Musical Speed and Musical Emotion: What Does a Listener Rate?

A melody can remain recognizable when accelerated while acquiring a different emotional description. Does it sound more vital, more uneasy, or less peaceful? Answering also requires distinguishing emotion perceived in music from emotion felt by its listener. Someone can describe a piece as sad and still enjoy hearing it. A rating of musical emotional character is not a direct measurement of that person's entire mood.

A study published in May 2026 by Ulvhild Helena Færøvik and Karsten Specht examines ratings of music under changes in tempo and major or minor mode. Rather than comparing entirely different songs, the researchers use altered versions of original compositions. This creates a more controlled way to investigate speed and mode without letting a familiar artist, lyrics, or a personal memory supply every difference. The result is not a dictionary covering all music. It is an online listening task with specified materials and measures. [1]

Five compositions become forty stimuli

Each of five original compositions is arranged at four tempi and in two modes. The tempi are sixty, one hundred, one hundred twenty, and one hundred fifty beats per minute; the modes are major and minor. Eight versions of each composition produce forty musical stimuli. Each lasts twenty-one seconds. Simple melodies, constant rhythms, and avoidance of surprising musical changes support comparison of the selected properties.

That approach differs from comparing a well-loved slow song with an unfamiliar fast one. Such a comparison could vary melody, voice, lyrics, recording quality, and familiarity alongside speed. Versions derived from the same original material reduce the risk that the tempo question disappears among all those other differences. The methodological aim is not to recreate the full richness of ordinary listening. It is to make a limited set of changes easier to examine.

Control comes with a trade-off. The authors acknowledge that simple MIDI piano stimuli offer less everyday musical richness. Some participants said they did not regard the stimuli as music, while others reported enjoying them greatly. Simplifying the material does not guarantee an identical experience for every listener. The experiment balances controllable features against the variety that normally surrounds music.

Three dimensions are not one happiness scale

Participants rate each sound on three dimensions from one to seven. The study uses higher-order factors from the music-specific GEMS-9 framework: sublimity, vitality, and unease. These terms encompass more than their shortest everyday definitions. Sublimity relates to qualities including wonder, transcendence, tenderness, nostalgia, and peacefulness. Vitality relates to power and joyful activation. Unease particularly involves tension.

The dimensions are not three positions on a single pleasant-to-unpleasant line. A piece could be rated high in vitality and also high in unease. Fast does not automatically mean happy, and slow does not exclusively mean sad. The framework attempts to separate features that broad everyday labels can conceal. Nor should an unease rating be mistaken for a clinical assessment of the listener's anxiety. The measured response concerns an appraisal of the sound.

The authors say the study targets perceived emotion, while acknowledging that participants were simply asked to give opinions and that the procedure does not produce a perfectly pure separation from felt emotion. That qualification matters. Ratings document a judgment about the music; they do not independently capture every accompanying bodily or subjective state. Using the same scale in different research contexts does not make instructions and the intended target of measurement irrelevant.

Listening at home leaves some conditions unverified

The methods report an initial sample of 1,280 people. Recruitment uses snowball sampling through social media, with variation in age and musical experience. The study defines musicians through both more than six years of playing an instrument and identification as an amateur, semi-professional, or professional musician. This is an operational definition for the analysis, not a universal boundary separating every possible musical ability.

Participants are instructed to wear headphones, but the researchers cannot verify compliance in the online setting. Listening to every stimulus in full is not mandatory either. After supplying all three ratings, a participant can advance or choose to replay a sound. A twenty-one-second file therefore does not imply twenty-one seconds of listening by everyone. The duration of the supplied material and the behavior of the listener remain different quantities.

Control sounds include voice recordings and stimuli resembling white, pink, and brown noise rather than silence alone. The authors explain that the noises were later found to have more restricted frequency ranges than their conventional counterparts. They should not consequently be described as perfect standardized examples of those noises. Their function is to provide a non-musical auditory comparison. They are analyzed separately, not inserted into the tempo-by-mode models.

The controls also prevent music from being compared only with an absence of auditory activity. But their inclusion does not solve every interpretive problem. A voice recording and a noise sample differ from a composition in several ways. The comparison helps describe ratings under this particular arrangement; it is not proof that all possible differences between musical and non-musical experience have been isolated.

Which emotional dimension changes most clearly?

For each participant, ratings of the five compositions within a tempo–mode combination are averaged. This produces eight condition values for each emotional dimension. The analysis concerns the pattern across these conditions rather than the distinctive impact of one particular composition. Faster tempi are associated with higher vitality ratings. Major versions receive higher vitality ratings than minor versions, and the increase with tempo is more pronounced in major mode. Speed and mode therefore need to be interpreted together.

For unease, the strongest main distinction concerns mode: minor versions are rated higher than major versions. Tempo contributes a smaller effect on this dimension. Slower versions are associated with higher sublimity ratings, but the reported sublimity effects are small. Saying that tempo changes emotional ratings does not mean that every dimension changes equally strongly. The dimension under examination affects what the finding means.

This is also why statistical significance and effect magnitude require separate attention. A large sample can make a small difference statistically detectable. The results section discusses effect sizes rather than treating every significant comparison as equally important. At the same time, a large sample reached through social media does not automatically become a random representation of an entire population. Recruitment and sample size address different questions.

The paper additionally contains a mismatch between the initial participant count and counts in several figure captions. This article uses the methods' stated initial count and does not invent a reconciled exact analysis sample. The paper explains that effective sample sizes vary across models because of missing responses. These reporting details are a reason for precision in description, not a license to declare that every finding has either been independently confirmed or invalidated without checking the underlying data.

The evidence-based conclusion is not that fast music makes everyone feel the same emotion. In these controlled compositions, tempo is especially informative for vitality ratings, while mode makes a particularly clear difference to unease ratings. The same alteration can have effects of different magnitudes across dimensions. A listener's emotional description is not simply a count of beats. It is a judgment made through the combination of musical features and available emotion terms. The study's contribution is to move beyond a single happy-song label and ask which property varies with which appraisal.

Sources

1. Færøvik UH, Specht K. (2026). The effect of tempi and mode on the rating of the perceived emotion in music. PLOS ONE 21(5): e0348069. Open primary research article.
2. Creative Commons Attribution 4.0 license.

Reviewed October 14, 2026. The research, © 2026 Færøvik and Specht, is credited under CC BY 4.0. This independently written explanation reorganizes the material and adds editorial interpretation. Audio, figures, and direct quotations are not reproduced. Raw data and supplementary files were not reanalyzed; no music-therapy or personal mood-alteration procedure is recommended.