Detection of partially filled gaps in noise and the temporal modulation transfer function

Area of Science:

• Auditory Neuroscience
• Psychoacoustics
• Signal Processing

Background:

• Understanding the human auditory system's temporal resolution is crucial for explaining speech perception.
• Previous research established basic gap detection capabilities, but detailed modeling of temporal processing remained incomplete.

Purpose of the Study:

• To investigate the detection thresholds for silent intervals (gaps) in broadband noise for normal-hearing listeners.
• To compare gap detection thresholds for decrements and increments in noise.
• To evaluate auditory system models based on gap detection and temporal modulation transfer functions.

Main Methods:

• Experiments measured gap detection thresholds in broadband noise for normal-hearing participants.
• Thresholds for noise decrements and increments were determined.
• Temporal modulation transfer functions were measured and compared with gap detection data.

Main Results:

• A gap detection threshold of approximately 2 ms was observed, independent of noise burst duration or level variations.
• Increment detection thresholds were slightly higher than decrement thresholds, consistent with a fixed peak-to-valley detection ratio model.
• Auditory system models, including Viemeister's three-stage model, showed good agreement with experimental gap detection and modulation data.

Conclusions:

• The auditory system exhibits precise temporal resolution, capable of detecting gaps as short as 2 ms.
• Existing models provide a robust framework for understanding temporal processing in hearing, with parameters like low-pass filter time constants (3-8 ms) being key.
• Simulations of established temporal models accurately replicate human performance in both gap detection and temporal modulation tasks.