Understanding speech in noise (SPiN), particularly when the masker is also speech, is a complex ability that relies on multiple acoustic and cognitive cues, among which amplitude modulation (AM) plays a crucial role. Previous studies have demonstrated that sensitivity to AM is crucial for encoding speech envelopes and segregating competing sound sources. Beyond basic AM detection (AMD), modulation-detection interference (MDI) reflects how the perception of a target modulation is disrupted by concurrent modulations in other frequency channels. While AMD and MDI have been studied extensively, their relationship to SPiN remains poorly characterized. This study quantifies MDI in adults and examines how it relates to individual differences in SPiN intelligibility.

To date, twenty normal-hearing adults aged 18–29 years have participated in the study. Data collection is ongoing, additional adolescent participants is expected to be included by the time of the conference. Speech intelligibility thresholds (SRT₅₀) were measured using a French version of the Coordinate Response Measure, in which participants have to identify a target sentence (color–number combination) among two conditions: two competing female voices (speech-on-speech condition) and envelope-modulated speech-shaped noise (eSSN). Participants also completed a 3-interval alternative forced-choice (3IAFC) AM detection task. The target and masker were each carried by distinct pure tones at different carrier frequencies. The target carrier was modulated at 2.62 Hz, corresponding to the dominant modulation rate of the CRM target voice, while the masker carrier was modulated in six conditions: (1) no masker, (2) unmodulated masker, (3) 2.62 Hz (same rate as target), (4) 3.55 Hz (dominant rate of CRM maskers), (5) 16 Hz, and (6) 32 Hz.

We expect the magnitude of MDI to depend upon the distance between target and masker modulation rates. Specifically, AM detection thresholds are expected to be lowest for unmodulated or high-rate (16 & 32 Hz) maskers and highest when the masker shares the same modulation rate as the target (2.62 Hz), reflecting maximal interference between overlapping amplitude modulation filter banks. The magnitude of this interference is anticipated to reflect the listener’s ability to segregate competing envelopes and the level of internal modulation noise. We also expect individual differences in MDI to relate to SPiN (SRT₅₀), particularly in the speech-on-speech condition, highlighting the importance of envelope-processing mechanisms for SPiN. Data collection in adults is ongoing, and these results will serve as a baseline for future developmental work investigating how MDI contributes to lasting SPiN difficulties in adolescents.