Binaural intelligibility prediction based on the speech transmission index

Although the speech transmission index (STI) is a well-accepted and standardized method for objective prediction of speech intelligibility in a wide range of environments and applications, it is essentially a monaural model. Advantages of binaural hearing in speech intelligibility are disregarded. In specific conditions, this leads to considerable mismatches between subjective intelligibility and the STI. A binaural version of the STI was developed based on interaural cross correlograms, which shows a considerably improved correspondence with subjective intelligibility in dichotic listening conditions. The new binaural STI is designed to be a relatively simple model, which adds only few parameters to the original standardized STI and changes none of the existing model parameters. For monaural conditions, the outcome is identical to the standardized STI. The new model was validated on a set of 39 dichotic listening conditions, featuring anechoic, classroom, listening room, and strongly echoic environments. For these 39 conditions, speech intelligibility [consonant-vowel-consonant (CVC) word score] and binaural STI were measured. On the basis of these conditions, the relation between binaural STI and CVC word scores closely matches the STI reference curve (standardized relation between STI and CVC word score) for monaural listening. A better-ear STI appears to perform quite well in relation to the binaural STI model; the monaural STI performs poorly in these cases.     

ODEON仿真与实际测量语言传输指数实验对比

近年来音质仿真技术的快速发展为语言传输指数STI的预测提供了一个潜在的解决方案。但这种方法的有效性如何,则是在使用该技术之前应该考虑的问题。本文对3个房间内音质仿真与实际测量STI进行实验对比,研究表明:在仿真模型与实际空间的声学等效较为准确的情况下,使用音质仿真软件ODEON计算得到的STI误差较小;混响时间的变化在背景噪声较高时可能会对仿真STI的准确性带来显著影响,随着混响时间的增加,仿真与实际测量STI的差值可能变大;信噪比的变化并不会给仿真STI的准确性带来显著影响;仿真脉冲响应与实际测量脉冲响应的频谱有一定差别,时域上的反射声序列也不相同,但这些差别对仿真STI的影响并不大;仿真过程中比较容易产生的信噪比误差对仿真STI产生了较大影响。由于影响音质仿真结果的因素较多,仿真模型与实际空间的声学等效也比较复杂,尤其是对于没有实际参照校准的房间来说,想要获得准确的STI预测结果是比较困难的。     

Blind estimation of reverberation time

The reverberation time (RT) is an important parameter for characterizing the quality of an auditory space. Sounds in reverberant environments are subject to coloration. This affects speech intelligibility and sound localization. Many state-of-the-art audio signal processing algorithms, for example in hearing-aids and telephony, are expected to have the ability to characterize the listening environment, and turn on an appropriate processing strategy accordingly. Thus, a method for characterization of room RT based on passively received microphone signals represents an important enabling technology. Current RT estimators, such as Schroeder's method, depend on a controlled sound source, and thus cannot produce an online, blind RT estimate. Here, a method for estimating RT without prior knowledge of sound sources or room geometry is presented. The diffusive tail of reverberation was modeled as an exponentially damped Gaussian white noise process. The time-constant of the decay, which provided a measure of the RT, was estimated using a maximum-likelihood procedure. The estimates were obtained continuously, and an order-statistics filter was used to extract the most likely RT from the accumulated estimates. The procedure was illustrated for connected speech. Results obtained for simulated and real room data are in good agreement with the real RT values.     

Verifying two commercial software implementations of impulse-response-based speech intelligibility measurements

A number of objective evaluation methods are currently used to quantify the speech intelligibility in a built environment, including the speech transmission index (STI), rapid speech transmission index (RASTI), articulation index (AI), and the percent articulation loss of consonants (%ALCons). Certain software programs can quickly evaluate STI, RASTI, and %ALCons from a measured room impulse response. In this project, two impulse-response-based software packages (WinMLS and SIA-Smaart Acoustic Tools) were evaluated for their ability to determine intelligibility accurately. In four different spaces with background noise levels less than NC 45, speech intelligibility was measured via three methods: (1) with WinMLS 2000; (2) with SIA-Smaart Acoustic Tools (v4.0.2); and (3) from listening tests with humans. The study found that WinMLS measurements of speech intelligibility based on STI, RASTI, and %ALCons corresponded well with performance on the listening tests. SIA-Smaart results were correlated to human responses, but tended to under-predict intelligibility based on STI and RASTI, and over-predict intelligibility based on %ALCons.     

Relationship between listening difficulty and acoustical objective measures in reverberant sound fields

The previous work [Morimoto et al., J. Acoust. Soc. Am. 116, 1607-1613] showed that listening difficulty ratings can be used to evaluate speech transmission performance more exactly and sensitively than intelligibility. Meanwhile, speech transmission performance is usually evaluated using acoustical objective measures, which are directly associated with physical parameters of room acoustic design. However, the relationship between listening difficulty ratings and acoustical objective measures was not minutely investigated. In the present study, a total of 96 impulse responses were used to investigate the relationship between listening difficulty ratings and several objective measures in unidirectional sound fields. The result of the listening test showed that (1) the correlation between listening difficulty ratings and speech transmission index (STI) is the strongest of all tested objective measures, and (2) A-weighted D(50), C(50), and center time, which are obtained from the impulse responses passed through an A-weighted filter, also strongly correlate with listening difficulty ratings, and their correlations with listening difficulty ratings are not statistically different from the correlation between listening difficulty ratings and STI.     

Effects of unattended speech on performance and subjective distraction: The role of acoustic design in open-plan offices

Unattended background speech is a known source of cognitive and subjective distraction in open-plan offices. This study investigated whether the deleterious effects of background speech can be affected by room acoustic design that decreases speech intelligibility, as measured by the Speech Transmission Index (STI). The experiment was conducted in an open-plan office laboratory (84 m²) in which four acoustic conditions were physically built. Three conditions contained background speech. A quiet condition was included for comparison. The speech conditions differed in terms of the degree of absorption, screen height, desk isolation, and the level of masking sound. The speech sounds simulated an environment where phone conversations are heard from different locations varying in distance. Ninety-eight volunteers were tested. The presence of background speech had detrimental effects on the subjective perceptions of noise effects and on cognitive performance in short-term memory and working memory tasks. These effects were not attenuated nor amplified within a three-hour working period. The reduction of the STI by room acoustic means decreased subjective disturbance, whereas the effects on cognitive performance were somewhat smaller than expected. The effects of room acoustic design on subjective distraction were stronger among noise-sensitive subjects, suggesting that they benefited more from acoustic improvements than non-sensitive subjects. The results imply that reducing the STI is beneficial for performance and acoustic satisfaction especially regarding speech coming from more distant desks. However, acoustic design does not sufficiently decrease the distraction caused by speech from adjacent desks.     

Subjective and objective speech intelligibility investigations in primary school classrooms

This work concerns speech intelligibility tests and measurements in three primary schools in Italy, one of which was conducted before and after an acoustical treatment. Speech intelligibility scores (IS) with different reverberation times (RT) and types of noise were obtained using diagnostic rhyme tests on 983 pupils from grades 2-5 (nominally 7-10 year olds), and these scores were then correlated with the Speech Transmission Index (STI). The grade 2 pupils understood fewer words in the lower STI range than the pupils in the higher grades, whereas an IS of ~97% was achieved by all the grades with a STI of 0.9. In the presence of traffic noise, which resulted the most interfering noise, a decrease in RT from 1.6 to 0.4 s determined an IS increase on equal A-weighted speech-to-noise level difference, S/N(A), which varied from 13% to 6%, over the S/N(A) range of -15 to +6 dB, respectively. In the case of babble noise, whose source was located in the middle of the classroom, the same decrease in reverberation time leads to a negligible variation in IS over a similar S/N(A) range.     

Fast estimation of speech transmission index using the reverberation time

The paper presents the function of STI in the domain of reverberation time. Through the application of the said function, we can quickly estimate the speech transmission index, knowing only the time of room reverberation. For that purpose we applied a known method which consists in physical estimation of speech intelligibility basing on the modulation transfer function (MTF) determined in a room. Then, the STI was described using a logarithmic function whereof argument was the room reverberation time. To verify the model, reverberation times of six rooms were measured. The selected rooms were very different deliberately. They had different cubature and shape. The selection included a small cuboid, lecture halls and a church. Then, the same rooms were modeled in the ODEON version 11.23 and their reverberation times were determined. Furthermore, the STI was determined in the ODEON and then compared with the reverberation time obtained in effect of fast estimation. The statistical verification with the use of correlation index and regression equation has demonstrated that the fast estimation yields results close to those obtained in the computer simulation in ODEON. We obtained the correlation index at the level close to 1. Furthermore, the test probability at the level lower than 0.05 bespeaks of a statistically significant linear relation for the confidence level of 95%.     

Experimental comparison between speech transmission index, rapid speech transmission index, and speech intelligibility index

During the acoustical design of, e.g., auditoria or open-plan offices, it is important to know how speech can be perceived in various parts of the room. Different objective methods have been developed to measure and predict speech intelligibility, and these have been extensively used in various spaces. In this study, two such methods were compared, the speech transmission index (STI) and the speech intelligibility index (SII). Also the simplification of the STI, the room acoustics speech transmission index (RASTI), was considered. These quantities are all based on determining an apparent speech-to-noise ratio on selected frequency bands and summing them using a specific weighting. For comparison, some data were needed on the possible differences of these methods resulting from the calculation scheme and also measuring equipment. Their prediction accuracy was also of interest. Measurements were made in a laboratory having adjustable noise level and absorption, and in a real auditorium. It was found that the measurement equipment, especially the selection of the loudspeaker, can greatly affect the accuracy of the results. The prediction accuracy of the RASTI was found acceptable, if the input values for the prediction are accurately known, even though the studied space was not ideally diffuse.     

Measurement of acoustical characteristics of mosques in Saudi Arabia

The study of mosque acoustics, with regard to acoustical characteristics, sound quality for speech intelligibility, and other applicable acoustic criteria, has been largely neglected. In this study a background as to why mosques are designed as they are and how mosque design is influenced by worship considerations is given. In the study the acoustical characteristics of typically constructed contemporary mosques in Saudi Arabia have been investigated, employing a well-known impulse response. Extensive field measurements were taken in 21 representative mosques of different sizes and architectural features in order to characterize their acoustical quality and to identify the impact of air conditioning, ceiling fans, and sound reinforcement systems on their acoustics. Objective room-acoustic indicators such as reverberation time (RT) and clarity (C50) were measured. Background noise (BN) was assessed with and without the operation of air conditioning and fans. The speech transmission index (STI) was also evaluated with and without the operation of existing sound reinforcement systems. The existence of acoustical deficiencies was confirmed and quantified. The study, in addition to describing mosque acoustics, compares design goals to results obtained in practice and suggests acoustical target values for mosque design. The results show that acoustical quality in the investigated mosques deviates from optimum conditions when unoccupied, but is much better in the occupied condition.     

Predicting speech metrics in a simulated classroom with varied sound absorption

By systematically varying the amount of sound absorption, and the location of the sound-absorbing material in a simulated classroom, it was possible to assess the accuracy of the prediction of speech metrics in quite simple acoustical environments. Predictions of speech level, early-to-late sound ratios (C50) and speech transmission index (STI) values were obtained analytically and with two hybrid ray-based computer programs, RAYNOISE 3.0 and ODEON 4.1. The RAYNOISE predictions were accomplished with a purely specular reflection model and also with a calibrated diffuse reflection model. ODEON uses a parameter called transition order, TO, to change the reflection procedure from purely specular to diffuse for reflections that have orders higher than TO. A parametric study was conducted to determine the best transition order for the ODEON prediction of speech metrics. It was found that the analytical predictions of speech level and C50 were on average accurate to about 1 just-noticeable difference (jnd), whereas the analytical predictions of STI were on average within 2 jnd's. ODEON predictions of speech level, C50 and STI were on average within 2 jnd's. RAYNOISE predictions of C50 and STI with the specular model were on average within 2 jnd's. However, the RAYNOISE predictions of speech level, with both types of reflection models, and the RAYNOISE predictions of C50 and STI with the diffuse model had average errors greater than 2 jnd's. The effects of the sound-absorption treatments on the measured speech metric values are also discussed.     

Prediction of optimal conditions for verbal-communication quality in eating establishments

This paper discusses the prediction of verbal-communication quality in eating establishments (EEs). EEs contain talkers and listeners who require high speech intelligibility at their tables, and high speech privacy between tables. Using catt-Acoustic, verbal-communication quality--quantified by speech transmission index (STI)--in models of three existing EEs was predicted. Talker voice-output levels were predicted using an existing empirical model accounting for the Lombard effect. With these, catt-Acoustic predicted impulse responses, speech levels and noise levels at primary and secondary listener positions, and the corresponding STIs. The untreated EEs were first modeled for various talker and listener positions, and occupancies. Then various treated configurations, involving reduced volume, increased absorption and barriers were studied to determine the effectiveness of the treatments. The results suggest that placing barriers around tables can be an effective way to achieve good verbal-communication quality. Increasing the absorption of the room surfaces or decreasing the ceiling height to control reverberation may not be effective. However, increasing the surface absorption and putting barriers around tables may achieve optimal speech conditions in EEs. Subdividing large EEs into smaller ones can also be effective.     

Using the Speech Transmission Index for predicting non-native speech intelligibility

While the Speech Transmission Index (STI) is widely applied for prediction of speech intelligibility in room acoustics and telecommunication engineering, it is unclear how to interpret STI values when non-native talkers or listeners are involved. Based on subjectively measured psychometric functions for sentence intelligibility in noise, for populations of native and non-native communicators, a correction function for the interpretation of the STI is derived. This function is applied to determine the appropriate STI ranges with qualification labels ("bad"-"excellent"), for specific populations of non-natives. The correction function is derived by relating the non-native psychometric function to the native psychometric function by a single parameter (nu). For listeners, the nu parameter is found to be highly correlated with linguistic entropy. It is shown that the proposed correction function is also valid for conditions featuring bandwidth limiting and reverberation.     

Accuracy of speech transmission index predictions based on the reverberation time and signal-to-noise ratio

This paper examines the accuracy of the speech transmission index (STI) calculated from the reverberation time (T) and signal-to-noise ratio (L_SN) of enclosed spaces. Differences between measured and predicted STIs have been analysed in two rooms (reverberant vs. absorbent), for a wide range of absorption conditions and signal-to-noise ratios (sixteen tests). The STI was measured using maximum length sequence analysis and predictions were calculated using either measured or predicted values of T and L_SN, the latter assuming diffuse sound field conditions. The results obtained for all the conditions tested showed that STI predictions based on T and L_SN tend to underestimate the STI, with differences between measured and predicted STIs always lower than 0.1 (on a 0.0–1.0 scale), and on average lower than 0.06. According to previous research, these differences are noticeable and therefore non-negligible, as 0.03 is the just noticeable difference in STI. The use of either measured or predicted values of T and L_SN provided similar STI predictions (i.e. non-noticeable changes), with differences between predictions that are on average lower than 0.03 for the absorbent room, and lower than 0.01 for the reverberant room.     

Relationship between listening difficulty rating and objective measures in reverberant and noisy sound fields for young adults and elderly persons

Listening difficulty ratings [Morimoto et al., J. Acoust. Soc. Am. 116, 1607-1613 (2004)] were obtained for 20 young adult listeners and 34 elderly listeners in reverberant and noisy sound fields simulated in an anechoic room. The listening difficulty ratings were compared with acoustical objective measures. The results and analyses showed the following: (i) The correlation between listening difficulty ratings and the revised speech transmission index (STI(r)), and that for the useful-detrimental ratio (U(50)) were high, regardless of the age of the listeners. (ii) STI(r) and U(50) need to be increased by 0.12 and 4.2 dB, respectively, to equalize the listening difficulty ratings for the elderly listeners with those for the young listeners. (iii) The estimation accuracies for STI(r) and U(50) can be improved by calculating them with the L(eq) of background noise linearly increased by 4 to 10 dB, which depends on the age of the listeners and the objective measures. However, the improvement was not statistically significant for the elderly listeners.     

Analysis of speech-based Speech Transmission Index methods with implications for nonlinear operations

The Speech Transmission Index (STI) is a physical metric that is well correlated with the intelligibility of speech degraded by additive noise and reverberation. The traditional STI uses modulated noise as a probe signal and is valid for assessing degradations that result from linear operations on the speech signal. Researchers have attempted to extend the STI to predict the intelligibility of nonlinearly processed speech by proposing variations that use speech as a probe signal. This work considers four previously proposed speech-based STI methods and four novel methods, studied under conditions of additive noise, reverberation, and two nonlinear operations (envelope thresholding and spectral subtraction). Analyzing intermediate metrics in the STI calculation reveals why some methods fail for nonlinear operations. Results indicate that none of the previously proposed methods is adequate for all of the conditions considered, while four proposed methods produce qualitatively reasonable results and warrant further study. The discussion considers the relevance of this work to predicting the intelligibility of cochlear-implant processed speech.     

Speech intelligibility of English,Polish, Arabic and Mandarin under different room acoustic conditions

This paper examines the impact of room acoustic conditions on the speech intelligibility of four languages (English, Polish, Arabic and Mandarin). Listening test scores (diagnostic rhyme tests, phonemically balanced word tests and phonemically balanced sentence tests) of the four languages were compared under four room acoustic conditions defined by their speech transmission index (STI = 0.2, 0.4, 0.6 and 0.8). The results obtained indicated that there was a statistically significant difference between the word intelligibility scores of languages under all room acoustic conditions, apart from the STI = 0.8 condition. English was the most intelligible language under all conditions, and differences with other languages were larger when conditions were poor (maximum difference of 29% at STI = 0.2, 33% at STI = 0.4 and 14% at STI = 0.6). Results also showed that Arabic and Polish were particularly sensitive to background noise, and that Mandarin was significantly more intelligible than those languages at STI = 0.4. Consonant-to-vowel ratios and languages’ distinctive features and acoustical properties explained some of the scores obtained. Sentence intelligibility scores confirmed variations between languages, but these variations were statistically significant only at the STI = 0.4 condition (sentence tests being less sensitive to very good and very poor room acoustic conditions). Overall, the results indicate that large variations between the speech intelligibility of different languages can occur, especially for spaces that are expected to be challenging in terms of room acoustic conditions. Recommendations solely based on room acoustic parameters (e.g. STI) might then prove to be insufficient for designing a multilingual environment.     

A method to determine the speech transmission index from speech waveforms

A method for computing the speech transmission index (STI) using real speech stimuli is presented and evaluated. The method reduces the effects of some of the artifacts that can be encountered when speech waveforms are used as probe stimuli. Speech-based STIs are computed for conversational and clearly articulated speech in several noisy, reverberant, and noisy-reverberant environments and compared with speech intelligibility scores. The results indicate that, for each speaking style, the speech-based STI values are monotonically related to intelligibility scores for the degraded speech conditions tested. Therefore, the STI can be computed using speech probe waveforms and the values of the resulting indices are as good predictors of intelligibility scores as those derived from MTFs by theoretical methods.     

Logatom corpus for the assessment of the intelligibility in Spanish speaking environments and its relation with STI measurements

A 1000 consonant–vowel–consonant structure logatoms corpus (CVC-structure), grouped in 10 phonetically equally balanced lists of 100 words each, was developed to satisfy the need of subjective assessment of speech intelligibility in American Spanish speaking environments. This corpus was tested and correlated with the Speech Transmission Index (STI) measurements to compare its articulation intelligibility score with other lists’ scores.Through the development of this work it was determined that in two different acoustically poor rooms that have the same STI (with STI < 0.50), the intelligibility score is lower when the articulation test is performed in a quiet room with high reverberation time than when it is performed in a very noisy room with low reverberation time. The final correlation curve of the American Spanish CVC-structure corpus was around 10% points higher than the CVC_EQB curve obtained by Steeneken and Houtgast in 2002.     

Prediction of verbal communication in noise—A development of generalized SIL curves and the quality of communication (Part 2)

The prediction of verbal communication is absolutely necessary for workplaces and living quarters disturbed by high noise levels. The methods to predict speech intelligibility (SIL, AI, STI, SNA) are assessed, taking into account both the speaker's sound level and the speech intelligibility at the hearer's position (Part 1). On the basis of this survey the development of generalized SIL curves is described and new SIL curves are proposed. Furthermore a method to determine the communication quality is described, which includes both the speaker's effort and the intelligibility reduction at the hearer's position (Part 2).