Glossary

110 terms from the lessons, in plain words. Inside a lesson, tap a dotted word to read its definition without leaving the page.

#

32-bit float: A sample format that stores a scale factor with every value, so levels above 0 dBFS inside the DAW are kept instead of clipped. Overs still clip at a fixed-point export, at the interface output and inside any plugin that clips.; Explained in Where 32-bit float headroom ends

A

ADSR: Attack, decay, sustain, release: the four stages most synth envelopes use to shape a note over time.; Explained in Attack time changes how a part feels
Aliasing: A frequency above the Nyquist limit being stored as a false, lower frequency that was never in the source. Once it is recorded, it cannot be filtered back out.; Explained in Why aliasing is a ghost frequency problem
Attack time: How quickly a compressor clamps down once the signal crosses the threshold, or how quickly a synth note rises to full level. Usually set in milliseconds.
Attention residue: Thoughts about a previous task that linger after you switch, so part of your attention is still elsewhere.
Auditory scene analysis: How the brain sorts a mixture of sound into separate sources, such as a voice, a kick and a pad.

B

Backbeat: The snare or clap on beats 2 and 4 of a bar in 4/4, the accent most pop, rock and hip-hop grooves are built on.
Bit depth: How many bits store each sample. Each bit adds about 6 dB between full scale and the quantization noise floor: about 96 dB at 16-bit, about 144 dB at 24-bit.; Explained in Bit depth sets the noise floor
BPM: Beats per minute: the tempo of a track.
Buffer size: How many samples the audio interface handles in one go. Smaller buffers mean less latency and more work for the computer.; Explained in Latency changes how a take feels
Bus compression: A compressor on a group or the mix bus, so every part feeding it shares one gain movement.; Explained in What bus compression glue really is

C

Cadence: The chord move that ends a phrase. Ending on the tonic, as in V to I, sounds closed. Stopping on the dominant, a half cadence, sounds like a question.
Call and response: One part plays or sings a phrase and another answers it, like a vocal line followed by a guitar fill.
Clip gain: A gain change applied to a region of audio itself. In most DAWs it acts before the plugins on the track, so unlike the fader it changes what they hear.; Explained in Fix the loud word before the compressor hears it
Clipping: Cutting off the top of a waveform at a fixed level. Hard clipping flattens the peaks; soft clipping rounds them, which adds fewer harsh harmonics.; Explained in Clipping can work. It can also cost you
Comb filtering: The hollow, notched sound you get when a signal is mixed with a slightly delayed copy of itself. The cancellations fall at evenly spaced frequencies, like the teeth of a comb.; Explained in Room reflections EQ your vocal before you do
Comping: Building one final take from the best parts of several recorded takes.; Explained in Comp vocals in phrases, not syllables
Compressor: A processor that turns a signal down once it goes above a threshold. The ratio sets how much: at 4:1, every 4 dB over the threshold comes out as 1 dB over.; Explained in Compression changes motion before level
Convolution: Combining a sound with an impulse response so it takes on the character of that room or device.; Explained in Convolution reverb copies a space
Correlation meter: A meter that shows how alike the left and right channels are, from +1 (identical) through 0 (unrelated) to -1 (one is the other flipped). Readings near -1 warn that the mix will lose level in mono.; Explained in What a correlation meter tells you
Crest factor: The gap between a signal's peak level and its average level. Heavy limiting makes it smaller.

D

dBFS: Decibels relative to full scale, the largest value a fixed-point file or converter can hold, so levels below it are negative numbers. Inside a floating-point mix a signal can read above 0 dBFS and stay intact; it clips where it leaves float or passes through a plugin that clips. dBTP is the same scale measured as true peak, and it can read above 0.
De-esser: A compressor that only reacts to sibilant frequencies, so harsh s sounds come down while the rest of the vocal stays put.
Deceptive cadence: A phrase that sets up a return to the tonic and lands somewhere else instead, most often V to vi.
Detune: Setting copies of an oscillator slightly off pitch from each other so they drift against each other and sound wider.
Dither: Very low-level noise added before reducing bit depth. It turns quantization error, which can sound like gritty distortion on quiet sounds, into a steady, smooth hiss.; Explained in Bit depth sets the noise floor
Double-tracking: Recording the same part twice and playing both. Small differences in timing and pitch make it sound thicker.; Explained in Why a vocal double works best unnoticed
Downbeat: The first beat of a bar.
Dynamic EQ: An EQ band that cuts or boosts only while the level in its band, or in a key signal, passes a threshold, and stays flat the rest of the time.; Explained in Dynamic EQ or multiband: match the problem
Dynamic range: The distance between the quiet and loud parts of a signal or a song.

E

Ear fatigue: Hearing that has adapted after a long or loud session, so judgments about level and brightness drift.; Explained in Fresh ears are a production tool
Early reflections: The first echoes from nearby surfaces, arriving in roughly the first 80 ms after the direct sound. They shape clarity and apparent width more than the sense of room size.; Explained in Early reflections place a sound, the tail sets the room
EQ: Equalization: turning chosen frequency ranges up or down.; Explained in EQ is attention design
Equal-loudness contours: Curves showing the level each frequency needs to sound as loud as a 1 kHz tone (ISO 226). They are steepest in the bass at low levels, so quiet monitoring hides low end.; Explained in Monitoring level changes the balance you hear
Exclusive license: A license that stops the producer from licensing the same work to anyone else from then on. What it covers, including earlier licenses, depends on the contract.

F

FFT: Fast Fourier Transform. The algorithm analyzers use to split a short block of audio into frequencies. Longer blocks show frequency more precisely and timing less precisely.; Explained in Fourier shows the ingredients
Formant: Resonant peaks shaped by the throat and mouth that make a voice sound like itself. Shifting them changes how large the voice seems.
Forward masking: A loud sound making a quieter sound that follows it harder to hear for a short time afterwards, up to roughly a couple of hundred milliseconds.
Fourier transform: Maths that describes any sound as a sum of sine waves, each with its own frequency, level and phase.; Explained in Fourier shows the ingredients
Fundamental: The lowest frequency of a pitched sound. It sets the note you hear.

G

Gain staging: Setting the level at each point in the signal chain so nothing clips and nothing sits so low that noise becomes audible.
Ghost note: A very quiet hit, often on snare or hi-hat, felt more than heard. It fills the groove between the main hits.
Group delay: How long a filter holds back each frequency, in milliseconds. Steep high-pass filters delay the lowest frequencies the most.; Explained in Filters are shape machines
Guide tones: The third and seventh of a chord, the notes that define its quality. Moving them by small steps carries a progression.

H

Haas effect: When the same sound reaches you from two places less than about 30 ms apart, you hear one sound, placed toward the one that arrived first. Producers delay one side by a few milliseconds to widen a part, which can turn hollow when summed to mono.
Habituation: The brain responding less and less to a sound that repeats without change.; Explained in Why contrast makes the chorus feel big
Half-time: A feel where the snare lands once per bar, on beat 3, so a fast tempo feels half as fast. Trap at 140 BPM is usually felt at 70.
Harmonics: Frequencies above a note's fundamental. In pitched sounds they sit at whole-number multiples of it, and their balance is a large part of timbre.
Headroom: The space between the loudest peak in a signal and the level where it would clip.
High-pass filter: A filter that removes energy below a cutoff frequency and lets everything above it through.; Explained in Filters are shape machines

I

Impulse response: A recording of how a room or device answers one very short click. Convolution uses it to place any sound in that space.; Explained in Convolution reverb copies a space
Inter-sample peak: A peak in the reconstructed waveform that falls between two samples, so the highest sample value under-reads it. True-peak meters find it by oversampling.; Explained in True peak bites after encoding
Intermodulation: Distortion that creates new frequencies at sums and differences of the tones going in, such as 1160 and 840 Hz from 1 kHz plus 80 Hz. Unlike harmonics, they are not related to the notes.
Inversion: A chord with a note other than its root in the bass, such as C major with E at the bottom, written C/E.

L

Latency: The delay between doing something, like pressing a key, and hearing the result.; Explained in Latency changes how a take feels
Limiter: A very fast compressor with a very high ratio. It stops peaks from going over a set ceiling.
Linear phase: A filter that delays every frequency by the same amount, so the waveform shape is kept. The cost is latency and a little ringing before sharp hits.; Explained in Filters are shape machines
Loudness normalization: A streaming service turning each track up or down so everything plays back at a similar loudness. A louder master is simply turned down further.; Explained in Streaming loudness myths waste masters
Loudness range: LRA, in LU: the spread between the quieter and louder parts of a programme, taken from the 10th to the 95th percentile of its gated short-term loudness (EBU Tech 3342). It tracks section-to-section change, not punch.; Explained in Loudness, PLR and LRA answer different questions
Low-pass filter: A filter that removes energy above a cutoff frequency and lets everything below it through.; Explained in Filters are shape machines
LU: Loudness Unit: a difference in loudness on the LUFS scale. 1 LU is the same step as 1 dB.
LUFS: Loudness Units relative to Full Scale. A loudness measurement weighted toward how the ear hears. Streaming services use integrated LUFS, an average over the whole track, to match playback level.; Explained in Stop treating LUFS like a target

M

Masking: One sound making another harder to hear, usually because they share frequencies at the same moment. A loud sound also masks what follows it for a short time (forward masking). The quieter sound is still there; the ear stops separating it.; Explained in Mud is often masking, not dirt
Mere exposure effect: People tend to like something more the more often they have encountered it, even without noticing. It is one reason a demo you have looped for weeks can sound right when it is not.
Microtiming: Small timing offsets, from a few to a few tens of milliseconds, that place notes ahead of or behind the grid.; Explained in Why tiny timing differences create human feel
Mid/side: A second way to describe a stereo signal. Mid is what left and right share, side is how they differ. Raising the side widens the image; mid alone is what a mono fold keeps.; Explained in Mid/side widening moves the centre too
Mono compatibility: How well a mix holds up when left and right are summed into one channel, as on many phone speakers, club systems and smart speakers.; Explained in Mono reveals what stereo hides
Multiband compression: Splitting a signal into frequency bands with crossover filters and compressing each band on its own.; Explained in Dynamic EQ or multiband: match the problem

N

Noise floor: The level of background noise in a system. Anything quieter than it is lost in the hiss.
Non-exclusive license: Permission to use a beat or track within set limits, while the producer keeps ownership and can license the same work to other people.
Nyquist frequency: Half the sample rate. It is the highest frequency that sample rate can store correctly.; Explained in Why aliasing is a ghost frequency problem

O

Oscillator: The part of a synth that generates the raw waveform, such as a sine, saw or square.
Oversampling: Running a process at a multiple of the session sample rate, then filtering and converting back. It gives distortion and clipping room to create high harmonics without folding them back as aliasing.

P

P-centre: The moment a syllable or sound is heard to land. For a sung syllable it sits near the start of the vowel, not at its first consonant.
Parallel compression: Blending a heavily compressed copy of a signal under the untouched original. The quiet detail comes up while the loud peaks keep most of their shape.; Explained in Parallel compression turns the quiet parts up
Peak to loudness ratio: The gap between a master's true peak and its integrated loudness. Heavy limiting makes it smaller. Plain gain normalization moves peak and loudness by the same amount, so PLR does not change: at a -14 LUFS target it is how far the loudest peak sits above -14. One stray peak can make it high on a flat master.; Explained in Loudness, PLR and LRA answer different questions
Peak to short-term loudness ratio: The gap between the highest true peak in a 3-second window and the short-term loudness of that window. It follows how far the peaks stand above the music moment by moment.; Explained in Loudness, PLR and LRA answer different questions
Phase: Where a wave is in its cycle at a given moment. Two copies of a sound that are out of phase partly cancel where one rises while the other falls.; Explained in Phase is timing with consequences
Polarity: Whether a waveform is upright or flipped. Flipping polarity inverts the wave at every frequency at once. It is not the same as a time delay.; Explained in Phase and polarity are not the same fix
Pre-chorus: The section that lifts the verse into the chorus. Its job is to build pressure the chorus can release.; Explained in A pre-chorus builds pressure by holding back
Pre-delay: The gap between the dry sound and the start of its reverb. More pre-delay keeps the source clear and in front.; Explained in Reverb moves emotion forward or back
Prediction error: The gap between what the brain expected to hear and what arrived. Small surprises that make sense in hindsight tend to feel rewarding.; Explained in Why expectation drives musical emotion
Prosody: The rhythm, stress and pitch of speech. In a song, good prosody puts musical stress where the words are naturally stressed.
Proximity effect: The bass boost a directional microphone adds as the source moves closer to it.; Explained in How proximity effect turns warmth into mud

Q

Q: How narrow a filter band or resonance is: centre frequency divided by bandwidth. Higher Q is narrower and rings longer.; Explained in Resonance can sing or wreck the mix
Quantize: Snapping recorded notes to the nearest grid position. Partial quantize moves them only part of the way.

R

Reference track: A finished record you compare your mix against to check balance, tone and loudness on the same speakers.; Explained in Reference tracks are for calibration, not imitation
Relative key: A major key and a minor key that share the same notes, such as C major and A minor.
Release time: How quickly a compressor lets go after the signal falls back below the threshold.
Resonance: A frequency where a system rings or builds up energy. On a filter, the resonance control boosts a narrow band around the cutoff.; Explained in Resonance can sing or wreck the mix
Room mode: A low frequency that fits a whole number of half-wavelengths between two walls, so it piles up in some spots and nearly vanishes in others. The first axial mode is about 343 / (2 × length) Hz.; Explained in Why your low end lies in a small room

S

Sample rate: How many times per second a digital system measures the waveform. 48 kHz means 48,000 measurements a second.; Explained in Sampling is taking photos of air
Saturation: Gentle distortion that adds harmonics and rounds off peaks, often modelled on tape, tubes or transformers.
Scale degree: A note's position in the key, counted from the tonic: in C major, C is 1, D is 2 and G is 5.
Short-term loudness: Loudness measured over a sliding 3-second window, useful for comparing sections. Momentary loudness uses 400 ms.; Explained in Stop treating LUFS like a target
Sibilance: The hiss of s, sh and similar sounds in a vocal, usually strongest somewhere between about 5 and 10 kHz.
Sidechain: Letting one signal control a processor on another, for example the kick triggering a compressor on the bass so the two take turns.; Explained in Sidechain lets one track control another
Spectrum analyzer: A meter that shows how much energy sits at each frequency, usually on a log scale from 20 Hz to 20 kHz.; Explained in Fourier shows the ingredients
Stems: Grouped exports of a song, such as drums, bass, music and vocals, used for mixing, mastering, remixing or licensing.
Swing: Delaying every second subdivision so the rhythm moves long-short instead of evenly. In most DAWs 50% is straight and about 66% is a triplet feel.; Explained in Swing explained without mystical language
Syncopation: Accenting a weak beat or the space between beats, so the rhythm pulls against the pulse.; Explained in Why syncopation wakes up the listener

T

Timbre: What lets you tell two sounds apart at the same pitch and level: the balance of harmonics and noise, and how they change over time.; Explained in Timbre tells the brain what arrived
Tonic: The home note of a key, and the chord built on it. A phrase that ends on it sounds finished.
Track delay: A per-track setting that plays a whole track a few milliseconds early or late without moving its notes on the grid.
Transient: The short, sharp start of a sound, like the crack of a snare or a pick hitting a string. It carries a lot of how hard a sound feels and how quickly the ear recognizes it.
Transient shaper: A processor that turns the start of each sound up or down, and its sustain up or down, by comparing a fast and a slow level follower. Many have no threshold, so quiet and loud hits are shaped alike.; Explained in Transient shaper or compressor for punch
True peak: The highest level a waveform reaches between samples once it is converted to analog or decoded from a lossy file. It can sit above the highest sample value, so meters estimate it by oversampling.; Explained in True peak bites after encoding

V

Varispeed: Changing playback speed so tempo and pitch move together, like tape running faster or slower. About 6% faster is one semitone higher.
Voice leading: How each note of one chord moves to the next. Small steps and shared notes make chord changes sound smooth.

W

Window function: A taper applied to each block of audio before an FFT, so the block edges do not smear energy into neighbouring frequencies. That smear is called spectral leakage.; Explained in How to read a spectrum analyzer