Glossary of Music Production Terms: A to Z for the Home Studio

A short, practical glossary of the words you meet in a DAW, a plugin window or a mixing forum, so you can read a manual or a tutorial without stopping at every second term.

By Delia Marchbanks, in Room & Monitoring

Production vocabulary piles up fast. A single plugin manual can mention knee, lookahead, sidechain and true peak on the same page, and each word hides an idea that matters once you understand it. The entries below are short on purpose: one to three sentences, enough to get the idea and move on. Where a topic deserves a full explanation, the entry points to the guide that covers it.

Terms are grouped by letter. Numbers given here are common starting points, not rules.

A to C

A

AAX. The plugin format used by Pro Tools. A plugin has to exist in AAX to load there, even if the same tool is available in VST3 or AU for other hosts.

AD/DA converter. The circuit that turns an analog voltage into numbers (analog to digital) and back again (digital to analog). Every audio interface contains both.

Attack. On a compressor, the time it takes to start reducing gain after the signal crosses the threshold. Short attack clamps transients, longer attack lets the front of each hit through. The guide on attack and release settings covers how to set it by ear.

AU (Audio Units). The plugin format native to macOS, used by Logic and several other Mac hosts.

Automation. Recorded changes to a parameter over time: a fader ride, a send that opens on one word, a filter sweep. See the automation guide for modes and habits.

Aux send. A path that copies part of a channel's signal to another track, usually a reverb or delay return, so several channels can share one effect.

B

Bit depth. The number of bits used to describe each sample. Each bit adds roughly 6 dB of dynamic range, so 16-bit gives about 96 dB and 24-bit about 144 dB in theory. More in sample rate and bit depth.

Bounce. Rendering a track, a group or the whole mix to a new audio file. Also called export or render.

Buffer size. The number of samples the computer processes in one block. Small buffers mean low latency and more CPU strain; large buffers are steadier but slower to respond. The buffer and latency guide gives the math.

Bus. Any path where several signals are summed together: a drum bus, a vocal bus, the mix bus.

C

Clip gain. A volume change applied to an audio region before any plugin on the channel. Useful for evening out a performance before compression.

Clipping. What happens when a signal goes past the maximum level a system can represent. In fixed-point audio the waveform tops are cut flat, which sounds harsh.

Comping. Building one final take from the best parts of several takes, phrase by phrase.

Compressor. A processor that turns down the signal when it gets louder than a set threshold, reducing the distance between loud and quiet parts. Circuit families differ in feel; see the compressor types guide.

Condenser microphone. A mic whose capsule is a thin charged diaphragm close to a backplate. It needs power, usually 48 V phantom, and tends to capture more detail and more room than a dynamic mic.

Crest factor. The difference between the peak level and the average level of a signal, in decibels. A raw drum recording might show 18 dB or more; a heavily limited master might show 6 dB or less.

D to H

D

DAW. Digital audio workstation, the recording and mixing software at the center of a home studio.

dBFS. Decibels relative to full scale. 0 dBFS is the highest level a digital file can hold, so every real level reads as a negative number.

dBu. An analog voltage reference where 0 dBu equals 0.775 volts RMS. Professional line level is usually +4 dBu. The gain staging guide explains how dBu and dBFS relate.

De-esser. A compressor that reacts only to sibilant energy, usually somewhere between 4 and 10 kHz, to tame harsh "s" and "sh" sounds. See de-essing vocals.

Decay. The time a reverb or a sound takes to fade away. On reverbs it is often labeled RT60, the time for the tail to fall by 60 dB.

Dither. Low-level noise added when reducing bit depth, for example from 24-bit to 16-bit, so the rounding error becomes a smooth hiss instead of gritty distortion on quiet passages.

Dry and wet. Dry is the unprocessed signal, wet is the processed one. A mix knob blends the two inside a single plugin.

Dynamic microphone. A mic with a diaphragm attached to a coil moving in a magnetic field. It needs no power, handles loud sources and picks up less of an untreated room.

E

EQ (equalizer). A processor that raises or lowers chosen frequency ranges. Common shapes are bell, shelf, high-pass and low-pass.

Expander. The opposite of a compressor: it makes signals below the threshold quieter, increasing dynamic range. A gate is an extreme expander.

F

Fader. The volume control of a channel. In most DAWs unity is 0 dB on the fader, meaning no change.

FET. Field-effect transistor. FET compressors, the 1176 being the classic reference, are known for very fast attack and an aggressive, forward sound.

Frequency response. How evenly a device reproduces different frequencies. Usually shown as a graph from 20 Hz to 20 kHz with a tolerance such as plus or minus 3 dB.

G

Gain staging. Keeping levels sensible at each point in the chain, from preamp to plugins to mix bus, so nothing clips and nothing sits buried in noise.

Gate. A processor that mutes or strongly lowers a signal when it falls below a threshold, used to remove bleed between drum hits or noise between phrases.

H

Haas effect. When two identical sounds arrive within roughly 30 ms of each other, the ear hears one sound coming from the side that arrived first. It is the basis of many widening tricks, and a common cause of mono problems.

Headroom. The space between the current peak level and the clipping point. A mix peaking at -6 dBFS has 6 dB of headroom.

High-pass filter. A filter that removes content below a chosen frequency and lets the rest pass. A high-pass at 80 to 100 Hz on a vocal clears rumble and handling noise.

I to M

I

Impulse response (IR). A recording of how a space or device responds to a very short click. Convolution reverbs and cabinet simulators load IRs to reproduce that space or speaker.

Insert. A plugin slot that processes the whole signal of a channel in series, as opposed to a send. EQ and compression usually go on inserts.

Integrated loudness. The average loudness of a whole song or program, in LUFS, with silent parts gated out.

K

Knee. How gradually a compressor moves from no compression to full ratio around the threshold. A hard knee switches abruptly, a soft knee eases in over several decibels.

L

Latency. The delay between a sound entering the system and coming back out of your speakers or headphones. Below about 10 ms round trip, most players stop noticing it.

Limiter. A compressor with a very high ratio and fast response that stops peaks from crossing a ceiling. It is usually the last processor on a master. See limiters and true peak.

Lookahead. A short delay inside a processor that lets its detector see a peak before it arrives, so it can react without overshoot.

Low-pass filter. The mirror of a high-pass: it removes content above a chosen frequency.

LUFS. Loudness Units relative to Full Scale, a loudness scale weighted to match how the ear perceives level. Streaming services normalize playback to a target, often around -14 LUFS integrated.

M

Mastering. The final stage before release: balancing tone and loudness, checking for problems, sequencing an album and preparing delivery files. The home mastering guide walks through it.

MIDI. A protocol that sends performance data (note, velocity, timing, controller moves) rather than sound. Drum programming is mostly MIDI; see MIDI drums vs audio samples.

Mid-side (M/S). A way of splitting a stereo signal into what both channels share (mid) and what differs between them (side), so each can be processed apart. Covered in mid-side processing.

Mono compatibility. How well a stereo mix holds up when both channels are summed to one. Phase problems show up as thin or missing parts in mono.

N to R

N

Noise floor. The level of background hiss and hum in a system when no signal is present. Good modern interfaces sit well below -100 dBFS.

Normalization. Raising or lowering a file so its peak or loudness hits a target. Streaming platforms normalize by loudness, which is why extra loudness at mastering earns less than many expect.

Nyquist frequency. Half the sample rate, the highest frequency a given rate can represent. At 48 kHz it is 24 kHz.

O

Optical compressor. A compressor that uses a light source and a light-sensitive cell to control gain. The LA-2A is the usual reference. Response is smooth and program dependent.

Oversampling. Running a plugin internally at a multiple of the session rate, such as 2x or 4x, to reduce aliasing from distortion or fast limiting, at the cost of CPU.

P

Pan law. The level drop applied to a signal panned to center so it does not seem louder than when panned to one side. Common values are -3 dB and -4.5 dB.

Parallel compression. Blending a heavily compressed copy of a signal with the untouched original. The parallel compression guide covers blend and alignment.

Phantom power. A DC voltage, normally 48 V, sent from a preamp to a condenser mic along the same cable that carries the audio.

Phase. The timing position of a waveform within its cycle. When two mics capture the same source at different distances, phase differences can cancel some frequencies.

Plugin delay compensation (PDC). A DAW feature that delays other tracks to match plugins that add latency, so everything stays in time.

Pre-delay. The gap between the dry sound and the start of a reverb. 20 to 40 ms keeps a vocal clear in front of its reverb. More in reverb types explained.

Proximity effect. The bass boost directional mics produce when the source gets very close, often several decibels below 200 Hz at a few centimeters. See home studio microphones.

Q

Q. The width of an EQ band. A low Q affects a wide range, a high Q a narrow one. A Q around 1 is broad and musical; above 8 is surgical.

Quantize. Snapping notes or audio events to a grid. Partial quantize moves them only part of the way to keep some feel.

R

Ratio. How strongly a compressor reduces level above the threshold. At 4:1, a signal 8 dB over the threshold comes out 2 dB over.

Release. The time a compressor takes to stop reducing gain once the signal falls back below threshold.

Reverb. The dense pattern of reflections that gives a sense of space. Plate, room, hall, chamber and spring are the classic families.

Room modes. Resonances at frequencies where a room's dimensions match half wavelengths of sound. They cause uneven bass, with peaks and nulls that change as you move. See practical acoustics.

Round robin. Cycling through several recordings of the same drum hit so repeated notes do not sound identical.

S to Z

S

Sample rate. The number of samples per second used to describe audio. 44.1 kHz and 48 kHz are the everyday choices.

Saturation. Gentle harmonic distortion, imitating tape, tubes or transformers, used to add density and help sounds read on small speakers.

Send. See aux send.

Sibilance. The hissing energy of consonants like "s", "sh" and "z", concentrated in the upper midrange and treble.

Sidechain. Feeding a compressor's detector from a different signal than the one it processes, for example letting the kick turn down the bass. See sidechain compression.

Stem. A submix of related parts (drums, bass, vocals, music) bounced to its own file. Stems are used for mastering, remixing and live playback; the remix guide explains how to approach them.

Stereo width. How spread a mix sounds between the speakers. Panning sets most of it; the stereo width guide covers the rest.

T

Threshold. The level above which a compressor, limiter or de-esser starts working, or below which a gate or expander does.

Transient. The short, sharp start of a sound, such as a stick hitting a snare or a pick hitting a string.

True peak. An estimate of the highest level the waveform reaches between samples once it is converted back to analog. A limiter ceiling of -1 dBTP leaves room for this and for lossy encoding.

V

VCA. Voltage-controlled amplifier, the gain element in many clean, precise compressors, including the classic console bus compressor.

Velocity. In MIDI, how hard a note is played, from 1 to 127. Drum instruments map velocity to volume and to different sample layers.

VST3. The current version of the most widespread plugin format on Windows and macOS.

W

Warping. Stretching or squeezing audio in time without changing pitch, to fit a new tempo or tighten a performance.

Wet. See dry and wet.

Z

Zero crossing. The point where a waveform passes through zero amplitude. Cutting and joining audio at zero crossings avoids clicks; short fades do the same job more reliably.

Quick answers

What is the difference between dBFS and LUFS?

dBFS describes level against the digital ceiling, sample by sample, and is what peak meters show. LUFS describes perceived loudness averaged over time, and is what streaming normalization uses.

Is a limiter the same as a compressor?

A limiter is a compressor set to an extreme: very high ratio, fast attack and a fixed ceiling. It is meant to catch peaks rather than shape dynamics musically.

Which terms matter most for a first mix?

Gain staging, EQ, compression basics (threshold, ratio, attack, release), panning and reverb sends cover most of a first mix. The rest can be learned as each problem appears.

Why do some plugins offer oversampling?

Distortion and fast gain changes create harmonics that can fold back below the Nyquist frequency as aliasing. Processing at a higher internal rate pushes those artifacts out of the audible range.