Effects Of Early Reflections Research Articles

How accurate dereverberation should be? Effects of underestimating and overestimating binaural room impulse responses

The sound arriving at the listeners' ears is influenced by the binaural room impulse response (BRIR) of the listening environment. Previous research (Srinivasan et al., 2017) has suggested that removing late reflections in BRIR improves speech understanding most when spatial cues are absent. However, in real world listening scenarios, it is difficult to differentiate between the effects of early reflections and late reverberation. Here, we present data from an experiment evaluating the effects of a simple static dereverberation technique on speech understanding for two reverberant environments (T60 = 1 and 2 s). The input speech signal was dereverbered by deconvolving the speech signal under three different conditions: (1) underestimating the effects of reverberation, (2) overestimating the effects of reverberation, and (3) correct estimation of the effects of reverberation. Effects of the three deconvolving techniques on identification thresholds and amount of release from masking will be discussed.

Using individually equalized playback of binaural recordings from an electronic orchestra to determine the effects of early reflections on proximity, localization, and loudness

Reflections in the first 50 ms are often assumed to enhance sound quality, but data supporting the assumption are lacking. We need to know when reflections are beneficial, and when they are not. Work by Lokki et al. may help to answer this question, but we have developed a binaural recording and playback technique that yields as good or better results. The key is individual equalization from the microphone to the eardrum of a listener. We have developed a quick and non-invasive method of headphone equalization that produces accurate timbre and frontal localization without head-tracking. This enables binaural impulse responses from a number of instrument positions in real halls to be manipulated by adding or deleting individual reflections and then evaluated with blind listening tests. Data from Boston Symphony show that early lateral reflections in the front of the hall are typically inaudible, but early medial and lateral reflections in the rear of the hall can reduce or eliminate proximity, clarity, impact, envelopment, and the ability to localize individual instruments. Deleting one or two of the earliest and strongest reflections improves all of those perceptions with no effect on loudness.

Speech intelligibility in reverberation with ideal binary masking: Effects of early reflections and signal-to-noise ratio threshold

Ideal binary masking is a signal processing technique that separates a desired signal from a mixture by retaining only the time-frequency units where the signal-to-noise ratio (SNR) exceeds a predetermined threshold. In reverberant conditions there are multiple possible definitions of the ideal binary mask in that one may choose to treat the target early reflections as either desired signal or noise. The ideal binary mask may therefore be parameterized by the reflection boundary, a predetermined division point between early and late reflections. Another important parameter is the local SNR threshold used in labeling the time-frequency units as either target or background. Two experiments were designed to assess the impact of these two parameters on speech intelligibility with ideal binary masking for normal-hearing listeners in reverberant conditions. Experiment 1 shows that in order to achieve intelligibility improvements only the early reflections should be preserved by the binary mask. Moreover, it shows that the effective SNR should be accounted for when deciding the local threshold optimal range. Experiment 2 shows that with long reverberation times, intelligibility improvements are only obtained when the reflection boundary is 100 ms or less. Also, the experiment suggests that binary masking can be used for dereverberation.

Effects of ear coverage and reflected sound on the localization of sound

Previous research has shown that listeners wearing helmets with greater ear coverage perform worse on localization tasks. However, the helmets studied were selected from those commonly in use and differed in suspension systems and profiles. Three versions of the same helmet design, differing only in the ear coverage (0%, 50%, 100%), were compared in this study. Sounds were presented from 12 azimuthal locations spaced 30 deg apart in the presence of a white noise masker. Twelve listeners completed a localization task while wearing each of the helmets as well as with no helmet in two different acoustic environments. As ear coverage increased, localization performance decreased. The effects of early reflections, though small, were consistent across the different coverage conditions. The effect of ear coverage and its interaction with the acoustic environment will be discussed in terms of its significance for real-world environments. In addition, head-related transfer-functions (HRTFs) were measured for each listener in each of the head conditions. A model of localization performance based on changes to the HRTF will be described and applied to the collected data.

Investigation of early reflections on the perception of audio

The effects of early reflections in a nonsymmetrical control room were measured using time delay spectrometry and MLS methods. The tests revealed that comb-filtering patterns were the dominant source of spectral distortion. The mixing console reflections were investigated to reveal the subjective impact of the comb-filtering at the mix position. Tests were performed with and without acoustic foam on the console to determine the influence of the early reflections and then compared with theoretical values for the filtering patterns. Results were compared and used to qualify the subjective perception of the distortion.

Experimental study on the effect of early reflections on players by binaural simulation technique.

Experimental study on the effect of early reflections on players by binaural simulation technique.

Evolution of a concert hall: Lateral reflections and the acoustical design for Wellington Town Hall

The evolution of research and acoustical design objectives which have been the basis for the preliminary design phase of the Wellington (NZ) Town Hall is reported. The undoubted success Christchurch Town Hall with four years of evaluation in intensive use, and with the progress of research together indicated the acoustical refinements and modifications to be included in the Wellington Town Hall design. There has become a growing recognition of the fact that the directional properties of early reflections are as important in the perception of sound in concert halls as their temporal distribution. Marshall's early papers on the importance of lateral reflections led to a comprehensive study on the effects of early reflections on subjective acoustical quality in concert halls by M. F. E Barron at Southampton. The study includes details of masking of both high‐ and low‐frequency orchestral sound in lateral reflections, integration time, quantitative prediction of “Spatial Impression,” image shifting, and the difference limen. Results of other studies are presented which give complementary results in listener preference for relative levels of lateral and frontal energy. While model studies and in situ evaluation have shown that conventionally fan‐shaped halls cannot usually provide the preferred levels of envelopmental energy, acoustical experience in Christchurch has shown that in halls of this type, the preferred relative levels of lateral sound are achievable. Specific design features are presented including the use of longitudinally oriented “Quadratic Residue Diffusers” which provide multiple lateral reflections. The result is an unpresidented design proposal.

Reflective Surfaces for the Hearing of Speech and Music in an Anechoic Chamber

Speech articulation tests in our anechoic chamber, with and without the benefit of three reflective surfaces (to be described) for reinforcing the speech, reveal that the percentage of typical speech words heard correctly is 93.3% with the reflectors and only 69.7% without them. The directional characteristics of speech around the head of the speaker were consistent with those for tones of different frequencies; thus, the sound-pressure level of a 13 oct band of speech, centered at 4000 Hz, dropped 14 dB when the speaker turned 180° away from the recording microphone. The results obtained in the anechoic room, without the reflectors, are consistent with those we obtained 40 years ago at a noise-free location on a dry lake bed. The results are demonstrated with data and tape recordings, which reveal qualitatively and quantitatively the beneficial effects of suitable reflectors. The effects of early reflections on the quality of music are also demonstrated by tape recordings. These effects may be as important as is optimal reverberation time.