Sound Quality Evaluation Research Articles

The Impact of Vocal Boost Manipulations on Musical Sound Quality for Cochlear Implant Users.

To evaluate the impact of vocal boost manipulations on cochlear implant (CI) musical sound quality appraisals. An anonymous, online study was distributed to 33 CI users. Participants listened to auditory tokens and assessed the musical quality of acoustic stimuli with vocal boosting and attenuation using a validated sound quality rating scale. Four versions of real-world musical stimuli were created: a version with +9 dB vocal boost, a version with -9 dB vocal attenuation, a composite stimulus containing a 1,000 Hz low-pass filter and white noise ("anchor"), and an unaltered version ("hidden reference"). Subjects listened to all four versions and provided ratings based on a 100-point scale that reflected the perceived sound quality difference of the music clip relative to the reference excerpt. Vocal boost increased musical sound quality ratings relative to the reference clip (11.7; 95% CI, 1.62-21.8, p=0.016) and vocal attenuation decreased musical sound quality ratings relative to the reference clip (28.5; 95% CI, 18.64-38.44, p < 0.001). When comparing the non-musical training group and musical training group, there was a significant difference in musical sound quality rating scores for the vocal boost condition (21.2; 95% CI: 1.76-40.7, p=0.028). CI-mediated musical sound quality appraisals are impacted by vocal boost and attenuation. Musically trained CI users to report greater musical sound quality enhancement with a vocal boost with respect to CI users with no musical training background. Implementation of front-end vocal boost manipulations in music may improve sound quality and music appreciation among CI users. 2 (Individual cohort study) Laryngoscope, 133:938-947, 2023.

Cortical Entrainment to Speech Produced by Cochlear Implant Users and Normal-Hearing Talkers.

The perceived sound quality of speech produced by hard-of-hearing individuals greatly depends on the degree and configuration of their hearing loss. A cochlear implant (CI) may provide some compensation and auditory feedback to monitor/control speech production. However, to date, the speech produced by CI users is still different in quality from that produced by normal-hearing (NH) talkers. In this study, we attempted to address this difference by examining the cortical activity of NH listeners when listening to continuous speech produced by 8 CI talkers and 8 NH talkers. We utilized a discriminative model to decode and reconstruct the speech envelope from the single-trial electroencephalogram (EEG) recorded from scalp electrode in NH listeners when listening to continuous speech. The correlation coefficient between the reconstructed envelope and original speech envelope was computed as a metric to quantify the difference in response to the speech produced by CI and NH talkers. The same listeners were asked to rate the perceived sound quality of the speech as a behavioral sound quality assessment. Both behavioral perceived sound quality ratings and the cortical entrainment to speech envelope were higher for the speech set produced by NH talkers than for the speech set produced by CI talkers.

Speech recognition as a function of the number of channels for Mid-Scala electrode array recipients.

This study investigated the number of channels needed for maximum speech understanding and sound quality in 15 adult cochlear implant (CI) recipients with Advanced Bionics (AB) Mid-Scala electrode arrays completely within scala tympani. In experiment I, CI programs used a continuous interleaved sampling (CIS)-based strategy and 4-16 active electrodes. In experiment II, CI programs used an n-of-m strategy featuring 16 active electrodes with either 8- or 12-maxima. Speech understanding and sound quality measures were assessed. For CIS programs, participants demonstrated performance gains using up to 4-10 electrodes on speech measures and sound quality ratings. For n-of-m programs, there was no significant effect of maxima, suggesting 8-maxima is sufficient for this sample's maximum performance and sound quality. These results are largely consistent with previous studies using straight electrode arrays [e.g., Fishman, Shannon, and Slattery (1997). J. Speech Lang. Hear. Res. 40, 1201-1215; Friesen, Shannon, Baskent, and Wang (2001). J. Acoust. Soc. Am. 110, 1150-1163; Shannon, Cruz, and Galvin (2011). Audiol. Neurotol. 16, 113-123; Berg, Noble, Dawant, Dwyer, Labadie, and Gifford (2020). J. Acoust. Soc. Am. 147, 3646-3656] and in contrast with recent studies looking at cochlear precurved electrode arrays [e.g., Croghan, Duran, and Smith (2017). J. Acoust. Soc. Am. 142, EL537-EL543; Berg, Noble, Dawant, Dwuer, Labadie, and Gifford (2019b). J. Acoust. Soc. Am. 145, 1556-1564], which found continuous improvements up to 16 independent channels. These findings suggest that Mid-Scala electrode array recipients demonstrate similar channel independence to straight electrode arrays rather than other manufacturer's precurved electrode arrays.

Tire test stand measurements for blocked forces identification and tire noise auralization

Tire/Road noise is a dominant contribution to a vehicle interior noise. In this paper, we present the process we developed to support automotive OEMs with road noise engineering during vehicle design and development. The process uses blocked forces identified from measurements performed on a single tire installed on a test stand. The forces are then combined with vehicle level transfer functions to predict the sound at the driver ears. The results are imported in a driving simulator for NVH to allow for auralization and sound quality evaluations. The process can be applied from the very early stages of the vehicle development using CAE simulation results and allows to evaluate the vehicle road noise performance long before the first prototype is available.

Auralization of road noise CAE simulation results for interactive sound quality evaluations

Road noise is one of the dominant contributions to a vehicle interior noise when driving below highway speed. In this paper, we present the process we developed to support automotive OEMs with road noise engineering during vehicle design and development. The process allows to auralize, in a driving simulator for NVH, road noise CAE simulation results obtained from coupled multibody, structural finite-elements and acoustic models. With this process, engineers can now listen to a vehicle sound and perform interactive sound quality evaluation for road noise even before physical prototypes are available.

Nonlinear overall annoyance level modeling and interior sound quality prediction for pure electric vehicle with extreme gradient boosting algorithm

A single psychoacoustic metric cannot reflect pure electric vehicle (PEV) interior overall sound quality (SQ) accurately. There exists a complex nonlinear mapping relationship between psychoacoustic indexes and SQ prediction. In this work, a nonlinear overall annoyance level modeling method using the extreme gradient boosting (XGBoost) algorithm is proposed. Firstly, comprehensive noise samples inside the cabins of PEV and conventional fuel vehicle (FV) at high speed are collected. Subjective and objective SQ evaluations are conducted for FV and PEV interior noise, respectively. Their differences are analyzed by comparison. The correlation among various SQ indexes is studied, revealing a significant multicollinearity disruption. Then, the relative importance ranking is determined by the random forests algorithm, which provides a basis for feature index selection. Finally, the XGBoost-based nonlinear overall annoyance level model for SQ evaluation is established. Using the XGBoost-based model, traditional multiple linear regression (MLR) model, and linear least absolute shrinkage and selection operator (LASSO) model, PEV SQ is predicted. The results show that the XGBoost-based model exhibits high prediction accuracy, improving 75.6% relative to the MLR model and 74.9% relative to the LASSO model. Meanwhile, it has excellent stability and goodness-of-fitting. Thus, it provides an alternative method for PEV SQ evaluation and prediction.

Predictive evaluation of spectrogram-based vehicle sound quality via data augmentation and explainable artificial Intelligence: Image color adjustment with brightness and contrast

In this study, a novel method for selecting the optimal data augmentation method in combination with explainable artificial intelligence techniques is presented. Accordingly, a convolutional neural network-based model was designed to quantitatively evaluate the affective sound quality characteristics of vehicle driving sounds, which were classified as professional knowledge. Virtual learning data were created by adjusting the color of the image to avoid damaging the physical features of the spectrogram. By implementing the explainable artificial intelligence technique, the spectrogram features were extracted using domain knowledge. In particular, the engine noise of the vehicle, which plays a significant role in determining the characteristics of the running sound of the vehicle, was selected as a physical characteristic called the engine order line in the spectrogram. The explainable artificial intelligence technique was used to select the most influential feature among those extracted from the spectrogram. By observing the changes in the selected characteristics according to the data augmentation method, an optimal data augmentation method is proposed according to each characteristic. Furthermore, an average classification accuracy of 94.22% was obtained using the proposed data augmentation method, which is an improvement of 1.55–5.55% over the existing data augmentation methods. Moreover, according to the dataset, the standard deviation of the classification accuracy was 2.13%, which yielded an optimum result.

NAL-NL2 Prescriptive Targets for Bone Conduction Devices With an Adaptation to Device Constraints in the Low Frequencies.

Objectives:To study the effectivity of a transformed NAL non-linear version 2 (NAL-NL2) gain prescription for percutaneous bone conduction devices (BCDs) and to investigate how to take into account output constraints for the lower frequencies.Design:The NAL-NL2 prescription was converted to a bone conduction prescription rule. Adaptations were needed, as this converted rule prescribes more output at low frequencies than the device delivers. Three adaptations with different audibility and compression were compared. Setting 1 (S1, “optimal audibility”) had most audibility due to adapted frequency-dependent compression, setting 2 (S2, “moderate audibility”) had moderate output reduction below 1 kHz, and setting 3 (S3, “reduced audibility, least distortion”) had most output reduction. Eighteen experienced BCD users rated their relative sound quality in paired comparisons for different sounds (own voice, mixed voices, traffic noise, and music). In addition speech intelligibility in quiet and noise were assessed.Results:The relative sound quality rating for the adapted prescriptions varied between the stimuli: more low-frequency sound was preferred for music (S1 over S3), and less low-frequency sound was preferred for the own voice (S2 and S3 over S1). No differences in quality rating were found for mixed voices or traffic noise. Speech intelligibility in quiet scores at 45 dB SPL was significantly lower for S3 than for S1. Speech intelligibility in noise was significantly reduced in all settings and S3 yielded significantly better speech intelligibility in noise than S1.Conclusions:With a moderate gain reduction for low frequencies to comply with device constraints the transformed NAL-NL2 prescription was found suitable for fitting BCDs. Perceived sound quality depended on the gain settings, but also on the sound spectra and how the sound was appreciated. A moderate gain reduction below 1 kHz seems to be the optimal adaptation as it has a neutral or positive relative sound quality for all stimuli without negative effects on Speech intelligibility. The NAL-NL2-BC prescribed a sufficient amount of gain, as indicated by the speech tests.

Research on the Sound Quality Evaluation Method Based on Artificial Neural Network

For the improvement of the traditional evaluation effect of the automobile sound quality, an evaluation model of automobile sound quality is constructed based on BP neural network. The first is to introduce the basic principle of the BP neural network in detail. The second is to use the MGC parameters to establish the vehicle interior sound conversion model. The converted sound characteristic parameters are taken into the WORLD model to synthesize the new sound signals. Furthermore, the wavelet decomposition method is used to remove noise from the synthesized sound signals. Finally, a sound evaluation model based on BP neural network is established. The sound quality of automobiles can be better evaluated by carrying out the ABX test and MOS test in the field of sound conversion. For the newly synthesized sound signal and the target sound signal, it can be seen that the newly synthesized sound signal is more inclined to the target sound signal, and the sound quality is better. In addition, the sound quality is tested through loudness, roughness, sharpness, and level A in the field of sound quality evaluation. The final results show that the quality of newly synthesized sound is better, and the average errors of sound signals meet the sound standard. Therefore, the constructed sound conversion model and the sound evaluation model are feasible and effective.

Changes in Speech Recognition Scores and Sound Quality According to Gain Adjustment in High Frequency Bands

Purpose: In this study, changes before and after adjustment in high frequency bands were assessed for aided threshold, psychoacoustic evaluation, word recognition scores (WRSs) in noise and sound quality to aid in electroacoustic adjustment, and to improve the satisfaction of hearing aids.Methods: The subjects were 16 sensorineural hearing impaired (25 ears) with 63.4 years old of mean age, 60.8% WRS and 46.6 months of aided experience. For the experiment, hearing aid gain of high-frequency bands adjusted to preferred gain was increased to a higher level. Aided threshold, Korean version of international outcome inventory for hearing aids (K-IOI-HA), Korean version of profile of hearing aid benefit-quick version (K-PHAB-Q), WRSs in noise for 10 dB and 5 dB signal-to-noise ratio (SNR) and sound quality were subsequently evaluated after two weeks and the results were quantitatively compared before and after gain adjustment in high-frequency bands.Results: This study results in as follows; first, improvement in aided threshold for frequencies above 2.0 kHz was shown. Second, K-IOI-HA and K-PHAB-Q in easy communication and localization categories resulted in significant enhancements. Third, WRS in 5 dB and 10 dB SNR was improved. Fourth, improvements were shown for occlusion and clarity of the sentence in the evaluation of sound quality, but significant results were not shown for sharpness and loudness of noise conditions.Conclusion: With sensorineural hearing loss, a sufficient gain in high frequency bands provides improved satisfaction to hearing aids and enhanced word recognition ability in various listening environments.

Sound quality evaluation of pure electric vehicle with subjective and objective unified evaluation method

The focusing points of the sound quality of pure electric vehicle (PEV) usually depend on their operating condition. In this paper, a subjective and objective unified evaluation model for sound quality evaluation in PEV based on operating condition is proposed. Firstly, a hierarchical comparison method (HCM) is applied to subjectively evaluate the acoustic test results. Then, four psychoacoustic parameters and one evaluation index of pure tonal noise are used to objectively evaluate the noise sample. And the indicators used in different operating conditions are not the same. Finally, the sound quality evaluation models based on backpropagation neural network (BPNN) algorithm and the support vector machine (SVM) algorithm are established. The result shows that the larger the sample size, the more accurate are both models. However, when the sample size is insufficient, the accuracy of SVM model is better than the BPNN model.

Effect of Frequency Response Manipulations on Musical Sound Quality for Cochlear Implant Users.

Cochlear implant (CI) users commonly report degraded musical sound quality. To improve CI-mediated music perception and enjoyment, we must understand factors that affect sound quality. In the present study, we utilize frequency response manipulation (FRM), a process that adjusts the energies of frequency bands within an audio signal, to determine its impact on CI-user sound quality assessments of musical stimuli. Thirty-three adult CI users completed an online study and listened to FRM-altered clips derived from the top songs in Billboard magazine. Participants assessed sound quality using the MUltiple Stimulus with Hidden Reference and Anchor for CI users (CI-MUSHRA) rating scale. FRM affected sound quality ratings (SQR). Specifically, increasing the gain for low and mid-range frequencies led to higher quality ratings than reducing them. In contrast, manipulating the gain for high frequencies (those above 2 kHz) had no impact. Participants with musical training were more sensitive to FRM than non-musically trained participants and demonstrated preference for gain increases over reductions. These findings suggest that, even among CI users, past musical training provides listeners with subtleties in musical appraisal, even though their hearing is now mediated electrically and bears little resemblance to their musical experience prior to implantation. Increased gain below 2 kHz may lead to higher sound quality than for equivalent reductions, perhaps because it offers greater access to lyrics in songs or because it provides more salient beat sensations.

Sound quality evaluation of pure electric vehicle with subjective and objective unified evaluation method

Sound quality evaluation of pure electric vehicle with subjective and objective unified evaluation method

Sound quality perception of cochlear implant recipients: low-frequency information and foreign-language effect

Objective This study aimed to adapt a method used in sound quality measurements named CI-MUSHRA (the multiple stimuli with hidden reference and anchor for cochlear implant users) to the Turkish language. The effect of low-frequency information and non-native musical stimuli on sound quality perception was investigated. Design Subjects completed the Turkish version of the MUSHRA test, called TR-MUSHRA, and the original CI-MUSHRA test. Participants also completed the Turkish monosyllabic word recognition test and the spectral temporal modulated ripple test (SMRT). Study sample 19 cochlear implant (CI) users and 16 normal-hearing (NH) adults were included. Results CI users demonstrated a lack of ability to detect the sound quality differences between original stimuli and stimuli with omitted low-frequency information up to 600 Hz in both tests. There was no significant main effect of the test version on sound quality ratings for the two groups. No significant correlation was found between mean sound quality scores, SMRT, and speech recognition in quiet and noise conditions. Conclusions Our study suggests that CI users perform poorly in discriminating high-pass filtered musical sounds regardless of the language of the musical stimuli. The TR-MUSHRA can be used as a reliable research tool to evaluate the perceived sound quality.

Cochlear Implant Compression Optimization for Musical Sound Quality in MED-EL Users.

Variations in loudness are a fundamental component of the music listening experience. Cochlear implant (CI) processing, including amplitude compression, and a degraded auditory system may further degrade these loudness cues and decrease the enjoyment of music listening. This study aimed to identify optimal CI sound processor compression settings to improve music sound quality for CI users. Fourteen adult MED-EL CI recipients participated (Experiment No. 1: n = 17 ears; Experiment No. 2: n = 11 ears) in the study. A software application using a modified comparison category rating (CCR) test method allowed participants to compare and rate the sound quality of various CI compression settings while listening to 25 real-world music clips. The two compression settings studied were (1) Maplaw, which informs audibility and compression of soft level sounds, and (2) automatic gain control (AGC), which applies compression to loud sounds. For each experiment, one compression setting (Maplaw or AGC) was held at the default, while the other was varied according to the values available in the clinical CI programming software. Experiment No. 1 compared Maplaw settings of 500, 1000 (default), and 2000. Experiment No. 2 compared AGC settings of 2.5:1, 3:1 (default), and 3.5:1. In Experiment No. 1, the group preferred a higher Maplaw setting of 2000 over the default Maplaw setting of 1000 (p = 0.003) for music listening. There was no significant difference in music sound quality between the Maplaw setting of 500 and the default setting (p = 0.278). In Experiment No. 2, a main effect of AGC setting was found; however, no significant difference in sound quality ratings for pairwise comparisons were found between the experimental settings and the default setting (2.5:1 versus 3:1 at p = 0.546; 3.5:1 versus 3:1 at p = 0.059). CI users reported improvements in music sound quality with higher than default Maplaw or AGC settings. Thus, participants preferred slightly higher compression for music listening, with results having clinical implications for improving music perception in CI users.

Studying critical frequency bands and channels for EEG-based automobile sound recognition with machine learning

As for subjective methods (such as scale scores) for evaluation of automobile sound quality, electroencephalogram (EEG) can objectively and real-time reflect human mental states. Thus, EEG was introduced here to study the method of automobile sound recognition in view of frequency bands and channels. An EEG acquisition instrument was utilized to simultaneously acquire the EEGs of subjects stimulated by engine, exhaust and in-car sounds, and the power spectra of EEG were extracted by means of Welch Algorithm to construct a feature matrix; Then, the accuracies of automobile sound recognition based on five machine learning models were compared to determine the optimal model and the best frequency band range; What’ more, considering the differences between subjects, the m-ReliefF algorithm was applied here to investigate the universality channel with the optimal weight. Our results indicated that the differences between the three kinds of automobile sounds can be effectively identified by means of the KNN model, whose mean of accuracy peaks up to 87.93%; In addition, the recognition accuracies of beta and gamma bands were far higher than that of alpha and theta bands. Also, the accuracies of top universality 9 and 12 channels were up to 89.9% and 92.8% respectively, which were even better than that of the original 63 channels. This study revealed that recognition accuracy and computational efficiency can be significantly improved through the precise localization of the frequency bands and channels related with the automobile sounds, and demonstrated that EEG-based evaluation of the automobile sound quality shall be feasible.

Beneficial loudspeaker/room/listener interactions

Sound reproduction of the highest quality requires an optimized combination of loudspeakers, rooms and listeners—electroacoustics, acoustics and psychoacoustics. Domestic and monitor loudspeakers have improved because of research investigating the physical sounds arriving at listeners in small rooms and correlating double-blind subjective ratings of sound quality with relevant anechoic measurements on loudspeakers. Loudspeakers designed using specific anechoic measurements and guidance (e.g., ANSI/CTA-2034-B) reliably yield high sound quality ratings in double-blind listening tests in normally reflective domestic rooms. The relative subjective ratings of loudspeakers remain substantially unchanged when comparison listening tests are conducted in rooms with very different acoustics, although the absolute sound quality has changed. Listeners clearly adapt to some of the essential perceptual characteristics of rooms. But there are limits. About 30% of subjective ratings of sound quality is attributable to bass quality. This is different for every arrangement of loudspeakers and listeners in every different room. Seat-to-seat variations at low frequencies are large. Absorption is traditionally used to damp bass resonances. However, for reproduced sound, multiple subwoofers, with or without signal processing, can manipulate room resonances. These schemes, with equalization, can substantially attenuate audible resonances while yielding much reduced seat-to-seat variation, using no passive acoustical devices.

Research on prediction model of tractor sound quality based on genetic algorithm

The noise comfort of a tractor is a key factor in the evaluation of tractor comfort, and it has a significant impact on the evaluation of the overall performance evaluation of a tractor. This paper used the rating scale method to evaluate the sound quality of each operation state and position of the tractor, and obtain the evaluation value of the tractor noise. In addition, objective parameters of the sound (including the sound pressure level, A-weighted sound pressure level, loudness, sharpness, roughness, and fluctuation strength) were calculated. By considering objective parameters of sound quality as input and sound quality evaluation values as output, we established back propagation neural network (BPNN) model and support vector regression (SVR) model. Furthermore, the initial weights and thresholds of the BPNN model, penalty parameters, insensitive loss parameters, and kernel function parameters of the SVR model were optimized by genetic algorithm (GA). After verification of the experimental results, it was observed that the GA could improve the prediction accuracy of the model and significantly reduce the extreme errors. Compared with the prediction results of other prediction models, the GA–SVR model predicted the sound quality value of tractor noise with higher accuracy and stability.

Effects of Adaptive Non-linear Frequency Compression in Hearing Aids on Mandarin Speech and Sound-Quality Perception

ObjectiveThis study was aimed at examining the effects of an adaptive non-linear frequency compression algorithm implemented in hearing aids (i.e., SoundRecover2, or SR2) at different parameter settings and auditory acclimatization on speech and sound-quality perception in native Mandarin-speaking adult listeners with sensorineural hearing loss.DesignData consisted of participants’ unaided and aided hearing thresholds, Mandarin consonant and vowel recognition in quiet, and sentence recognition in noise, as well as sound-quality ratings through five sessions in a 12-week period with three SR2 settings (i.e., SR2 off, SR2 default, and SR2 strong).Study SampleTwenty-nine native Mandarin-speaking adults aged 37–76 years old with symmetric sloping moderate-to-profound sensorineural hearing loss were recruited. They were all fitted bilaterally with Phonak Naida V90-SP BTE hearing aids with hard ear-molds.ResultsThe participants demonstrated a significant improvement of aided hearing in detecting high frequency sounds at 8 kHz. For consonant recognition and overall sound-quality rating, the participants performed significantly better with the SR2 default setting than the other two settings. No significant differences were found in vowel and sentence recognition among the three SR2 settings. Test session was a significant factor that contributed to the participants’ performance in all speech and sound-quality perception tests. Specifically, the participants benefited from a longer duration of hearing aid use.ConclusionFindings from this study suggested possible perceptual benefit from the adaptive non-linear frequency compression algorithm for native Mandarin-speaking adults with moderate-to-profound hearing loss. Periods of acclimatization should be taken for better performance in novel technologies in hearing aids.

Sound quality evaluation for luxury refrigerator door closing sound

In this study, we carried out subjective evaluation tests employing 19 refrigerator door closing sounds to quantify the luxury feeling. By applying factor analysis to the subjective evaluation results, the sound quality of the refrigerator door closing sound was found to be expressed by the following two factors: overall loudness and the pitch of the sound. Subsequently, luxury feeling evaluation model was obtained through multiple regression analysis. As the result, the luxury feeling of the door closing sound was evaluated to be high when the sound was softer and had lower pitch. Then, we prepared several luxury door closing sounds according to the obtained evaluation model through a filter processing and conducted subjective evaluation tests again to verify the evaluation model. The result shows that the amplitude increased sound at low frequency band under 100 Hz, which was calculated to be high luxury by the evaluation model, was actually evaluated as the best among the presented sounds through the subjective test. And the luxury sound quality evaluation method was confirmed to be useful to quantify and estimate the sound quality of the refrigerator door closing sound.