THD Distortion Audibility Test: Can You Hear It?
Overview
THD Distortion Audibility Test. Click Start Test to run the THD Distortion Audibility Test, a blind staircase that plays a clean sine tone against versions carrying progressively less harmonic distortion — pick whichever one sounds different using the First or Second buttons. The test narrows in on the exact THD threshold where you can no longer hear the difference, a number that says as much about your playback gear as it does about your ears. The vocal range test tool takes under a minute and needs nothing but your browser.
Every amplifier, speaker and pair of headphones comes with a THD figure, but few buyers know whether they could ever hear it. A THD distortion audibility test answers that by playing a clean tone, adding measured distortion in small steps and finding the point where your hearing notices the change. Here you learn how the audibility of distortion works, what masks it and how to try the test yourself.
Total Harmonic Distortion and Why Its Audibility Matters
Distortion is any unwanted change to an audio signal as it passes through a device. Total harmonic distortion measures one type: the extra frequencies your equipment adds at whole-number multiples of the original tone, expressed as a percentage of the sound you wanted. A lower figure means a cleaner signal, yet specifications on a data sheet say nothing about what your ears can detect. That gap between a measurement and perception is the reason audibility research exists, and it applies to every recording you play, whether analog, vinyl or digital. A spec chasing ever-lower numbers can easily go beyond what anyone hears, which hurts neither sound quality nor fidelity, but does mislead buyers. The audio codec blind training tool sharpens your ear with short, repeatable rounds.
Harmonics, the Fundamental and Odd Harmonics
Play a 220 Hz sine wave and the fundamental sits at 220 Hz. When a device bends that wave, harmonics appear at 440 Hz (second), 660 Hz (third), 880 Hz (fourth) and so on across the spectrum. Clipping a tone produces mostly odd harmonics, which are the 3rd, 5th and 7th, because the wave drifts toward a square shape. Each harmonic has its own level, and the total comes from combining them all.
Why a Pure Tone Reveals Distortion
A pure tone has no natural harmonics, so anything extra stands out against a silent background. That makes a pure sine signal the most sensitive probe for timbre changes, while a busy mix hides them. Loudness matters too: because your sensitivity peaks in the midrange, a louder-sounding fundamental covers more of the added harmonics.
Auditory Masking: How Music Hides Distortion
Auditory masking is the reason a measurable flaw can be inaudible. A loud sound raises your hearing threshold for nearby frequencies, so a weaker distortion component close to it is simply not heard. Masking is strongest near the masker and weaker far above or below it, which is why low harmonics hide inside loud music while high-order harmonics can stay exposed. Real program material also changes constantly, so each moment offers different cover. Each experiment with listeners and loudspeaker playback of music generally finds that detection needs far more distortion than a sine tone does, and the reverse holds for headphones playing a quiet test tone. The perception of the ear, not the meter, decides the outcome. Try the eq ear training online for a quick way to test your own listening discrimination.
Masking Curves and the Absolute Threshold of Hearing
Researchers plot masking curves to predict this. The curve shows how loud a nearby component must be before the ear notices it. Below the absolute threshold of hearing, a component is inaudible even in silence, and the normalized masking pattern shifts that line upward around any strong tone. Comparing each harmonic level in dB against the curve at a given SPL tells you whether it can be heard. The frequency range from 2 kHz to 5 kHz is where the ear is most forgiving of weaker components and where outer and middle ear response shapes the curve.
Hearing Changes with Age and Level
In psychoacoustics, one study after another shows that the hearing sensitivity of older listeners falls at high frequencies, which hides high-order harmonics sooner. Listening volume matters just as much, so a result at a quiet level may not hold at a loud one. Any THD result therefore applies only to the listener's age and the level used.
THD Audibility Thresholds by Frequency and Level
There is no single number for audibility thresholds, because they depend on the tone, its frequency, the playback level and the type of distortion. As a rule, the same THD percentage is easier to hear when it is made of high-order harmonics than of the second or third harmonic alone. Most ranges quoted for music sit between 0.1% and a few percent, and a clean pure sine tone can reveal distortion at lower values.
The decibel scale lets you compare the figures. Each step down by a factor of ten in percentage is 20 dB lower, so the values below match levels on a spectrum analyzer.
| THD (percent) | Level below the fundamental | What it usually means for listeners |
|---|---|---|
| 10% | -20 dB | Obvious on nearly any program material |
| 1% | -40 dB | Audible with a pure tone, often masked in music |
| 0.1% | -60 dB | Rarely audible outside tuned listening tests |
| 0.01% | -80 dB | Far below the masking threshold in practice |
The formula behind the table is simple. Combine the harmonic amplitudes, then divide by the fundamental:
$$\text{THD} = \frac{\sqrt{V_2^{2} + V_3^{2} + V_4^{2} + \dots}}{V_1} \times 100\%$$
For a worked case, suppose a 220 Hz tone shows a 2nd harmonic of 0.30%, a 3rd of 0.40% and a 5th of 0.12%. Then \(\sqrt{0.30^{2} + 0.40^{2} + 0.12^{2}} \approx 0.51\%\), or -45.8 dB. These harmonics fall at 440, 660 and 1100 Hz, so the third and fifth land where your ear is more sensitive.
Sine Tone Detection Threshold at Low Frequencies
Lower the sine tone frequency and the detection threshold gets stricter. The standard 1kHz test is the easiest on equipment: the tone sounds loud, which masks the harmonics, and few of them fall in the audible range. A 125Hz tone puts many more harmonics in the zone where your hearing is sensitive, and high order components, including those from -6 dBFS signals, become easier to catch. Testing across frequency shows far more than a single 1 kHz reading. Room noise matters here too: hiss from a noisy signal chain can cover the weakest harmonics, so a fair result needs a quiet playback signal.
Hard Clipping and Two-Tone Signals
Hard clipping is the harshest ordinary distortion, so researchers build hard clipped test files with an audio editor and measure THD for each gain step. A two-tone signal helps above 1 kHz, where a single sine cannot properly reveal the threshold. Brief transients hide it better than steady notes, and when audible it appears only as a slight change in timbre.
How to Run a THD Distortion Audibility Test at Home
A THD test at home needs a set of sound files with known distortion, a room with low background noise and honest expectations. Generate or download WAV files of one tone at rising distortion, for example 0%, 0.01%, 0.1%, 0.5%, 1% and 5%, each at the same loudness. Use the total harmonic distortion test files from a reputable test-tone site or build them yourself in an audio editor.
- Set a comfortable volume with the 0% file and leave the gain alone.
- Play each file in order, listening for a change in tone color or a rough edge.
- Note the first file where you hear a difference every time.
- Repeat with another frequency, such as 125Hz and then 1kHz.
Your own equipment limits the result. If your headphones already add 0.1% distortion, a file labelled 0% reaches your ears with 0.1%, and you cannot discriminate anything lower. Speakers and subwoofers add more. Treat the outcome as an upper bound of your system, not of your hearing.
Blind Test Setup and Listening Test Steps
A sighted comparison lies to you. For a proper blind test, have a friend or a tool shuffle the files so you do not know which is which. In each listening test round, play a pair of the same tone at different THD levels and guess the more distorted one. Ten rounds with at least nine correct is strong evidence. Repeat with loudspeakers or amplifiers swapped to see whether the equipment changes the listening result.
Reading Your Results Against the Threshold
If you reliably hear 0.5% but not 0.1%, your audibility threshold sits between them for that tone and level. Reviewers who report audible differences at much lower values should be asked about the signal, the level and whether the test was blind. Only a consistent, audible result counts.
Distortion Test Results with Music: Why Specs Mislead
A distortion test with real music tells a different story from one with test tones, because masking, dynamics and the harmonic structure of instruments all change what you hear. In the 1970s, adding heavy negative feedback drove amplifier THD to tiny numbers, yet transient intermodulation distortion made some designs sound harsh. A lower figure did not guarantee better sound quality. Specifications also hide dynamic range: an amplifier measured at one fixed output may behave differently at the loud peaks of a dynamic recording.
Low-Order Distortion, Tubes and Even Harmonics
Low-order distortion, such as second and third harmonics, is the mildest kind, and the ear tolerates it well. Tube circuits often add even harmonics that follow the note in musical intervals, so they can sound warm. Intermodulation distortion is different: it creates sum and difference frequencies that are not musically related, and an amplifier can trade one for the other. A single THD figure cannot capture that difference, which is why audibility tests use music as well as tones.
Nonlinear Distortion Measurements: SINAD and THD+N
Nonlinear distortion adds new components that were never in the source. Two common measurements summarize it. THD+N adds the residual noise to the harmonics. SINAD expresses signal plus noise plus distortion against noise plus distortion, and for a clean amplifier the two are close to the same number inverted. In the worked case above, adding 0.20% of random noise lifts the 0.51% figure to about 0.55%, or -45.2 dB.
Both measurement types use one frequency and level, and neither applies a weighting that mirrors how the ear works. A good measurement sweeps the audio band at high power and then steps the level down at a single frequency, so make sure a spec sheet says where it was taken before it tells you anything about sound.