Binaural recordings of 400 Hz pure-tones played from a loudspeaker in my dining room. Analyses show that ITD and ILD resemble ‘ideal’ values before recent reflections arrive, after which values begin to drift. Contributions to spatial hearing are primarily from cues emerging at the onsets of the tones, before early reflections arrive.
Interaural time differences (ITDs) computed across zero-crossings after passing each recorded tone through a 400 Hz auditory filter (circles in upper panel). Red and blue lines in the lower panel show right and left amplitude envelopes, respectively.
- • ITDs fluctuate broadly prior to when each pure-tone starts (before 0 ms).
- • Values then contract for ~20 ms as amplitude rises, as would occur in a quiet, anechoic, setting.
- • Due to acoustic reflections and background sounds, values begin to drift and fluctuate after ~20 ms, becoming more pronounced over the duration of each tone (200 ms) .
When listening to recordings over headphones, it becomes evident that cues near the onsets of the pure-tones contribute substantially to spatial hearing. Below, for instance, are recordings for 0° azimuth, cropped 1 second before and after the recorded tone’s onset. What a listener is likely to hear is a single compact auditory ‘image’, from a frontal direction, that remains in roughly the same position when channels for the left and right ears are reversed.
channels reversed
That cues near the onsets of the pure-tones contribute more to spatial hearing than later cues is further evident when recordings are cropped so that they start 50 ms after each recorded tone’s onset, thereby removing cues at the tone’s onset. What a listener is likely to hear, in this case, is a distinct leftward auditory ‘image’ and a distinct rightward ‘image’ when channels for the left and right ears are reversed.
channels reversed
Pure-tones (400 Hz) with 10 ms on/off ramps (Hanning) were played from a single loudspeaker in my dining room well above average ambient noise levels (~36 dB). Ten recordings were obtained at random times over 2-3 days using the Neumann KU100 dummy head rotated in 10° azimuthal increments during each recording session.
The Neumann KU100 dummy head and speaker were positioned 2.8 m apart and 1.35 m above a hardwood covered concrete floor in my 5.25 X 4.2 X 2.5 m dining room. A dinning table, 2 X 1 X 0.75 m, was positioned between the microphone and speaker, all slightly offset in the room so that the left microphone was 1.7 m from a large shuttered window along the lefthand wall and 2.4 m from a solid righthand wall. A passageway behind the speaker allowed sound to travel through 2 rooms and a foyer to another passageway behind the microphone. All of these factors contribute to a binaural room impulse response (BRIR) that is complex and somewhat long.
Binaural room impulse responses (BRIR) for single loudspeaker recorded with the Neumann KU100 dummy head at 0° azimuth.
Tap or move cursor over element to pause slideshow
