Music, Imagery & Creativity: What Improvisation Research Shows

The short answer

Imagining music engages much of the same auditory machinery used to hear it, which is why a remembered melody can feel almost audible. Music also commonly evokes visual imagery in listeners, though how much varies widely between people and with the music. Studies of musical improvisation - most famously fMRI scans of jazz pianists - report a characteristic pattern of altered prefrontal activity during spontaneous performance compared with rehearsed playing. Those studies involve small samples of highly trained musicians performing under unusual constraints, and reviews of the field emphasise inconsistency across studies. They describe what happens in a trained brain during improvisation. They do not show that listening to music, or watching visuals generated from music, makes people more creative.

Why this matters for sound and music

Creativity is where sound-and-brain writing most often leaves the evidence behind. The real findings about imagery and improvisation are more specific and more interesting than the claim that music unlocks creative potential.

Imagining music uses the machinery of hearing

Most people can bring a familiar tune to mind with some sense of its pitch, timbre and timing. Imaging work on this 'mental concert' finds that auditory imagery engages secondary auditory areas overlapping those recruited during actual listening, with frontal regions involved in generating and maintaining the imagined material.

The overlap is partial rather than total, and imagery is generally weaker and less detailed than perception. Musicians tend to show more vivid and more accurate auditory imagery than non-musicians, consistent with the wider finding that training sharpens auditory representations.

This is a useful anchor for the section, because it establishes that internally generated sound is not merely metaphorical. It is a real, measurable process with a partly shared neural basis.

• Auditory imagery

  Hearing music in the mind; overlaps with perceptual auditory areas.

• Visual imagery

  Seeing scenes or colours while listening; common but variable.

• Generation

  Frontal regions initiate and hold imagined material in mind.

• Expertise

  Trained musicians typically report and demonstrate more precise imagery.

Music-evoked visual imagery

Many listeners spontaneously experience visual imagery during music - landscapes, memories, abstract movement or colour. Reviews of this literature find it to be a common and legitimate object of study, associated with absorption in the music and with emotional response.

The variability is substantial. Some listeners report rich visual scenes; others report almost none while still finding the music deeply affecting. Familiarity, whether the music has lyrics, listening context and individual imagery ability all contribute. Because most studies rely on self-report, quantitative comparisons across studies should be treated cautiously.

Note that this is imagery generated by the listener, and it is distinct from synesthesia, which requires involuntary, consistent and idiosyncratic experiences. It is also distinct from a visualizer, which supplies an external image derived from the audio signal rather than one produced by the listener's own mind.

What improvisation studies found

The best-known study scanned jazz pianists playing a specially built keyboard while alternating between memorised sequences and free improvisation. During improvisation the researchers observed reduced activity in lateral prefrontal regions associated with self-monitoring and evaluation, alongside increased activity in medial prefrontal cortex associated with self-generated, internally driven behaviour. The intuitive reading - internal editing quietens while self-expression rises - is appealing and widely repeated.

The caveats are equally important. The sample was six professional musicians. Improvising in a scanner, lying still, on an unfamiliar instrument, is not improvising on a stage. And the comparison is between spontaneous and rehearsed performance in trained experts, not between creative and uncreative people.

Later reviews of the neuroscience of improvisation describe a field with genuinely interesting convergences but meaningful inconsistency across studies in prefrontal findings, attributable partly to differences in task design, expertise level and analysis choices. The honest summary is that improvisation reliably differs from rehearsed performance in prefrontal engagement, while the precise pattern and its interpretation remain open.

• Sample sizes are typically under twenty and often under ten.
• Participants are usually highly trained musicians, limiting generalisation.
• Scanner constraints alter the task substantially.
• Prefrontal findings differ across studies; reviews report inconsistency.

Where the claims outrun the evidence

Three inferential leaps are common. The first is from expert improvisers to everyone: a finding about trained jazz pianists says little about a listener with no musical training. The second is from producing music to consuming it - nothing in the improvisation literature concerns listening. The third is from domain-specific musical creativity to general creativity, which the studies never measured.

There is separate research on whether background music affects creative task performance, and it is genuinely mixed, with effects depending on the task, the music, the noise level and the individual. It does not support a simple prescription.

For SonicSenses this sets a clear boundary. Making visuals from audio can be an engaging creative activity, and people do use it as one. That is a statement about what users do with the tool, not a claim that using it enhances creativity, imagery ability or brain function. Our tools are creative and educational, not cognitive interventions.

What we know

• Auditory imagery engages auditory cortical regions that overlap with those used in perception.
• Musicians generally show more precise auditory imagery than non-musicians.
• Music-evoked visual imagery is common, though highly variable between listeners.
• Improvisation is associated with altered prefrontal activity compared with rehearsed performance in trained musicians.

What remains uncertain

• Whether the reported prefrontal pattern during improvisation replicates consistently across tasks and expertise levels.
• How scanner conditions distort improvisation relative to real performance.
• What determines why some listeners experience rich visual imagery and others almost none.
• Whether any of these findings generalise beyond trained musicians.

What this does not prove

• That listening to music makes you more creative.
• That watching audio-reactive visuals enhances imagery or creativity.
• That improvisation research identifies a general creative state anyone can enter.
• That musical creativity findings transfer to creativity in other domains.

Practical meaning

• If music brings images to mind, that is a common and well-documented experience, not a special ability.
• Improvisation research is about performing, not listening - treat claims that reverse this as unsupported.
• Creating visuals from your own audio is a creative practice worth enjoying on its own terms.
• Be sceptical of any product promising to unlock creativity through sound or visuals.

Frequently asked questions

Does imagining music use the same brain areas as hearing it?

Partly. Auditory imagery engages secondary auditory regions overlapping those active during listening, with frontal areas involved in generating and holding the imagined material. The overlap is partial, and imagined music is usually less detailed than heard music.

Why do I see images when I listen to music?

Music-evoked visual imagery is common and is associated with absorption and emotional engagement. It varies widely between people and pieces. It is generated by you, which distinguishes it both from synesthesia and from an external visualizer.

What did the jazz improvisation brain study actually show?

In six professional pianists, free improvisation was associated with reduced lateral prefrontal activity and increased medial prefrontal activity compared with playing memorised material. It is a small, influential study conducted under scanner constraints, and later reviews report inconsistency across the wider improvisation literature.

Does listening to music make you more creative?

The evidence does not support that as a general claim. Improvisation research concerns musicians producing music, not listeners. Studies of background music and creative task performance are mixed and depend on the task, the music and the person.

Can a visualizer boost creativity?

There is no evidence for that. Making and watching visuals from audio can be an enjoyable creative activity, but SonicSenses does not claim to enhance creativity, imagery or cognition.

References & further reading

1. Zatorre, R. J., & Halpern, A. R. (2005). Mental concerts: musical imagery and auditory cortex. Neuron DOI: 10.1016/j.neuron.2005.06.013
2. Halpern, A. R., & Zatorre, R. J. (1999). When that tune runs through your head: a PET investigation of auditory imagery for familiar melodies. Cerebral Cortex DOI: 10.1093/cercor/9.7.697
3. Limb, C. J., & Braun, A. R. (2008). Neural substrates of spontaneous musical performance: an fMRI study of jazz improvisation. PLoS ONE DOI: 10.1371/journal.pone.0001679
4. Beaty, R. E. (2015). The neuroscience of musical improvisation. Neuroscience & Biobehavioral Reviews DOI: 10.1016/j.neubiorev.2015.01.004
5. Taruffi, L., & Küssner, M. B. (2019). A review of music-evoked visual mental imagery: conceptual issues, relation to emotion, and functional outcome. Psychomusicology: Music, Mind, and Brain DOI: 10.1037/pmu0000226

Related articles

• Visualizing Sound: How the Brain Integrates What We Hear and See

  How hearing and vision influence each other, and what a visualizer really shows.

• Musical Training & Neuroplasticity

  What years of practice change, and what remains association rather than cause.

• Music, Reward & Dopamine

  Why music is rewarding, and why dopamine language is usually wrong.