Which instrument should play here? Decoding predicted musical timbre from EEG signals during omission

Area of Science:

• Neuroscience
• Cognitive Science
• Music Perception

Background:

• The human brain actively predicts musical elements like harmony, melody, and rhythm.
• Previous electroencephalographic (EEG) studies confirmed pitch prediction from EEG signals based on predictability.
• The specificity of prediction for other musical features, like timbre, remained underexplored.

Purpose of the Study:

• To investigate if predicted musical instruments (timbre) can be decoded from EEG signals during auditory omission.
• To examine the specificity of brain prediction within the timbre dimension of music.

Main Methods:

• Thirty-five participants listened to melodies with simple (high predictability) or complex (low predictability) timbre rules.
• EEG data were recorded during the omission of tones with expected timbres (celesta, electric piano, marimba, organ).
• Support vector machine (SVM) analysis was used to decode timbre information from EEG signals.

Main Results:

• No significant difference in the omitted stimulus potential (oN1) amplitude between high and low predictability conditions.
• SVM successfully decoded musical timbre from EEG signals above chance in the high predictability condition.
• Timbre decoding was not successful above chance in the low predictability condition.

Conclusions:

• EEG signals contain information about predicted musical timbre during omission.
• This timbre prediction is not reflected in traditional event-related potentials like oN1.
• The brain appears to predict musical timbre specifically, similar to its prediction of pitch.