Target-masker onset asynchrony modulates linguistic release from masking
Speech-in-speech recognition is a challenge that listeners often encounter. Performance in such situations is improved if the competing speech is in a language different from the target—a phenomenon called Linguistic Release from Masking (LRM). LRM occurs due to a combination of energetic masking th...
Saved in:
| Published in: | The Journal of the Acoustical Society of America Vol. 155; no. 3_Supplement; p. A80 |
|---|---|
| Main Authors: | , , , |
| Format: | Journal Article |
| Language: | English |
| Published: |
01-03-2024
|
| Online Access: | Get full text |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | Speech-in-speech recognition is a challenge that listeners often encounter. Performance in such situations is improved if the competing speech is in a language different from the target—a phenomenon called Linguistic Release from Masking (LRM). LRM occurs due to a combination of energetic masking that arises from the physical overlap between the target signal and the masker and informational masking that arises from cognitive, attentional, and other factors. The contributions of informational masking to LRM is poorly understood. In the current study, we manipulated informational masking by varying the target-masker onset asynchrony while holding energetic masking constant. Typical, monolingual English listeners transcribed BKB sentences with either English or Mandarin two-talker babble in the background. The target sentences started either simultaneously with the masker, 500 ms after the masker, or 1000 ms after the masker. Listening accuracy was higher for the Mandarin than for the English masker, replicating typical LRM. However, the size of LRM was moderated by onset asynchrony with larger effects for longer lag times. This finding suggests that the detrimental effect of English masker accrues with lag. Implications of this finding for the role of informational masking in speech-in-speech listening will be discussed. |
|---|---|
| ISSN: | 0001-4966 1520-8524 |
| DOI: | 10.1121/10.0026877 |