Tell me about sound recording, and the clearest starting point is this: sound recording is the process of turning changing air pressure into a stored representation that can later recreate the audible pattern. A microphone converts acoustic vibration into an electrical signal; analogue or digital equipment shapes and stores that signal; a playback system converts the stored representation back into electrical energy; and a loudspeaker turns the electrical signal back into pressure waves. The complete system therefore moves through four domains—acoustic, electrical, stored representation and acoustic again. Understanding sound recording means following what changes at every handoff without losing sight of the original pressure pattern.

People searching for how sound recording works usually want to know how microphones capture sound, what preamps do, why gain matters, how analogue tape differs from digital recording, what sample rate and bit depth mean, why clipping sounds harsh, what mono and stereo are, how equalisation and compression work, why rooms change a recording, what mixing and mastering actually do, why some recordings sound noisy or distorted, and how files such as WAV, FLAC, MP3 and AAC differ. These questions are connected by one central idea: every recording system must preserve enough information about amplitude, timing and frequency content for playback to reconstruct a convincing version of the original sound.

This guide explains sound recording from first principles. It covers acoustics, microphones, transducers, preamplifiers, gain staging, analogue recording, tape, digital conversion, sampling, quantisation, dither, clocks, latency, multitrack recording, stereo, room acoustics, signal flow, equalisation, compression, reverb, editing, mixing, mastering, audio file formats, loudness, noise, distortion, troubleshooting and practical applications. It also includes worked examples, misconceptions and diagnostics. The aim is not to turn every reader into an audio engineer, but to build a model strong enough to explain why recordings sound the way they do and how the entire chain from performance to playback fits together.

The 50-second explanation

A sound source vibrates and creates pressure changes in air. A microphone diaphragm moves with those pressure changes and converts the motion into an electrical signal whose voltage changes over time. That small signal is amplified by a preamp. In an analogue system, the changing voltage may be stored directly as magnetisation on tape or motion in a record groove. In a digital system, an analogue-to-digital converter measures the signal many times per second and represents each measurement with numbers.

Those numbers can be stored, copied, edited and processed without the cumulative generation loss associated with many analogue copying steps. During playback, a digital-to-analogue converter turns the numbers back into a changing voltage. An amplifier drives a loudspeaker, whose cone pushes and pulls the air to recreate pressure variations. The listener hears a reconstructed acoustic waveform, not the original sound travelling through time.

Every stage can alter the result. The microphone changes tonal balance and directionality. The room adds reflections. The preamp can add noise or distortion. Poor gain staging can clip. Digital conversion has finite resolution. Compression and equalisation reshape dynamics and frequency balance. Loudspeakers and headphones colour playback. A recording is therefore the outcome of a chain, not a single device.

Core definitions

Sound wave

A sound wave is a pattern of pressure variation travelling through a medium such as air. Frequency describes how rapidly the pressure oscillates and strongly relates to perceived pitch. Amplitude describes the size of the pressure change and strongly relates to perceived loudness. Complex sounds contain many frequencies and time-varying components at once.

Signal

In audio, a signal is a representation of sound that changes over time. It may be acoustic pressure, electrical voltage, magnetic flux, groove displacement or digital numbers. Engineers care about signal integrity: how accurately one form preserves the information needed to create the next.

Transducer

A transducer converts energy from one form to another. A microphone converts acoustic energy into electrical energy. A loudspeaker performs the reverse conversion. Transducers are therefore the physical gateways between sound in air and signals in equipment.

Gain

Gain is amplification applied to a signal. It changes electrical level before later processing or recording. Setting gain correctly matters because signals that are too small approach the noise floor, while signals that are too large overload circuits or digital converters.

Decibel

The decibel is a logarithmic unit used to express ratios of level. Audio uses several reference systems, including sound-pressure level in air and digital full-scale level inside digital systems. Because decibels are logarithmic, a small-looking numerical change can represent a substantial change in power or amplitude.

How microphones turn sound into electricity

A microphone begins with a diaphragm that moves when sound pressure changes. The mechanical motion is then converted into an electrical signal. Different microphone designs use different physical mechanisms, which affects sensitivity, frequency response, transient behavior, self-noise and practical use.

Dynamic microphones

A moving-coil dynamic microphone attaches a coil of wire to a diaphragm and places it in a magnetic field. When the diaphragm moves, the coil moves through the field and generates a voltage by electromagnetic induction. Dynamic microphones are robust, tolerate high sound pressure and do not require external power for the transducer itself.

Condenser microphones

A condenser microphone uses a diaphragm and backplate as a capacitor. Sound changes the distance between them, changing capacitance and producing a corresponding electrical variation. The capsule requires polarising or electret charge and active electronics, so studio condensers commonly use phantom power. They often offer high sensitivity and detailed high-frequency response.

Ribbon microphones

A ribbon microphone uses a very thin conductive ribbon suspended in a magnetic field. Air motion moves the ribbon and induces voltage. Classic ribbons often produce a smooth response and figure-eight pickup pattern, but traditional designs can be mechanically delicate.

For a deeper microphone-specific route, see Tell Me About Microphones.

Polar patterns and microphone placement

Microphones do not always capture sound equally from every direction. An omnidirectional microphone responds broadly around itself. A cardioid microphone is most sensitive to the front and rejects more sound from the rear. Figure-eight microphones respond to front and back while rejecting the sides. Other patterns sit between these families.

Directionality changes the ratio between direct sound and room sound. A cardioid microphone placed close to a singer captures more direct voice and less room reflection than an omnidirectional microphone placed several metres away. This means microphone placement is part of the recording, not merely a logistical decision.

Distance also changes tonal balance. Many directional microphones exhibit proximity effect, increasing low-frequency response when placed close to the source. Moving a microphone by a few centimetres can therefore alter bass response, attack, ambience and the balance among different parts of an instrument.

Room acoustics: the sound reaches the microphone before the equipment begins

A microphone records not only the source but also the room. Sound travels directly from the source to the microphone and also reflects from walls, ceiling, floor and objects. Early reflections combine with the direct signal and can cause comb filtering. Later reflections blend into reverberation.

A small untreated room can create strong low-frequency resonances called room modes. Some bass notes become exaggerated while others nearly disappear at particular positions. Moving the microphone or listener can change the result dramatically because the interference pattern changes spatially.

Acoustic treatment uses absorption and diffusion to control reflections. Absorption removes some acoustic energy, while diffusion scatters reflections in many directions. Treatment is different from soundproofing: treatment changes sound within a room; soundproofing reduces transmission between rooms.

Preamplifiers and gain staging

Microphone signals are usually small, so a preamplifier raises them to a level suitable for conversion or further processing. The preamp itself contributes noise and has a maximum level before distortion. Good gain staging keeps the wanted signal comfortably above circuit noise while preserving headroom for unexpected peaks.

In digital recording, there is no benefit in recording every source almost at 0 dBFS. Modern converters have enough dynamic range to leave healthy headroom. Peaks that exceed digital full scale clip because there is no representable value above the maximum. Once a waveform is clipped during conversion, turning it down later does not restore the missing shape.

Gain staging continues through the entire chain. Plugins, buses and analogue processors can overload even when the final output meter looks safe. The useful habit is to watch levels at each stage rather than only at the end.

Analogue recording

Magnetic tape

Analogue tape stores sound by magnetising particles on a moving strip in patterns corresponding to the electrical audio signal. Playback heads sense those magnetic variations and convert them back into voltage. Tape speed, track width, formulation and machine alignment all affect frequency response, noise and distortion.

Tape has a finite noise floor and gradually saturates as level rises. Mild saturation can compress peaks and add harmonics in a way some listeners find musically pleasing, while excessive saturation causes obvious distortion. Repeated analogue copying adds generation loss because every pass introduces more noise and response error.

Groove recording

A phonograph record stores audio as microscopic variations in a spiral groove. During playback, a stylus follows the groove and converts motion into an electrical signal. Physical constraints on groove spacing, bass excursion and high-frequency velocity influence mastering decisions for vinyl.

Digital recording: turning voltage into numbers

An analogue-to-digital converter measures the incoming voltage at discrete times and assigns each measurement a numerical value. Two key settings describe this process: sample rate and bit depth. The resulting stream of numbers can be stored in memory or files.

Digital audio does not store a staircase-shaped sound that remains jagged at playback. When sampling rules are obeyed and proper reconstruction filtering is used, the converter can reconstruct a continuous band-limited waveform from the samples. The samples are measurements that define the waveform, not tiny rectangular pieces of sound.

Converter quality still matters. Analogue input stages, anti-alias filters, clocking, linearity and noise affect performance. But modern professional and consumer converters can be extremely accurate when used within their design limits.

Sample rate and the Nyquist principle

Sample rate is the number of measurements taken per second. A rate of 48 kHz means 48,000 samples each second. According to the sampling theorem, a band-limited signal can be reconstructed when sampled at more than twice its highest frequency. The halfway point is called the Nyquist frequency.

This is why common audio sample rates are comfortably above twice the upper range of typical human hearing. Anti-alias filters prevent frequencies above the supported range from folding into false lower frequencies during conversion.

Higher sample rates can simplify some filtering or processing and may be useful in specialised production, but they also increase storage and processing load. They do not automatically make every recording audibly better. Microphone quality, room acoustics, performance and monitoring usually matter more.

Bit depth, quantisation and dynamic range

Bit depth determines how many discrete amplitude values are available for each sample. A 16-bit system has 65,536 possible values; a 24-bit system has far more. More bit depth lowers the theoretical quantisation noise floor and provides greater practical headroom during recording.

Quantisation is the rounding involved when a continuous voltage is represented by one of the available numerical values. With proper dither, quantisation error behaves like very low-level noise rather than correlated distortion. This lets digital systems represent signals below the least-significant undithered step statistically over time.

Twenty-four-bit recording is useful in production because engineers can record conservative levels without approaching the system noise floor. Final consumer formats may use lower bit depths after mixing and mastering.

Clocks, jitter and synchronisation

Digital converters need timing references so samples are taken and reproduced at the intended intervals. The sampling clock determines when each measurement occurs. Timing error is called jitter. In well-designed modern systems, jitter is usually far below levels that dominate ordinary recording problems, but clocking becomes important when several digital devices exchange audio simultaneously.

If multiple converters operate from unrelated clocks, their sample timing drifts. Digital interfaces therefore designate one clock source or recover timing from an incoming stream. Incorrect synchronisation can produce clicks, dropouts, pitch changes or complete loss of lock. The engineering lesson is simple: digital audio depends on both the sample values and when those values are interpreted.

Multitrack recording

Multitrack recording stores different sources on separate tracks so they can be adjusted independently later. A drum kit may use many microphones, while voice, guitar and keyboards occupy other tracks. The recording system keeps all tracks aligned to the same timeline.

Separation creates flexibility. Engineers can change vocal level without changing drums, repair one performance section, apply different equalisation to each instrument and automate changes across the song. The trade-off is complexity: every extra microphone introduces phase relationships, routing decisions and more material to organise.

Digital audio workstations combine recording, editing, processing, automation and mixing on one timeline. They also allow non-destructive editing, where the software stores edit instructions without altering the original audio file.