12 VOLT ENCYCLOPEDIA · VOLUME VI
Speaker Distortion Graphs: What Harmonics and Intermodulation Tell You
Learn to read speaker distortion evidence without mistaking a smooth response curve—or one small percentage—for the whole story.
A smooth response line is only part of the story. Frequency response describes how strongly a speaker reproduces frequencies. Distortion measurements ask what else it adds while doing that job. Neither graph should be read without knowing the test conditions.
Start with the signal that was actually played
Suppose a test sends a clean 1 kHz tone. A linear system reproduces that frequency without generating new ones. A real driver can also produce energy at multiples such as 2 kHz and 3 kHz. These are the second and third harmonics. A graph may show each harmonic separately, combine them into total harmonic distortion, or express their level relative to a reference.
Audio Precision distinguishes THD from THD+N: the latter includes noise as well as distortion. A percentage therefore needs its definition, measurement bandwidth and reference. Two plots bearing the word “distortion” can be describing different quantities. Do not compare their percentages as if the labels were interchangeable.
Swipe horizontally to view the full diagram, or focus the diagram and use the arrow keys.
Read the axes before judging the speaker
The horizontal axis might show the frequency of the test tone, not the frequency at which every plotted harmonic was radiated. At a 1 kHz test point, the H3 trace concerns a product at 3 kHz. That distinction matters near a driver’s bandwidth limits. Read the software documentation or test author’s method before interpreting a peak.
REW displays the fundamental, individual harmonics, total harmonic distortion and a measured noise floor. Its documentation also explains that sweep length and measurement method affect how accurately distortion can be separated. A trace approaching the noise floor should not be treated as precise evidence of an exceptionally clean speaker.
Compare the same job, not merely the same voltage
Imagine two midranges measured with the same input voltage. If one produces a higher sound level, it is being asked to do more acoustic work at that setting. Now imagine matching their sound levels instead. That answers a different question: how cleanly does each deliver the output you need? Neither comparison is inherently dishonest, but the test must identify which comparison it makes.
Make a comparison card before choosing a winner: microphone distance and angle, acoustic level or input voltage, baffle or enclosure, crossover settings, frequency range and the plotted metric. A missing item is a question to ask, not an invitation to fill in a favorable assumption. A speaker tested quietly on a large laboratory baffle is not automatically comparable to a driver tested loudly through a restrictive factory grille.
| Before comparing | Ask for | Why it matters |
|---|---|---|
| Drive level | Input voltage and/or acoustic level | Distortion behavior changes with demand |
| Geometry | Distance, angle, baffle or enclosure | The measurement describes that setup |
| Filtering | Applied HPF/LPF and test range | A protected driver is not a full-range test |
| Metric | Individual harmonics, THD or THD+N | They do not include identical components |
| Confidence | Noise floor and repeatability | Tiny values can approach the test limit |
Why one clean tone does not settle everything
Music contains many simultaneous frequencies. Intermodulation testing applies multiple tones and examines additional components created by their interaction. Klippel describes two-tone measurements that reveal amplitude and frequency modulation, and explains that some nonlinear behavior is not fully characterized by harmonic measurements alone. That is why “low THD” should not be expanded into “no distortion of any kind.”
Swipe horizontally to view the full diagram, or focus the diagram and use the arrow keys.
Keep the result tied to the intended use
For a door midbass, examine the region it will actually reproduce at the output you expect. An impressive high-frequency measurement does not settle its low-frequency capability. For a tweeter, do not extrapolate results taken with a protective high-pass into permission to run it lower. Treat the tested configuration as part of the result.
Klippel’s measurement overview separates linear response, nonlinear distortion and other measurement categories. That organization is useful for buyers: several tests answer several questions. A distortion graph does not replace dispersion, fitment, usable bandwidth or installation checks.
A practical reading sequence
- Identify the driver, test setup and intended operating range.
- Check the fundamental output and whether competing tests use comparable level.
- Read individual harmonic trends rather than only one headline percentage.
- Look for a documented noise floor and repeatable behavior.
- Check whether multi-tone evidence exists when a broad passband is important.
- Use the result alongside installation requirements and controlled listening, not instead of them.
This is a guide to interpreting published evidence, not a request to run destructive sweeps at home. High-level tones can damage drivers and hearing. If a report omits its conditions, the right conclusion is “not enough information,” not “the lower number wins.”
