple sample peak metering (dBFS).
ITU-R BS.1770ITU-R BS.1770 Loudness Standard#: An international standard published by the International Telecommunication Union that defines the algorithm for measuring audio loudness. It specifies a frequency-weighted (K-weighted) filter that models human perception of loudness, a gating algorithm that excludes quiet passages, and the LUFS unit of measurement. All major streaming platforms base their loudness targets on this standard.

K-WeightingK-Weighted Filter#: A frequency weighting filter defined in ITU-R BS.1770 that approximates the frequency sensitivity of human hearing. It applies a pre-filter that emphasizes high frequencies (above 1 kHz) and de-emphasizes bass, reflecting the fact that humans perceive mid and high frequencies as louder than low frequencies at the same amplitude.

LimiterPeak Limiter#: A compressor with an extremely high ratio (effectively infinite:1) that prevents the signal from exceeding a set ceiling. A true peak limiter also accounts for inter-sample peaks. Limiters are the final step in a mastering chain, ensuring the output does not exceed the maximum allowable level.
LRALoudness Range#: A measure of the statistical variation in loudness throughout an audio file, expressed in LU (Loudness Units). Low LRA indicates a highly compressed, consistent signal. High LRA indicates wide dynamic variation. Streaming platforms do not typically impose an LRA requirement, but it is used in mastering to characterize the feel of a mix.
LUFSLoudness Units Full Scale#: The unit used to measure perceived loudness of audio as defined by ITU-R BS.1770. Numerically equivalent to LKFS (Loudness, K-weighted, relative to Full Scale). Streaming platforms normalize content to a target LUFS value — typically between −14 LUFS and −16 LUFS integrated — so all tracks play back at a consistent perceived volume.

MasteringAudio Mastering#: The final stage of audio post-production in which a mix is processed for consistent playback across different systems and distribution platforms. For DJ mixes, mastering typically involves loudness normalization, true peak limiting, EQ corrections, and metadata tagging before delivery to streaming platforms.
Momentary Loudness#: Loudness measured over a very short window (400 milliseconds) defined in ITU-R BS.1770. It reflects instantaneous perceived loudness and is useful for detecting loud transient events in a mix, but is not used by streaming platforms for normalization decisions.

NormalizationAudio Normalization#: The process of adjusting the overall level of an audio file to meet a specific target. Simple peak normalization raises or lowers the gain so the loudest sample hits a target dBFS. Loudness normalization (used by streaming platforms) adjusts gain based on the integrated LUFS measurement so that perceived loudness matches a target.

RMSRoot Mean Square#: A mathematical method of measuring average signal power, commonly used in older metering standards to approximate loudness. RMS does not incorporate frequency weighting and is less accurate than LUFS for representing perceived loudness. Most modern mastering standards have moved to LUFS (ITU-R BS.1770) for loudness measurement.

Sample Rate#: The number of audio samples captured or played back per second, expressed in Hz or kHz. CD quality is 44,100 Hz (44.1 kHz). Streaming platforms typically accept 44.1 kHz or 48 kHz. Higher sample rates (88.2 kHz, 96 kHz) are used in professional production but are downsampled for distribution.
Short-Term Loudness#: Loudness measured over a 3-second sliding window defined in ITU-R BS.1770. Useful for monitoring the loudness of a specific section of a mix (such as a drop or breakdown) without the averaging effect of integrated loudness.

True Peak#: The maximum level of the audio signal as it would appear after digital-to-analog conversion, taking into account inter-sample peaks that exceed the digital sample values. True peak measurement is defined in ITU-R BS.1770-4. Streaming platforms require files to have a true peak ceiling of −1 dBTP or lower to prevent clipping on playback.

WaveformAudio Waveform#: A visual representation of an audio signal that shows how amplitude changes over time. The horizontal axis represents time; the vertical axis represents amplitude. Waveform visualizations are used in DAWs and audio editing software to inspect the structure, dynamics, and loudness of a recording.