Fact-checked by Grok 2 weeks ago

Audio signal

An audio signal is an electrical representation of sound waves, capturing variations in acoustic pressure as either an analog waveform—typically a continuous voltage or current that mirrors the sound's amplitude over time—or a digital sequence of numerical samples obtained through sampling and quantization processes.^[1]^[2] These signals generally occupy the frequency range audible to humans, from approximately 20 Hz to 20 kHz, though the exact upper limit can vary slightly with age and individual hearing sensitivity.^[3] In audio engineering and digital signal processing, audio signals serve as the core medium for recording, transmission, reproduction, and manipulation of sound, enabling applications such as music production, speech recognition, noise reduction, and audio compression.^[4]^[5] Key characteristics include amplitude (related to loudness, often measured in decibels), frequency content (analyzed via Fourier transforms for spectral properties), and duration, all of which are processed using techniques like filtering, equalization, and effects to enhance quality or achieve artistic goals.^[6] Analog audio signals are susceptible to noise and degradation during transmission, prompting the widespread adoption of digital formats for greater fidelity and storage efficiency in modern systems.^[7]

Fundamentals

Definition

An audio signal is a representation of sound, typically in the form of an electrical voltage or current that varies over time in analogy to the pressure changes of acoustic waves produced by sound.^[8]^[7] This representation carries information about sounds within the frequency range perceivable by humans, approximately 20 Hz to 20 kHz.^[9] Audio signals can also take mechanical forms, such as vibrations in a medium, but electrical forms predominate in modern applications due to their ease of processing and transmission.^[10] The origins of audio signals trace back to advancements in telephony and radio during the late 19th and early 20th centuries. A pivotal development was Alexander Graham Bell's invention of the telephone in 1876, which enabled the electrical transmission of speech by converting sound waves into varying electric currents through a diaphragm and electromagnet system.^[11] This breakthrough laid the foundation for audio signal technology, later extended to radio broadcasting in the early 1900s, where modulated electrical signals carried voice and music wirelessly.^[12] Unlike video signals, which require higher bandwidths—often several megahertz—to convey visual information with spatial and temporal detail, or data signals optimized for efficient digital information transfer regardless of perceptual fidelity, audio signals emphasize accurate reproduction of natural auditory experiences within limited bandwidth constraints.^[13]^[14] Common examples include the output from a microphone, which converts acoustic pressure into an electrical voltage proportional to the sound wave, and the input to a loudspeaker, which drives a diaphragm to recreate those pressure variations.^[15]^[10]

Audible Range

The audible range of human hearing defines the perceptual boundaries for audio signals, typically encompassing frequencies from 20 Hz to 20 kHz under normal conditions.^[16] This range represents the spectrum where sound waves are detectable by the human ear, with sensitivity peaking between 1 kHz and 4 kHz.^[17] However, individual variations exist, influenced by factors such as age; presbycusis, or age-related hearing loss, can begin as early as the thirties or forties and typically becomes more pronounced after age 60, progressively diminishing sensitivity to high frequencies and affecting speech discrimination and environmental sound perception.^[18] In terms of intensity, the dynamic range of human hearing spans approximately 120 dB, from the threshold of hearing at 0 dB sound pressure level (SPL), where the quietest detectable sounds occur, to the pain threshold between 120 and 140 dB SPL, beyond which exposure causes discomfort or damage.^[19]^[20] This vast range allows perception of subtle whispers to intense noises like jet engines, though the ear's response is nonlinear, with greater resolution in mid-frequencies and lower intensities. Signals outside these limits—infrasound below 20 Hz or ultrasound above 20 kHz—fall beyond the audible spectrum and are not classified as audio for human perception, though they find applications in seismic monitoring for earthquake detection and medical imaging for diagnostic visualization, respectively.^[21]^[22]^[23] Perceived loudness within the audible range is quantified using units like the phon and sone, which account for the ear's frequency-dependent sensitivity rather than physical intensity alone. The phon scale equates a sound's loudness to that of a 1 kHz tone at a specific decibel level; for instance, a tone perceived as equally loud to a 60 dB SPL 1 kHz sound is rated at 60 phons.^[24] The sone provides a more linear measure of subjective loudness, where a doubling of perceived volume corresponds to a roughly 10-phon increase, facilitating comparisons across frequencies. These standards derive from equal-loudness contours, originally mapped in the 1933 study by Harvey Fletcher and Wilden A. Munson, which experimentally determined the sound pressure levels required for tones of varying frequencies to sound equally loud using headphone presentations to listeners.^[25] These contours reveal the ear's reduced sensitivity at low and high frequencies, especially at lower intensities, and form the basis for modern ISO standards on loudness perception.^[26]

Characteristics

Amplitude and Levels

In audio signals, amplitude denotes the maximum positive or negative deviation of the waveform from its equilibrium (zero) level, quantifying the signal's magnitude and directly influencing perceived loudness, as greater deviations produce stronger pressure variations in the medium.^[27]^[28] Electrical audio signals are typically measured in volts (V), representing the voltage across a load, though logarithmic scales like decibels (dB) are preferred for their ability to compress the vast dynamic range of human hearing (approximately 120 dB) into manageable values.^[29] The decibel scale for voltage ratios is defined by the formula

\mathrm{dB} = 20 \log_{10} \left( \frac{V_1}{V_0} \right),

where V_1 is the measured voltage and V_0 is the reference voltage, allowing relative comparisons such as a doubling of voltage equating to +6 dB.^[30] Professional audio equipment operates at a nominal line level of +4 dBu, corresponding to 1.228 V RMS (root mean square), which provides a standardized reference for balanced interconnections in studios and live sound systems.^[29] In contrast, consumer-grade devices use a lower -10 dBV line level, equivalent to 0.316 V RMS, suited for unbalanced RCA connections in home entertainment systems.^[29] To avoid clipping—where signal peaks exceed the system's maximum capacity, introducing harsh harmonic distortion—professional audio chains incorporate 20-24 dB of headroom above nominal levels, ensuring transients do not distort while maintaining clean reproduction.^[31] This headroom is critical because peak levels (instantaneous maxima) can exceed RMS levels (effective average for power) by 10-20 dB in typical program material, such as music with dynamic drums or vocals.^[32]

Frequency and Phase

Audio signals are composed of frequency components that determine their pitch and timbre. The fundamental frequency represents the lowest frequency in a periodic waveform, corresponding to the perceived pitch of a tone, while overtones are higher-frequency components that are integer multiples of the fundamental, forming the harmonic series. These harmonics, such as the second harmonic at twice the fundamental frequency and the third at three times, contribute to the unique character of different sounds, like the richness in a musical note from a violin.^[33] Spectrum analysis reveals these frequency components by decomposing the time-domain signal into its frequency-domain representation. The Fourier transform achieves this conceptually by expressing the signal as a sum of sine and cosine waves at various frequencies, amplitudes, and phases, allowing identification of the fundamental and harmonic strengths without deriving the mathematical process. This analysis is essential for understanding how audio signals carry complex tonal information beyond simple periodicity.^[34]^[35] Phase in audio signals refers to the relative timing or position of waveform cycles, typically measured in degrees (0° to 360°) or radians (0 to 2π). It quantifies the shift between corresponding points on two or more waves of the same frequency, with a phase shift given by the formula \phi = 2\pi f t, where f is frequency and t is time delay. This property becomes evident when signals interfere, altering their combined waveform.^[36]^[37] Phase distortion, where different frequencies experience varying phase shifts, can lead to comb filtering—a series of frequency notches and peaks resembling a comb—resulting in unnatural coloration or loss of clarity in the audio. In applications like stereo imaging, precise phase alignment between left and right channels enhances spatial perception, while in reverb effects, controlled phase differences simulate room reflections for depth. These implications highlight phase's role in maintaining signal integrity and perceptual quality.^[38]^[39]^[40] The full audio bandwidth spans approximately 20 Hz to 20 kHz, capturing the range of human hearing for high-fidelity reproduction, whereas telephone systems limit to 300 Hz to 3400 Hz to conserve transmission bandwidth, trading off clarity and naturalness for efficiency in speech conveyance. This narrower band suffices for intelligibility but omits low bass and high sibilance, illustrating trade-offs in audio applications.^[41]^[42]^[43]

Analog Representation

Waveforms

Analog audio signals are typically represented by continuous waveforms that vary in voltage or current over time, mirroring the acoustic pressure fluctuations of sound waves. The simplest and most fundamental waveform is the sine wave, which corresponds to a pure tone with no harmonics, producing a smooth, continuous oscillation at a single frequency.^[44] This waveform is ideal for testing audio systems because it lacks additional frequency components that could complicate analysis. In contrast, a square wave consists of abrupt transitions between high and low voltage levels, resulting in a Fourier series dominated by odd harmonics that decrease in amplitude with increasing order, often imparting a harsh, buzzy timbre due to the prominence of these higher-frequency components.^[45] The sawtooth wave, characterized by a linear rise followed by a sharp drop, features a richer harmonic content with both even and odd harmonics of equal phase that diminish progressively, making it particularly useful in analog synthesizers for generating bright, full-bodied tones resembling brass or string instruments.^[45] More complex waveforms arise in natural audio sources like speech and music, where multiple frequency components interact to form distinctive acoustic profiles. In human speech, the waveform is modulated by vocal tract resonances known as formants, with the first formant (F1) approximately 250–850 Hz and the second formant (F2) approximately 700–2700 Hz, contributing significantly to vowel intelligibility and overall timbre.^[46] Musical waveforms, such as those from percussion or plucked strings, often include transient attacks—rapid onsets with high-energy bursts—and subsequent decays, where amplitude falls exponentially, shaping the perceived instrument identity through these temporal dynamics.^[47] These waveforms are generated in analog audio systems primarily through transducers, such as dynamic microphones, where sound pressure waves cause a diaphragm to vibrate, converting mechanical motion into an electrical voltage via electromagnetic induction or other principles.^[48] However, nonlinearities in transducers or amplifiers can introduce distortions that alter the original waveform. Harmonic distortion occurs when additional frequency components at integer multiples of the fundamental are generated, quantified by total harmonic distortion (THD), defined as

\text{THD} = \frac{\sqrt{\sum_{n=2}^{\infty} H_n^2}}{H_1} \times 100\%

where H_1 is the amplitude of the fundamental and H_n are the amplitudes of the harmonics.^[49] Intermodulation distortion, another common type, arises when multiple input frequencies interact in a nonlinear device to produce sum and difference frequencies not present in the original signal, potentially creating dissonant artifacts that degrade audio fidelity.^[50]

Transmission Standards

Analog audio signals are transmitted over wired connections using either balanced or unbalanced lines, with the choice depending on the required noise immunity and distance. Balanced lines employ two signal conductors with equal impedance to ground, carrying signals of opposite polarity that allow differential receivers to reject common-mode noise through cancellation. This differential signaling provides superior noise rejection, often achieving common-mode rejection ratios (CMRR) exceeding 100 dB in well-designed systems. The XLR connector is the standard for balanced audio transmission, featuring three pins: two for the differential signal pair and one for ground, enabling robust connections in professional environments like studios and live sound.^[51] In contrast, unbalanced lines use a single signal conductor with a ground return, making them susceptible to electromagnetic interference, particularly over longer distances. The RCA connector is commonly used for unbalanced transmission, with two contacts: one for the signal and one for ground (shield), suitable for consumer applications such as home audio systems where runs are short. Unbalanced connections lack the noise-cancellation benefits of differential signaling, resulting in poorer performance in noisy environments.^[51] Impedance matching ensures efficient signal transfer between sources and loads. Historically, a 600 Ω standard originated from telephone systems and was adopted in early audio equipment for consistent power transfer, serving as the reference for volume units (VU) where 0 VU corresponds to 1 mW in a 600 Ω load. Modern professional audio favors bridging configurations, with low output impedances (around 50–200 Ω) driving high input impedances (10 kΩ or greater) to minimize loading effects and voltage loss. This shift, recommended by IEC standards since 1978, improves signal integrity without the need for exact matching.^[52]^[51] For wireless transmission, analog audio is modulated onto carrier waves using amplitude modulation (AM) or frequency modulation (FM). In AM, the audio signal varies the amplitude of a carrier in the medium frequency band (typically 535–1705 kHz), while FM varies the carrier frequency in the VHF band (88–108 MHz), offering better noise resistance and fidelity for broadcast audio. FM stereo transmission, standardized globally, encodes left and right channels using a 38 kHz subcarrier within the 15 kHz audio bandwidth, allowing receivers to demodulate high-quality stereo signals.^[53] Cabling imposes practical limits on transmission due to resistive and capacitive effects. Capacitive losses between conductors form a low-pass filter, attenuating high frequencies over long runs; for instance, cable capacitance of 30–50 pF/m combined with source impedance can reduce the -3 dB point at 20 kHz for runs exceeding 100 m. Balanced lines mitigate this and noise better than unbalanced, supporting maximum lengths up to 100 m for line-level signals at 20 kHz with minimal degradation, provided low-capacitance cables (under 50 pF/m) are used. Unbalanced lines are limited to shorter distances, typically under 10–15 m, to avoid significant high-frequency roll-off and hum pickup.^[54]

Digital Representation

Sampling and Quantization

Sampling and quantization are the core processes in converting an analog audio signal into its digital representation, discretizing the continuous-time waveform in both time and amplitude domains. Sampling involves measuring the signal's amplitude at regular intervals, known as the sampling rate, to capture its temporal variations without loss of information. According to the Nyquist-Shannon sampling theorem, formulated by Harry Nyquist in 1928 and rigorously proved by Claude Shannon in 1949, the sampling rate must exceed twice the highest frequency component in the signal to enable perfect reconstruction from the samples, preventing distortion known as aliasing.^[55]^[56] For human-audible audio, which typically extends up to 20 kHz, a sampling rate greater than 40 kHz is required; in practice, this is achieved with rates like 44.1 kHz for compact disc (CD) audio, allowing faithful reproduction of frequencies up to 22.05 kHz while incorporating a safety margin against filter imperfections.^[57] Aliasing occurs when the sampling rate is insufficient, causing higher frequencies to masquerade as lower ones in the digital domain, which can introduce inaudible or audible artifacts if not mitigated by anti-aliasing filters prior to sampling. The theorem ensures that a bandlimited signal—here, audio constrained below half the sampling rate—can be reconstructed using a low-pass filter, preserving the original waveform's integrity. Higher sampling rates, such as 96 kHz used in high-resolution audio formats, provide greater headroom for frequencies beyond the audible range, facilitating advanced processing like oversampling in digital-to-analog converters while reducing aliasing risks in non-linear operations.^[58] Quantization follows sampling by mapping each continuous amplitude sample to the nearest discrete level from a finite set, determined by the bit depth of the digital system. The quantization step size \Delta is given by \Delta = \frac{\text{full scale}}{2^n}, where n is the number of bits per sample and full scale represents the maximum amplitude range (e.g., from -V to +V). For 16-bit audio, common in CD specifications, this yields $2^{16} = 65,536 levels, providing fine granularity for amplitude representation across the dynamic range. In high-resolution formats, 24-bit quantization expands to $2^{24} \approx 16.8 million levels, enhancing precision for subtle signal details and reducing perceptible distortion in quiet passages.^[57]^[58] The primary error introduced by quantization is noise, modeled as uniform random distribution across each step, with power \sigma_q^2 = \frac{\Delta^2}{12}. This quantization noise limits the signal-to-noise ratio (SNR), approximated for a full-scale sinusoidal input as \text{SNR} \approx 6.02n + 1.76 dB, where each additional bit improves SNR by about 6 dB, doubling the effective dynamic range. For 16-bit audio, this yields an SNR of roughly 98 dB, sufficient for most listening environments, while 24-bit extends it to about 146 dB, approaching the limits of human hearing sensitivity. To mitigate the harsh, correlated nature of quantization distortion—especially at low signal levels—dithering adds low-level, uncorrelated noise (typically 1-2 bits below the least significant bit) before quantization, randomizing errors and linearizing the process to mask artifacts as benign hiss rather than distortion.^[59]^[60]^[61]

Binary Encoding

Binary encoding of audio signals involves representing quantized samples as binary data for storage and transmission, primarily through uncompressed formats like Pulse Code Modulation (PCM) or compressed alternatives. PCM serves as the foundational uncompressed standard for digital audio, where each sample is encoded as a fixed-length binary word without further processing. In PCM, audio data for multichannel configurations, such as stereo, is typically stored with interleaving, alternating samples from each channel (e.g., left channel followed by right channel) to maintain temporal alignment.^[62]^[63] The bit depth in PCM determines the number of discrete amplitude levels and the overall dynamic range. A 16-bit PCM encoding provides 65,536 quantization levels (2^16), yielding a theoretical dynamic range of 96 dB, sufficient for consumer audio applications like CDs. In contrast, 24-bit PCM offers 16,777,216 levels (2^24), extending the dynamic range to 144 dB, which is preferred in professional recording for greater headroom and reduced quantization noise.^[64]^[45] Common file formats encapsulate this binary data with structural metadata. The WAV (Waveform Audio File Format) stores uncompressed PCM data in a RIFF-based container, using little-endian byte order and including headers that specify parameters like sample rate, bit depth, and number of channels. For example, the 'fmt' chunk in WAV defines the audio format, with the data chunk holding the interleaved PCM samples. AIFF (Audio Interchange File Format), developed by Apple, similarly supports uncompressed PCM but employs big-endian byte order for its IFF-based structure, with headers conveying metadata such as sample rate in the 'COMM' chunk.^[65]^[63]^[66] For efficient storage, lossy compression formats like MP3 (MPEG-1/2 Audio Layer III) apply perceptual coding to discard inaudible audio components based on human psychoacoustics, achieving bitrates as low as 128 kbps while approximating the original signal quality. This reduces file sizes dramatically compared to uncompressed PCM (e.g., from 1.4 Mbps for CD-quality stereo), making it suitable for distribution, though it introduces irreversible data loss.^[67]

Signal Flow

Audio Chain

The audio chain refers to the sequential path an audio signal follows through an interconnected system, from initial capture to final reproduction, ensuring coherent transmission and playback. This chain encompasses both analog and digital components, though the focus here is on the high-level flow without delving into conversion specifics. Proper management of the chain minimizes degradation, such as noise or distortion, while optimizing signal integrity across stages.^[68] A typical audio chain begins at the source, where sound is transduced into an electrical signal via a microphone for acoustic inputs or directly from an instrument like a guitar. The signal then proceeds to a preamplifier (preamp), which boosts the low-level microphone signal (often in the microvolt range) to line level (around 1 volt) for compatibility with downstream equipment. Following preamplification, the signal enters the mixing or processing stage, where multiple sources are combined, balanced, and routed—typically in a console or digital audio workstation (DAW)—to create a cohesive output. From there, it advances to the amplification stage, where a power amplifier increases the signal strength to drive output transducers. Finally, the chain culminates at the output device, such as speakers or headphones, converting the electrical signal back to acoustic sound waves for listener perception.^[68] Gain staging is a critical practice throughout the audio chain, involving the adjustment of signal levels at each stage to prevent clipping (overloading) or excessive noise amplification while maintaining optimal headroom—typically targeting peaks around -6 dBFS and averages near -18 dBFS in digital contexts. This ensures the signal remains clean and dynamic without cumulative degradation; for instance, excessive gain early in the chain can introduce noise that propagates downstream, whereas insufficient gain may require compensatory boosts later, amplifying unwanted artifacts. Conceptual block diagrams of the audio chain often illustrate this as a linear series of blocks: Source → Preamp (with gain control) → Mixer/Processor (level balancing) → Amplifier (power scaling) → Output, emphasizing unidirectional flow and interconnection points for monitoring or intervention.^[69] Audio chains can operate in mono, utilizing a single channel for basic reproduction suitable for voice announcements or simple recordings, or in stereo, employing two channels (left and right) to simulate spatial imaging through amplitude panning, where signal distribution between channels creates perceived directionality and width. Stereo enhances immersion by mimicking human binaural hearing, providing a sense of depth and positioning not achievable in mono. Extensions to surround sound, such as the 5.1 format, further expand this by incorporating five full-range channels (front left, center, front right, surround left, surround right) plus a low-frequency effects (LFE) channel, enabling more enveloping spatial audio for cinema and home theater applications.^[70]^[71] Latency, or delay through the chain, varies significantly by implementation: analog chains exhibit near-zero delay due to direct electrical propagation without buffering, ideal for live performance. In contrast, digital chains introduce buffering in DAWs or processors, typically resulting in 10-50 ms of round-trip latency depending on buffer size (e.g., 512 samples at 44.1 kHz yields about 11 ms), which can affect real-time monitoring but is often imperceptible in non-interactive playback.^[72]^[73]

Processing Techniques

Audio signal processing techniques are essential for modifying signals to enhance quality, correct imperfections, or achieve artistic effects within the audio chain. These methods manipulate frequency content, dynamic range, spatial characteristics, and unwanted artifacts, often applied in real-time or offline using analog or digital systems. Common applications include music production, broadcasting, and live sound reinforcement, where precise control improves clarity and listener experience.

Equalization (EQ)

Equalization adjusts the amplitude of specific frequency bands to balance tonal characteristics or compensate for room acoustics and equipment responses. Parametric equalizers, a widely adopted type, allow independent control of center frequency f_c, gain, and bandwidth via the quality factor Q, defined as the ratio of the center frequency to the bandwidth (Q = f_c / \text{BW}).^[74] Higher Q values produce narrower, more precise boosts or cuts, enabling surgical corrections without affecting adjacent frequencies, as demonstrated in early implementations using cascaded biquad filters.^[75] This flexibility has made parametric EQ a standard tool since its introduction in professional audio in the 1970s, outperforming fixed graphic equalizers in accuracy for corrective tasks.^[76]

Dynamics Processing

Dynamics processing controls the amplitude variations in an audio signal to manage loudness consistency and prevent overload. Compression reduces the dynamic range by attenuating signals exceeding a set threshold, using a compression ratio (e.g., 4:1, meaning input exceeding the threshold by 4 dB is reduced to 1 dB output) to smooth peaks while preserving transients.^[77] Key parameters include attack time (onset of gain reduction, typically 1-30 ms to allow punchy transients) and release time (recovery after the signal drops, 50-500 ms to avoid pumping effects).^[77] Limiting, a form of compression with an infinite ratio, caps peaks to avoid clipping, commonly applied in mastering to maintain headroom without distortion.^[77]

Effects

Effects processors add spatial or temporal modifications to enrich audio signals, simulating environments or creating rhythmic elements. Reverb simulates acoustic reflections, with convolution methods convolving the dry signal with measured impulse responses from real spaces for accurate, high-fidelity decay tails.^[78] Algorithmic reverb, such as feedback delay networks (FDNs), generates synthetic reflections using delay lines and feedback matrices, offering computational efficiency and adjustable parameters like room size and decay time.^[78] Delay effects produce echoes by recirculating a delayed copy of the signal, controlled by a feedback coefficient (0-1, where 1 yields infinite repeats and lower values create decaying repeats).^[79] These techniques enhance depth in mixes, with FDNs achieving dense echo patterns for natural-sounding reverb.^[79]

Noise Reduction

Noise reduction techniques suppress unwanted interference like hum (typically at 60 Hz from power lines) or hiss (high-frequency tape noise), preserving the core signal. Analog systems, such as Dolby A (professional) and Dolby B (consumer), employ pre-emphasis and companding: signals are compressed during recording to boost quiet parts above noise floor, then expanded on playback for up to 10-20 dB reduction without introducing artifacts.^[80] Digital methods like spectral subtraction estimate noise spectrum during silent periods and subtract it from the signal's magnitude spectrum in the frequency domain, effectively mitigating broadband noise but potentially causing musical noise if over-applied.^[81] Multi-stage spectral subtraction refines this by applying subtraction iteratively across frequency bands, improving signal-to-noise ratios in non-stationary audio like speech or music.^[81]

References

[1]
Home - DSP Audio Effects - Research Guides Home at Arkansas ...
An analog audio signal is a continuous electrical representation of sound waves in the air, typically carried by variations in voltage, current, or charge.
[2]
Linked Open Data for Digital Audio Formats - National Archives
Sep 11, 2025 · Digital audio is defined as an audio waveform that has been created as, or converted into, digital form and encoded as a sequence of numerical ...
[3]
Understanding Sound - Natural Sounds (U.S. National Park Service)
Jul 3, 2018 · Humans with normal hearing can hear sounds between 20 Hz and 20,000 Hz. Frequencies above 20,000 Hz are known as ultrasound.Missing: signal | Show results with:signal
[4]
Introduction to Audio Signal Processing - RIT Press
Jan 20, 2017 · Audio signal processing is used to convert between analog and digital formats, to cut or boost selected frequency ranges, to remove unwanted noise, to add ...
[5]
Audio Signal Processing for Music Applications | Stanford Online
Audio signal processing is an engineering field that focuses on the computational methods for intentionally altering sounds, methods that are used in many ...
[6]
[PDF] Digital Audio Systems - Stanford CCRMA
Digital audio systems use a form of binary coding to represent the audio data. The traditional system uses a fixed-point representation (pulse code ...
[7]
1. Sounds, Signals, and Recordings
A signal, or, to be more explicit, an analog signal, is a voltage or current that goes up and down in time analogously to the changing pressure at a fixed point ...
[8]
[PDF] Chapter 2: Basics of Signals
Signals are physical variables of interest, like voltage or position, often represented as a function of time, such as f(t). Examples include audio, images, ...
[9]
The Audible Spectrum - Neuroscience - NCBI Bookshelf - NIH
Humans can detect sounds in a frequency range from about 20 Hz to 20 kHz. (Human infants can actually hear frequencies slightly higher than 20 kHz.)Missing: signal | Show results with:signal
[10]
[PDF] Introduction to Electronics
Audio signals are usually time-varying voltage signals. Current (I): flow of charge through a circuit. (Amperes = coulombs/sec; the coulomb is a measure of ...
[11]
Alexander Graham Bell patents the telephone | March 7, 1876
In this first telephone, sound waves caused an electric current to vary in intensity and frequency, causing a thin, soft iron plate–called the diaphragm–to ...
[12]
Radio and Telephony Technologies During the G.I. Generation
Two great scientific minds both laid claim to the creation of the radio: Nikola Tesla and Guglielmo Marconi. It should be noted that this first iteration of ...
[13]
Sum and Difference Frequencies, Sidebands and Bandwidth
The high signal frequencies associated with video broadcasting require higher bandwidths for channels assigned to television.
[14]
[PDF] Linear Modulation and Demodulation
High fidelity audio requires 20 kHz. Video requires a few MHz. The means of transforming one or more of these signals so that they occupy a portion of the RF ...
[15]
Audio Fundamentals - Help Wiki
Feb 18, 2022 · They are used to take multiple analog signals (voltage) from a mixer or microphone, and convert it to digital signals (1's and 0's). It is also ...<|control11|><|separator|>
[16]
Extended High Frequency Thresholds in College Students - NIH
Human hearing is sensitive to sounds from as low as 20 Hz to as high as 20000 Hz in normal ears. However, clinical tests of human hearing rarely include ...
[17]
Human Hearing
The range of human hearing is generally considered to be 20 Hz to 20 kHz, but it is far more sensitive to sounds between 1 kHz and 4 kHz. For example, ...
[18]
Hearing Loss in Adults: Differential Diagnosis and Treatment - AAFP
Jul 15, 2019 · Presbycusis characteristically involves gradual onset of bilateral high-frequency hearing loss associated with difficulty in speech ...
[19]
Presbycusis - StatPearls - NCBI Bookshelf - NIH
It is the most common cause of hearing loss worldwide and is estimated to affect approximately two-thirds of Americans aged 70 or older. The hallmark of ...Introduction · Etiology · History and Physical · Evaluation
[20]
https://ec.europa.eu/health/opinions/en/hearing-loss-personal-music-player-mp3/figtableboxes/table-1.htm
[21]
Examples of sound pressure levels in relation to hearing threshold ...
... hearing threshold and pain threshold (in dB SPL). The range of human ear's audible sounds goes from 0 dB SPL (hearing threshold) to 120-140 dB SPL (pain ...Missing: dynamic | Show results with:dynamic
[22]
https://blog.quakelogic.net/introducing-the-sis-1-infrasound-sensor-precision-in-low-frequency-detection/
[23]
Introducing the SIS-1 Infrasound Sensor: Precision in Low ...
Jul 14, 2024 · Seismic Research: Infrasound sensors can detect low-frequency waves generated by earthquakes, providing early warning systems that can save ...
[24]
Ultrasound
Ultrasound probes, called transducers, produce sound waves that have frequencies above the threshold of human hearing (above 20KHz), but most transducers in ...
[25]
Loudness Units: Phons and Sones - HyperPhysics
If a given sound is perceived to be as loud as a 60 dB sound at 1000 Hz, then it is said to have a loudness of 60 phons. 60 phons means "as loud as a 60 dB, ...
[26]
[PDF] Loudness, Its Definition, Measurement and Calculation
Loudness, Its Definition, Measurement and Calculation*. By HARVEY FLETCHER and W. A. MUNSON. An empirical formula for calculating the loudness of any steady ...
[27]
Equal-Loudness Contour - an overview | ScienceDirect Topics
Although the Fletcher-Munson curves provided an accurate measure of the relative loudness of tones, their shape was too complicated for use with an analog sound ...
[28]
https://cs.wellesley.edu/~cs203/lecture_materials/dynamics/dynamics.pdf
[29]
[PDF] Dynamics and Loudness
Loudness depends upon the amplitude of the sound pressure. • Instinctively, this makes sense. A sine wave with higher amplitude will sound louder than a sine ...
[30]
Pro Audio Reference (D) - Audio Engineering Society
-10 dBV Standard voltage reference level for consumer and some pro audio use (e.g. TASCAM), equal to 0.316 Vrms.Missing: formula | Show results with:formula
[31]
Decibels Express the Ratio of Two Voltage Values for Power Gain
Decibels (dB) is a logarithmic unit for measuring the ratio of two values of power or root-power, such as electric voltage or current.Missing: AES | Show results with:AES
[32]
What You Need to Know About Headroom - InSync - Sweetwater
Jun 14, 2022 · Unless your channel's meters can alert you to intersample distortion, you should leave a few dB of headroom to avoid this. Many mastering ...
[33]
Amplitude - Sonarworks Blog
RMS amplitude refers to the working average voltage level of an audio signal. In professional audio devices, line level (+4 dBu) is defined as 1.228 volts RMS.Missing: deviation | Show results with:deviation
[34]
[PDF] Mathematics of Music 1. INTRODUCTION 2. HISTORY OF FOURIER ...
the harmonic series. The harmonic series is the series of tones created by multiplying a fundamental frequency by integers. This can be done based on any ...
[35]
FFTs and spectrograms – Understanding Sound - Pressbooks.pub
Frequency domain graphs show much of each frequency is present in a sample. Frequency is plotted along the x-axis and amplitude is plotted along the y-axis.
[36]
Fourier Series with Sound - ComPADRE
The Fourier spectrum shows which frequencies are present and the relative strength of each frequency. The lowest frequency present is called the fundamental ...
[37]
Phase Difference and Phase Shift in an AC Circuit
The difference or phase shift as it is also called of a Sinusoidal Waveform is the angle Φ (Greek letter Phi), in degrees or radians that the waveform has ...
[38]
Phase angle (deg) φ = time delay Δ t × frequency f × 360
Sound delay time = Time Delay Δ t (time difference) oscilloscope - Phase lag istime shift. Connection between phase, phase angle, frequency, time of arrival Δ ...
[39]
https://www.soundonsound.com/techniques/phase-demystified
[40]
Phase Demystified - Sound On Sound
Variable Phase Devices. The comb filtering that occurs due to phase cancellation can often be a desirable effect, and while a simple delay can be used to ...
[41]
What is Comb Filtering and how to avoid it - Blog | Resources - Q-SYS
Jun 23, 2021 · Comb filtering occurs when certain frequencies are either amplified or attenuated by the superposition of a delayed version of the original audio signal onto ...
[42]
Acoustic compression in Zoom audio does not compromise voice ...
Oct 31, 2023 · The dominant limitation of telephone audio (both landline and mobile) is its reduced bandwidth compared to studio audio. Landline telephones ...
[43]
Effect of bandwidth extension to telephone speech recognition ... - NIH
Compared to speech in face-to-face conversational settings, telephone speech does not convey information above 3400 Hz, which is useful in the identification of ...
[44]
Audio Frequency Range Explained - Gear4music
Aug 14, 2023 · ... Hz (20 kHz) would be a very high-pitched sound. Our auditory system is sensitive to a wide range of frequencies, typically from 20 Hz to 20 kHz.
[45]
Practically perfect pitch
people using sine waves as the stimuli to evaluate how readily frequencies can be identified. Sine waves pro- vide no harmonics or cues other than frequency ( ...Missing: audio | Show results with:audio
[46]
Pro Audio Reference (S) - Audio Engineering Society
sawtooth wave A periodic waveform characterized by a 50% duty cycle and a Fourier series consisting of both even- and odd-ordered, equal phase, sinusoidal ...
[47]
Frequencies, bandwidths and magnitudes of vocal tract and ...
May 20, 2016 · In speech, the frequency of the first formant F1 typically has a value between about 300 and 800 Hz that increases with mouth opening. The ...
[48]
Temporal characterization of experimental recorder attack transients
Jan 19, 2017 · In this work, the attack transient is considered to be the transition between no sound and a stable oscillation. It starts when the musician ...
[49]
A high-quality digital radio-frequency capacitor microphone with ...
Mar 27, 2020 · Capacitor microphones are widely used to transduce sound waves into electrical voltages. The converting capacitor can either be operated in ...
[50]
Analysis of total harmonic distortion of miniature loudspeakers used ...
Mar 8, 2021 · Total har- monic distortion (THD) is a key evaluation criterion for the performance of a miniature loudspeaker defined as the ratio of the root ...
[51]
Pro Audio Reference (I) - Audio Engineering Society
The test specifies two equal-amplitude closely spaced high frequency signals. Common test tones are 19 kHz and 20 kHz for full audio bandwidth units. While ...
[52]
[PDF] 1 Design of High-Performance Balanced Audio Interfaces
Design of High-Performance Balanced Audio Interfaces. Bill Whitlock. Jensen Transformers, Inc. 9304 Deering Avenue. Chatsworth, CA 91311. High signal-to-noise ...
[53]
[PDF] The Measurement of Audio Volume
Mar 31, 2025 · Madison Ave., New York 22, N. Y. a A standard impedance of 600 ohms was chosen originally since, keeping in mind the telephone plant, there ...
[54]
FM Radio - Federal Communications Commission
FM, or frequency modulation, encodes audio signals on a carrier frequency. FM stations operate in the 88-108 MHz band.Missing: stereo | Show results with:stereo
[55]
[PDF] driving loooong cables - EAW
"TERMINATING" LONG CABLE RUNS. There is still a common practice of putting 600 Ohm or similar terminating resistors at the end of long lines. You can now see ...
[56]
[PDF] Certain topics in telegraph transmission theory
Synopsis—The most obvious method for determining the distor- tion of telegraph signals is to calculate the transients of the tele- graph system.
[57]
[PDF] Communication in the Presence of Noise* - MIT Fab Lab
Shannon: Communication in the Presence of Noise ity than the message space. The type of mapping can be suggested by Fig. 3, where a line is mapped into a ...
[58]
[PDF] Red Book (audio CD standard) - Wikipedia, the free encyclopedia
Mar 25, 2007 · Bit rate = 44100 samples/s × 16 bit/sample × 2 channels=1411.2 kbit/s (more than 10 MB per minute). Sample values range from -32768 to +32767.
[59]
High Resolution Audio | Sony CA
There is no single standard for High-Resolution Audio, but the most commonly used specifications are 24-bit/96 kHz (3.2x more data transmitted than CD) and 24- ...<|control11|><|separator|>
[60]
[PDF] Taking the Mystery out of the Infamous Formula, "SNR=6.02N + 1.76 ...
Oversampling in conjunction with quantization noise shaping and digital filtering are the key concepts in sigma-delta converters, although oversampling can be ...<|control11|><|separator|>
[61]
Quantization Noise and Amplitude Quantization Error in ADCs
Apr 22, 2019 · In this article, we'll look at the conditions under which we are allowed to use a noise source to model the quantization error.
[62]
What is Dithering? Using Dithering to Eliminate Quantization Distortion
Dec 4, 2022 · This signal processing technique, known as dithering, deliberately adds a noise signal with appropriate PDF (probability density function) and ...
[63]
G.711 : Pulse code modulation (PCM) of voice frequencies
### Summary of PCM Definition for Audio Signals (G.711)
[64]
WAVE PCM soundfile format
The WAVE file format is a subset of Microsoft's RIFF specification for the storage of multimedia files.
[65]
Audio Bit Depth: Everything you need to know - SoundGuys
Dec 17, 2024 · 16-bit audio provides 96dB of dynamic range, while 24-bit provides 144 dB. However, in practical listening situations, this difference is ...
[66]
WAVEFORMAT structure (mmreg.h) - Win32 - Microsoft Learn
Aug 15, 2022 · Waveform-audio data is PCM. Number of channels in the waveform-audio data. Mono data uses one channel and stereo data uses two channels. Sample ...
[67]
AIFF / AIFC Sound File Specifications - McGill University
Sep 20, 2017 · The file header and its numerical fields still use big-endian byte order, but the PCM data for AIFF-C/sowt uses little-endian byte order.
[68]
[PDF] The Theory Behind Mp3
Perceptual coding is a lossy process and therefore it is not possible to regain this information when decompressing. This is fully acceptable since the ...
[69]
Basic Signal Flow Overview - Berklee Online
Jan 11, 2024 · Signal flow refers to the path that an audio signal takes from its source to the output. Understanding signal flow is critical for audio engineers.
[70]
Guide to gain staging in audio production - Avid
May 30, 2024 · Gain staging is the process of managing the volume levels of audio signals at each stage in the audio signal path, ensuring optimal sound ...
[71]
[PDF] Presented at ^uD, the 101stConvention 1996 November 8-11 Los ...
The primary method, panning, simply involves an amplitude control that adjusts the proportion of the mono signal applied to the left and right stereo channels.
[72]
Surround Sound: What It Is, How It Works, and Why Dolby Atmos ...
Surround sound system configurations: 5.1, 7.1, and Dolby Atmos. A 5.1 surround sound system utilises six audio channels, while a 7.1 system employs eight ...<|control11|><|separator|>
[73]
Chapter Five: Digital Audio - DACs - Introduction to Computer Music
A typical buffer of 512 samples creates a latency of ~11 ms at 44.1k, yet a 4096-sample buffer creates a ~93 ms delay, which can be noticeable.Missing: signal | Show results with:signal
[74]
Audio Engineering Society - Convention e-Brief
Oct 20, 2011 · In computer based Digital Audio Workstation (DAW) systems the major latency (typically around 5 to 10ms) is caused by the buffer size [3], so ...
[75]
Pro Audio Reference (Q) - Audio Engineering Society
Q (upper-case) Quality factor. Filters. The selectivity factor defined to be the ratio of the center frequency f divided by the bandwidth BW. See the RaneNote ...
[76]
AES Convention Papers Forum » Parametric Equalization
May 1, 1972 · This presentation concerns the application of new equalization techniques to professional audio control. The device utilized is a parametric ...
[77]
Design of digital parametric equalizer based on second-order function
The quality factor Q of filters doesn't vary with frequency, and thus they have high practical value.
[78]
[PDF] Digital Dynamic Range Compressor Design— A Tutorial and Analysis
Dynamic range compression maps audio signal dynamic range to a smaller range, reducing higher peaks. Key parameters include threshold, ratio, attack, and ...Missing: seminal | Show results with:seminal
[79]
Fifty Years of Artificial Reverberation | IEEE Journals & Magazine
Feb 29, 2012 · The first artificial reverberation algorithms were proposed in the early 1960s, and new, improved algorithms are published regularly.
[80]
https://secure.aes.org/forum/pubs/conventions/?elib=1905
[81]
AES Convention Papers Forum » A Noise Reduction System for ...
Mar 1, 1971 · The B-Type noise reduction system is designed for the reduction of hiss in consumer tape products. Employing techniques similar to those ...
[82]
Multi-stage spectral subtraction for enhancement of audio signals
Abstract: This paper describes a technique based on spectral subtraction (SS) for enhancing audio signals which have nonstationary properties.