Sound reproduction, audio measurements & fidelity: the 10 basic rules

May 4, 2025

Audio measurements, problems related to audio fidelity or sound reproduction can be a mess for "newbies". Without a compass it's easy to get lost in the noisy jungle of flawed reasoning, false analogies and junk philosophies (not to mention marketing frauds). The following is a list of the most important ground rules covering the science and logic behind sound reproduction and audio measurement.

1. Audio science is an applied science

Audio science (science of sound reproduction) is an applied science and not a fundamental science. However, as other applied fields it has roots in fundamental sciences. - Any claim about amplifiers, DACs, audio formats or loudspeakers is a claim about hearing, physics or signal behavior.

2. Audible difference... is only a difference

An audible difference is only a difference and not the final answer ("trust your ears" nonsense). By testing with music we can only make claims about the sound and not about the technology or the real cause. The latter requires special experiments.

We can also add that the analytical and focusing ability of our hearing system is very limited, especially when compared to vision. Separating one part of an image from the rest is a simple task, but separating one part of the music from the rest - for example, the sound of an instrument in an orchestral music - can be extremely difficult. Comparing minor audible differences is a difficult task as well (when differences become small, belief can mess up perception and the wire drama begins... ).

3. Blind test is a solution to only one problem!

Our senses can fool us, especially when the perceived differences are minor. Blind test, or its most commonly used form, the ABX test, is a universal tool for eliminating self-deception arising from our expectations and unreliable perception (unreliable auditory memory, confirmation bias). Blind test doesn't provide a solution for how to design a test free of flaws, how to control all variables (how to avoid false positive results) and if the test is free of errors, correlation still doesn't imply causation.

Blind testing methodology is the most misunderstood topic in audio. Rejection of blind testing is absurd, on the other hand, searching for the truth with double blind tests and/or statistical methods is a hopeless and completely pointless endeavour. Testing with music only makes sense if we already have adequate knowledge of the system to be tested. Blind tests shouldn't be used as a primary method, only as a secondary method to support auditory measurements. Moreover, the exclusive use of tests can lead not only to misconceptions but also to superficial knowledge.

Fortunately, our knowledge does not come from "listening tests", but from experiments based on measurement and modeling.

4. Goal is to understand

Goal is to understand the "mechanism" and not just generating or citing test results. Real advances in audio technology come from a deeper understanding of signal behavior, human hearing and sound sources (from loudspeakers to musical instruments). The difference between conventional microphone measurements, dummy head measurements and human hearing is another important subject (two-ear averaging, critical band averaging, precedence effect).

Key to science is the distinction between knowledge without comprehension (~test results) and knowledge with comprehension (~mechanism). The problem is that only the process of model validation with experiments can contribute to the knowledge with comprehension. The rest of the scientific tools can only contribute to the knowledge without comprehension at most (correlation analysis, data science in general). Mere accumulation of knowledge without comprehension can lead to a special form of illusion of knowledge.

Without comprehension results can hide the truth even if these results are acceptable. Even objective testing does not necessarily lead to an understanding of the mechanism, on the other hand, understanding is always accompanied by a reliable test method.

In science (and mainly in human sciences) there is an interesting problem related to null hypothesis testing: it's impossible to completely prove a null hypothesis (examples of null hypothesis: humans can't hear frequencies above 20 kHz, 'X' medicine has no harmful side-effects). The problem arises from the limited number of data samples and is similar to the problem of inductive reasoning (generalization). By understanding the mechanism we can limit the questions open to debate.

Audio Phoolosophy
What has really led the audiophile hobby down a wrong path is the belief that one can gain knowledge simply by listening to music and changing components and doesn't have to do anything else. However, testing with music only makes sense if we already have adequate knowledge.
Moreover, audiophiles are silenced by typical salesman nonsense, such as "you can't hear the difference because you don't have a highly resolving system". The Emperor's New Clothes-style fear of exclusion and "divide and conquer" is still a good strategy to control people and keep them in line. And after a lot of money and time has been wasted, cognitive dissonance takes over the control and locks the human mind: "It's easier to fool people than to convince them that they have been fooled". (The famous quote attributed to Mark Twain is a striking connection between cognitive dissonance and fraud.)

5. Scientific models: key to measurements

The logic behind audio measurements isn't that complicated: audio measurements are signal measurements, measuring only those characteristics that can affect what we hear. Audio fidelity of a component is determined objectively by measuring the change(s) in the signal as it passes through the component and comparing with the corresponding threshold(s).

Audio measurements are only useful if they correlate with perception. Unfortunately, some old-school measurement methods taken directly from electrical measurements with minor modifications correlate poorly with hearing (SNR, SINAD). However, understanding how we hear can be very helpful in interpreting less accurate measurements. If we know the limitations of SINAD and SNR, we will not fool ourselves with SINAD and SNR.

Objective assessment of audio fidelity requires a three-level analysis (model):

an accurate model of the system being measured (system behavior, causes and diagnosis of signal changes in amplifiers, cables, speakers, rooms...),
a hearing model (how we hear, what we can hear, differences between age groups),
characteristics of audio signals, sound sources (musical instruments, human voice, etc.).

Hearing model: psychoacoustics studies the human hearing with special test tones, creates hearing models and determines various thresholds.

6. What we can hear...

What we can hear is determined by the Absolute Thresholds of Hearing (ATH) and auditory masking. Masking means that in the neighborhood of a (loud) tone the hearing threshold is raised. Signals below the threshold are inaudible.

Not only the audibility of pure tones or compex tones, but even the audibility of nonlinear distortion, noise or resonances is related to masking and ATH. (In fact, there is a third mechanism: adaptation or compression, a shift in the non-masked threshold.)

7. "Purity" of the signal is irrelevant

It is not the purity of the signal that matters... In an audio system, the goal is to preserve and transmit the signal in such a way that accumulated errors cannot be heard or low enough to not affect playback fidelity.

The shape of the signal itself is also irrelevant (square wave response, impulse response). We can't assess fidelity by looking at the waveform.

This also means that the aversion to software resampling and the hype around bitperfect playback that is typical today is just another nonsense.

8. Audio measurements & fidelity - the main categories

(A short overview.)

Audio fidelity of a component is determined by frequency response, nonlinear distortion curves and noise (necessary for expressing dynamic range). Time domain measurements (impulse response, phase shift, group delay) are secondary, as DACs and amplifiers have negligible phase distortion in the audible range. In multi-way loudspeakers the phase distortion is orders of magnitude higher, but even this is not audible in the impulse response. Time domain measurements are essential only in room acoustics and differ significantly from phase/group delay measurements.

Audio fidelity measurements describe to what extent an audio system can reproduce the original performance (live, electronic). Poor frequency response can change the timbre (certain harmonics are emphasized while others are attenuated). Distortion can also change the timbre (in a different way) and high level of distortion can hide low-level details. Noise can also hide low-level signals. Usually a system or component with very poor audio fidelity will not sound good. Very poor audio fidelity can even be painful to listen to!

In amplifiers and DACs, crosstalk between channels can also be interpreted as an extra measurement, although It's very rare to find a system with audible crosstalk. There is no need to worry about jitter in digital audio systems either. (Jitter is the most common means of frightening in audiophilia. Jitter is the fluctuation of the clock signal but measures as nonlinear distortion. If jitter is periodic then artifacts will appear in the spectrum of the analog output as sidebands on both side of the audio signal. Jitter could be heard as harsh non-harmonic distortion.)

I recommend Ethan Winer's article on audio parameters (Audiophoolery). Ethan's article organized differently and includes some myth busting.

9. Thresholds & "worst-case" thresholds

Just-noticeable distortion, just-noticeable noise, just-noticeable level difference varies according to the harmonic and temporal properties of the audio signal. There always exists a particular signal with the lowest threshold, which can be identified as "worst-case".

For example, nonlinear distortion is best heard with pure tones and two-tone tones. Detection is more difficult with complex time-varying signals and pulses. In contrast, time-related errors are more audible in pulses or pulse series. (Yes, some measurement signals are worst-case signals.)

The definition of "transparency" and transparent system is closely related to the concept of 'worst-case'. For example, transparency is reached for nonlinearity when distortion is not audible with worst-case type test signals. The only exception is the frequency response of loudspeakers and headphones, where the “worst-case” (just-noticeable level difference with pure tone) is a too strict criterion.

Audibility threshold and the relationship between threshold and waveform are so important principles that cannot be ignored in this field. Any question related to fidelity is about threshold and the relationship between threshold and waveform.

10. Fidelity of the playback system, the format and the recording are different things!

Audiophiles tend to confuse these concepts, especially the fidelity of the format with the fidelity of the (actual) recording. This results in many misconceptions: analog is better than digital, FLAC is better than streaming, just to name some.

As a final word...

The list could go on with many other simple rules, but I considered these to be the most important ("try to think in terms of instruments and not musical styles" is also very helpful).

Csaba Horváth