Fig 1.

Inter-trial correlation analysis (ITC) A. Traces of single trials from one sample electrode, filtered between 1-10 Hz. The top panel shows the similarity in the time course of the neural response in two single trials (blue and red traces) where the same stimulus was presented versus two trials in which different stimuli were presented in the Single Talker condition. The bottom panel demonstrates that a similar effect is achieved by shifting attention in the Cocktail Party condition. Two trials in which attention was focused on the same talker elicit similar neural responses, whereas trials in which different talkers were attended generate different temporal patterns in the neural response, despite the identical acoustic input. B. Top: Phase coherence spectrum across all channels for Single Talker trials where the same stimulus was presented (red) vs trials where different stimuli were presented (chance level; gray). Phase-coherence for repetitions of the same stimulus was significant only at frequencies <7 Hz. Bottom: Percentage of electrodes with significant phase-ITC (red) and power-ITC (blue) in the Single Talker (empty bars) and Cocktail Party (full bars) conditions, in each frequency band. Significant phase-ITC was found dominantly in the low-frequency range (delta and theta), whereas significant power-ITC was mostly limited to the high-gamma range. C. Location of sites with significant LF phase-ITC (left) and HG power-ITC (right) in both conditions. The colors of the dots represent the ITC value at each site. D. Top: Average ITC in the Single Talker condition across trials in which the same stimulus was presented (black) vs. trials in which different stimuli were presented (gray). Bottom: Average ITC in the Cocktail Party condition across trials in which the same talker was attended (black), trials in which the same pair of talkers was presented but different talkers were attended (blue) and trials in which different pairs of talkers were presented. Here and elsewhere errorbars reflect SEM.