We effortlessly distinguish a cello from a piano, or a clarinet from a trombone. Even when instruments play the same note at the same volume, each maintains its unique sonic identity.
This quality is known as timbre.
But where does this distinction originate? Is it solely within the sound itself, or does each instrument leave a distinct imprint on brain activity?
A new study, published in 2026 in the journal NeuroImage, revealed that a machine learning algorithm can identify which of five timbres a person is currently hearing with greater-than-chance accuracy, based on just a short segment of an electroencephalogram (EEG).
Electrical brain activity retains information not only about a sound's pitch and loudness, but also about its unique character.
One Note — Five Sonic Worlds
The study was conducted by specialists from Memorial University of Newfoundland in Canada: Pravina Satkunarajah, Sarah Power, and Benjamin Zendel.
The article was published online on May 19, 2026, and appeared in the August issue of NeuroImage.
Ten individuals with normal hearing participated in the experiment, including musicians, self-taught players, and those with no musical background.
Participants listened to the same note, A3, with a fundamental frequency of 220 hertz. This note was presented in five different timbres:
- piano;
- clarinet;
- cello;
- trombone;
- a pure sinusoidal tone without overtones.
Each sound lasted one second. The volume of all variants was normalized to ensure it provided no simple clue for either the participants or the subsequent analysis.
Each of the five sounds was presented 200 times, and brain activity was continuously recorded using EEG and 73 electrodes.
Therefore, the primary variable was timbre — the sonic characteristic by which we identify a sound's source.
What is Timbre?
Timbre is often referred to as the “color” of a sound, but this definition doesn't encompass its full complexity.
It's not merely a hue painted over a note.
It is the biography of sound.
The same note played on a piano, clarinet, or cello shares an identical fundamental pitch. However, numerous additional frequencies, or overtones, resonate alongside it. The quantity, strength, and distribution of these overtones create an instrument's unique spectral signature.
Other distinguishing characteristics also include:
- attack — the speed at which a sound begins;
- its rise and decay;
- the distribution of energy among harmonics;
- brightness;
- roughness;
- the resonance of the instrument's materials and body.
For a cello, sound is produced through the interplay of the bow, strings, wooden body, and the air enclosed within it.
With a clarinet, the musician's breath vibrates the reed, and the air column inside the instrument generates resonance.
On a piano, a hammer strikes a string, and the soundboard and body then amplify and transform the resulting vibration.
Timbre, therefore, is neither an embellishment nor a byproduct.
It represents the genesis of sound, the material through which vibration travels, and the motion that brings it into existence.
Can an Instrument Be Identified by EEG?
Researchers investigated whether a single EEG segment could reveal which timbre a participant was currently hearing.
This is no simple task. Humans process comprehensive acoustic information, whereas the algorithm only analyzes electrical signals recorded from the scalp.
Four machine learning algorithms were employed for data processing:
- linear discriminant analysis;
- Gradient Boosting;
- support vector machines;
- k-nearest neighbors.
The researchers compared several data types: raw EEG segments, characteristics of evoked potentials, and harmonic and spectral features.
If the algorithm were merely guessing one option out of five, the expected accuracy would be approximately 20%.
However, when analyzing raw EEG segments, all four models yielded better-than-chance results:
- Gradient Boosting — approximately 35%;
- linear discriminant analysis — approximately 34%;
- support vector machines — approximately 33%;
- k-nearest neighbors — approximately 26%.
While the accuracy was far from perfect, the primary objective was not flawless instrument recognition. Instead, it aimed to verify whether timbre information is present within an individual brain response.
The algorithm did not hear the original sound or analyze its acoustic recording; it only received EEG data.
Nevertheless, the electrical signals contained enough discernible information to identify the perceived timbre with greater-than-chance accuracy.
Study in PubMed
The Cello Left the Most Distinct Trace
The scientists also analyzed the N1 and P2 components of the auditory evoked potential—brief changes in the brain's electrical activity that occur shortly after a sound appears.
The N1 component was examined approximately 100 to 200 milliseconds after sound onset, while P2 was analyzed from 150 to 290 milliseconds.
Responses to different timbres varied in their onset time and amplitude.
The cello elicited the most pronounced P2 response, while the pure sinusoidal tone generated the weakest. The N1 response to the cello emerged earlier, whereas the clarinet produced the latest response.
This does not imply that the brain “prefers” the cello.
These differences may relate to attack characteristics, spectral composition, roughness, and harmonic distribution. For instance, the clarinet sound in the experiment exhibited the smoothest onset among the instrumental tones utilized.
Nevertheless, the results highlight a crucial point: different timbres do not elicit entirely identical neural responses.
When processing an instrument's sound, the brain distinguishes more than just pitch and loudness.
Humans Performed Significantly Better Than the Algorithm
The experiment participants also consciously identified the sounds they heard, achieving an average recognition accuracy of approximately 94%.
Participants identified the trombone most accurately, in approximately 97% of cases. The pure tone was recognized with about 96% accuracy, the cello with 95%, the clarinet with 92%, and the piano with 89%.
This comparison is significant.
The algorithm does not yet decode perception as confidently as a human recognizes a familiar sound. This study merely represents the initial step, demonstrating that timbre information is present even in short, isolated EEG segments and can be partially decoded.
Could Hearing Aids Understand Our Attention?
One potential practical application of this research pertains to hearing aids.
While modern devices are well-suited for speech amplification, music continues to pose a challenge for them. When multiple instruments play concurrently, numerous frequencies and dynamic nuances can blend together after processing, losing their natural quality or becoming harsh.
In the future, a hearing aid linked to brain activity monitoring could theoretically identify which instrument a person is trying to focus on and selectively amplify that specific sound.
For instance, if a listener focuses their attention on a cello within an orchestra, the device could help extract its melodic line from the overall musical stream.
However, such an application remains a distant prospect.
Only ten individuals participated in the experiment. Furthermore, isolated one-second tones were used, whereas real music is considerably more complex, involving instruments playing simultaneously, changing pitch, loudness, dynamics, and articulation.
Classification accuracy also remained moderate, ranging from 26% to 35%.
Therefore, this isn't a ready-to-use mind-reading technology, but rather an initial confirmation of a fundamental possibility: a brief brain response contains discernible information about the perceived timbre.
What Did This Discovery Add to the "Sound of the Planet"?
It revealed that consciousness perceives sound not merely as an impersonal frequency.
The brain differentiates vibrational properties linked to material, touch, breath, and the very manner in which sound is produced. A single note, when played through various instruments, transforms into a multitude of independent sonic worlds, each leaving its own distinct trace in neural activity.
Music doesn't arise solely from what is heard.
Rather, it is born from how, through what, and through whom the vibration passes.
A piano, cello, clarinet, and trombone can all play the same note, yet consciousness encounters each sound as a distinct presence.
Timbre is the individuality of vibration. It is the memory of material, the imprint of movement.
It's the way matter enters into music.
And the more attentively we listen, the clearer we perceive that the Planet doesn't merely repeat a single note.
Instead, it unveils an infinite diversity—expressed through wood, metal, strings, the musician's breath, and the perceiving consciousness of the listener.
