Perceptual compensation for effects of reverberation in speech identification

Abstract

Listeners were asked to identify modified recordings of the words "sir" and "stir," which were spoken by an adult male British-English speaker. Steps along a continuum between the words were obtained by a pointwise interpolation of their temporal-envelopes. These test words were embedded in a longer "context" utterance, and played with different amounts of reverberation. Increasing only the test-word's reverberation shifts the listener's category boundary so that more "sir"-identifications are made. This effect reduces when the context's reverberation is also increased, indicating perceptual compensation that is informed by the context. Experiment 1 finds that compensation is more prominent in rapid speech, that it varies between rooms, that it is more prominent when the test-word's reverberation is high, and that it increases with the context's reverberation. Further experiments show that compensation persists when the room is switched between the context and the test word, when presentation is monaural, and when the context is reversed. However, compensation reduces when the context's reverberation pattern is reversed, as well as when noise-versions of the context are used. "Tails" that reverberation introduces at the ends of sounds and at spectral transitions may inform the compensation mechanism about the amount of reflected sound in the signal.

Amplitude modulation detection by human listeners in reverberant sound fields: Effects of prior listening exposure

“…Such enhancement is consistent with recent neural data that demonstrate the presence of reverberation-resistant coding of AM in the response properties of single neurons in the inferior colliculus (Delgutte et al, 2012;Kuwada et al, 2012). Because other studies have demonstrated that speech understanding in reverberant soundfields may be modified (Watkins, 2005) or improved (Brandewie and Zahorik, 2010;Srinivasan and Zahorik, 2013) with prior listening exposure to reverberation, and that this effect appears to most strongly influence the processing of the amplitude envelope of the speech signal (Watkins et al, 2011), it is important to determine whether AM sensitivity could be similarly affected by prior listening exposure to reverberation. This is the goal of the present study.…”

Revisiting perceptual compensation for effects of reverberation in speech identification

“…The influence of reverberation on the identification of the stop-consonant ͓t͔ was assessed using a continuum of 11 words, changing in steps from plain "sir" ͑step 0͒ to plain "stir" ͑step 10͒. The words were generated by calculating the envelopes of "sir" and "stir" and imposing different ratios of these envelopes on the waveform of "sir" ͑refer to Watkins, 2005c, for details͒. The test-words of the continuum were perceived as "sir" at low step numbers and as "stir" at high step numbers.…”

“…Watkins also characterized the effect as extrinsic because it depends on information about the reverberation from surrounding speech and not the word itself. In his studies, Watkins investigated how the proposed compensation mechanism may pick up reverberation information from the context in order to increase speech intelligibility ͑e.g., Watkins, 2005cWatkins, , 2005bMakin, 2007a, 2007b͒. The "tails" that reverberation adds to the offsets of sounds seem to be particularly important.…”

“…The attenuation of amplitude modulations in a room is an accurate predictor of speech intelligibility, and this correlation is reflected in methods for assessing intelligibility, such as the speech transmission index ͑STI; Steeneken and Houtgast, 1980͒. In a series of studies ͑e.g., Watkins, 2005cWatkins, , 2005bMakin, 2007a, 2007b͒, Watkins considered the reduction of amplitude modulations in investigations of the impact of reverberation on speech intelligibility. He tested the ability of listeners to identify the stop-consonant ͓t͔ in the test-word "stir" when the word was embedded in a carrier sentence and different amounts of reverberation were added.…”

Evidence for a neural source of the precedence effect in sound localization

“…Thus, provided the delay between two or more successive similar or identical sounds is sufficiently brief, a single sound is perceived near the location cued by the leading sound (Wallach et al 1949). In addition to providing for accurate sound source localization, the PE is thought to aid sound identification in reverberant environments, including (in humans) speech intelligibility (e.g., Brandewie and Zahorik 2010;Watkins 2005). The PE is a thus a key facility of normal hearing.…”