Publications | Dr. Randy F Fela

2026

ICASSP

Conformal Uncertainty for No-Reference MOS Prediction under Distribution Shift: A Leakage-Controlled Study

R. F. Fela

In ICASSP 2027 (submitted, under review), 2026

Abs HTML Code

No-reference MOS predictors (DNSMOS P.835, NISQA, UTMOS) return a point estimate with no signal of when to trust it. This study tests whether feature-space distance from a reference distribution can flag unreliable predictions and set per-utterance conformal interval widths, using a leakage-controlled design that separates feature-reference fitting, content-disjoint calibration, and evaluation. Feature-space distance detects some corpus shifts but does not reliably flag which individual predictions are wrong. Mean domain-detection AUROC is 0.615 to 0.690 across three distance scores, while mean AUROC for detecting the highest-error quartile within a domain is only 0.433 to 0.499, near chance.
EMNLP

Perceptually Better, Semantically Worse: Measuring Speech Enhancement Impact on LLM-Based Voice Systems

R. F. Fela and P. Mowlaee

In Proceedings of the 2026 Conference on Empirical Methods in Natural Language Processing: Industry Track, 2026

Abs HTML Code Poster

Speech enhancement (SE) front-ends improve perceptual quality metrics such as PESQ and STOI, but can silently damage the semantic content that downstream LLMs rely on. We introduce Output Divergence Rate (ODR), the fraction of clips whose LLM-predicted intent changes after enhancement relative to the clean baseline, and benchmark five conditions on 2,974 SLURP test clips with 77 intents. MetricGAN+ more than doubles ODR (0.318 vs. 0.135 for noisy speech) despite improving PESQ, and unmitigated echo reaches an ODR of 0.836 through wrong-speaker transcription. PESQ shows the strongest pooled correlation with ODR (Spearman rho = -0.467), yet within any single condition no metric exceeds |rho| = 0.34. Findings replicate across ASR architectures (Whisper large-v3 and wav2vec2-large).

2024

AES

Evaluation of Own Voice Perception while Using In-ear Headsets

M. Takeuchi, J. Marozeau, R. F. Fela, and 2 more authors

The AES 4th International Conference on Audio and Music Induced Hearing Disorders, Ålborg, Denmark. More Information can be found here , May 2024

Abs HTML PDF

Singing or speaking while using a headset or hearing aid can present significant challenges due to the occlusion effect, alongside variances in processing latency and auditory distortions inherent to these devices. Such challenges are particularly pronounced in activities like singing, where precise control over one’s vocal quality is paramount for performance. This study investigates the factors detrimental to the perceptual quality of one’s own voice when using in-ear headsets and how each factor influences one’s perception of self-voice. Prior literature reviews lead to occlusion, sidetone filters, latency, and distortion as the main factors. Twenty-three assessors participated in a subjective evaluation, which was followed by statistical data analysis. Further investigation was carried out through an assessor screening procedure and text analysis was performed on responses to collected open-ended questions. The study found that occlusion and distortion are the influential factors for the changes in self-voice perception, whereas latency is a determinant. Sidetone filtering becomes a significant factor when occlusion effects are minimized.

2023

JAES

Comparison of Full Factorial and Optimal Experimental Design for Perceptual Evaluation of Audiovisual Quality

R. F. Fela, N. Zacharov, and S. Forchhammer

Journal of the Audio Engineering Society. More Information can be found here , Jan 2023

Abs HTML PDF

Perceptual evaluation of immersive audiovisual quality is often very labor-intensive and costly because numerous factors and factor levels are included in the experimental design. Therefore, the present study aims to reduce the required experimental effort by investigating the effectiveness of optimal experimental design (OED) compared to classical full factorial design (FFD) in the study using compressed omnidirectional video and ambisonic audio as examples. An FFD experiment was conducted and the results were used to simulate 12 OEDs consisting of D-optimal and I-optimal designs varying with replication and additional data points. The fraction of design space plot and the effect test based on the ordinary least-squares model were evaluated, and four OEDs were selected for a series of laboratory experiments. After demonstrating an insignificant difference between the simulation and experimental data, this study also showed that the differences in model performance between the experimental OEDs and FFD were insignificant, except for some interacting factors in the effect test. Finally, the performance of the I-optimal design with replicated points was shown to outperform that of the other designs. The results presented in this study open new possibilities for assessing perceptual quality in a much more efficient way.

2022

ICMEW

Perceptual Evaluation on Audio-visual Dataset of 360 Content

R. F. Fela, A. Pastor, P. Le Callet, and 3 more authors

In 2022 IEEE International Conference on Multimedia and Expo Workshops (ICMEW), Taipei, Taiwan. More Information can be found here , Jul 2022

Abs arXiv HTML PDF

To open up new possibilities to assess the multimodal perceptual quality of omnidirectional media formats, we proposed a novel open source 360 audiovisual (AV) quality dataset. The dataset consists of high-quality 360 video clips in equirectangular (ERP) format and higher-order ambisonic (4th order) along with the subjective scores. Three subjective quality experiments were conducted for audio, video, and AV with the procedures detailed in this paper. Using the data from subjective tests, we demonstrated that this dataset can be used to quantify perceived audio, video, and audiovisual quality. The diversity and discriminability of subjective scores were also analyzed. Finally, we investigated how our dataset correlates with various objective quality metrics of audio and video. Evidence from the results of this study implies that the proposed dataset can benefit future studies on multimodal quality evaluation of 360 content.
JAES

Assessor Selection Process for Perceptual Quality Evaluation of 360 Audiovisual Content

R. F. Fela, N. Zacharov, and S. Forchhammer

Journal of Audio Engineering Society. More Information can be found here , Oct 2022

Abs HTML PDF

For accurate and detailed perceptual evaluation of compressed omnidirectional multimedia content, it is imperative for assessor panels to be qualified to obtain consistent and high-quality data. This work extends existing procedures for assessor selection in terms of scope (360? videos with high-order ambisonic), time efficiency, and analytical approach, as described in detail. The main selection procedures consisted of a basic audiovisual screening and three successive discrimination experiments for audio (listening), video (viewing), and audiovisual using a triangle test. Additionally, four factors influencing quality of experience, including the simulator sickness questionnaire, were evaluated and are discussed. After the selection process, a confirmatory study was conducted using three experiments (audio, video, and audiovisual) and based on a rating scale methodology to compare performance between rejected and selected assessors. The studies showed that (i) perceptual discriminations are influenced by the samples, the encoding parameters, and some quality of experience factors; (ii) the probability of symptom occurrence is considerably low, indicating that the proposed procedure is feasible; and (iii) the selected assessors performed better in discrimination than the rejected assessors, indicating the effectiveness of the proposed procedure.

2021

MMSP

Perceptual Evaluation of 360 Audiovisual Quality and Machine Learning Predictions

R. F. Fela, N. Zacharov, and S. Forchhammer

In 2021 IEEE 23rd International Workshop on Multimedia Signal Processing (MMSP), Tampere, Finland. More Information can be found here , Oct 2021

Abs arXiv HTML PDF

In an earlier study, we gathered perceptual evaluations of the audio, video, and audiovisual quality for 360 audiovisual content. This paper investigates perceived audiovisual quality prediction based on objective quality metrics and subjective scores of 360 video and spatial audio content. Thirteen objective video quality metrics and three objective audio quality metrics were evaluated for five stimuli for each coding parameter. Four regression-based machine learning models were trained and tested here, i.e., multiple linear regression, decision tree, random forest, and support vector machine. Each model was constructed using a combination of audio and video quality metrics and two cross-validation methods (k-Fold and Leave-One-Out) were investigated and produced 312 predictive models. The results indicate that the model based on the evaluation of VMAF and AMBIQUAL is better than other combinations of audio-video quality metric. In this study, support vector machine provides higher performance using k-Fold (PCC = 0.909, SROCC = 0.914, and RMSE = 0.416). These results can provide insights for the design of multimedia quality metrics and the development of predictive models for audiovisual omnidirectional media.

2020

QoMEX

Towards a Perceived Audiovisual Quality Model for Immersive Content

R. F. Fela, N. Zacharov, and S. Forchhammer

In 2020 Twelfth International Conference on Quality of Multimedia Experience (QoMEX), Dublin, Ireland. More Information can be found here , May 2020

Abs arXiv HTML PDF

This paper studies the quality of multimedia content focusing on 360 video and ambisonic spatial audio reproduced using a head-mounted display and a multichannel loudspeaker setup. Encoding parameters following basic video quality test conditions for 360 videos were selected and a low-bitrate codec was used for the audio encoder. Three subjective experiments were performed for the audio, video, and audiovisual respectively. Peak signal-to-noise ratio and its variants for 360 videos were computed to obtain objective quality metrics and subsequently correlated with the subjective video scores. This study shows that a Cross-Format SPSNR-NN has a slightly higher linear and monotonic correlation over all video sequences. Based on the audiovisual model, a power model shows a highest correlation between test data and predicted scores. We concluded that to enable the development of superior predictive model, a high quality, critical, synchronized audiovisual database is required. Furthermore, comprehensive assessor training may be beneficial prior to the testing to improve the assessors’ discrimination ability particularly with respect to multichannel audio reproduction.

2019

BuildSim

The Effects of Orientation, Window Size, and Lighting Control to Climate-Based Daylight Performance and Lighting Energy Demand on Buildings in Tropical Area

R. F. Fela, S. S. Utami, R. A. Mangkuto, and 1 more author

In Proceedings of Building Simulation 2019: 16th Conference of IBPSA, Rome, Italy. More Information can be found here , Sep 2019

Abs DOI HTML PDF

The use of daylight offers a positive contribution to building energy savings yet also provoke a shortfall anticipation such as the large amount of light and visual discomfort. Moreover, building energy demand will also depend on the lighting control system combined with occupants’ behaviour of which is still rarely studied. This paper addressed the evaluation of climate-based daylight performance in a simulated reference office with a single unglazed opening measuring various window-to-wall ratio on one of the short façades facing different orientations. Modelling and simulation were performed in Radiance and Daysim using weather data of Yogyakarta and Bandung, Indonesia. Six different lighting control conditions were included. The following metrics observed are: average daylight factor, spatial daylight autonomy, spatial useful daylight illuminance, simplified daylight glare probability and lighting energy demand due to the lighting control system. Sensitivity analysis using multiple linear regression was also included. Excluding the average daylight factor, the results yielded similar trends for each metrics in two cities. However, the trend of sDA and sUDI-e in Yogyakarta are getting higher than Bandung in contrast with lighting energy demand. Moreover, combination switch-off occupancy & dimming system shows the most efficient lighting control system in all conditions.
BuildSim

Effect of Facade Thickness on Daylight Performance in a Reference Office Building

R. A. Mangkuto, R. F. Fela, and S. S. Utami

In Proceedings of Building Simulation 2019: 16th Conference of IBPSA, Rome, Italy. More Information can be found here , Sep 2019

Abs DOI HTML PDF

Current daylighting simulation practice often ignores the importance of façade thickness, despite its real influence on the total building performance. This research therefore aims to determine the effect of various facade thickness on daylight performance in a simulated reference office. An isolated reference office with an unglazed opening on one of the short façade was taken as case study, modelled and simulated in Radiance and Daysim, using weather data of Bandung, Indonesia. The thickness of the window façade was varied within 0.05 0.50 m, and the following daylight metrics were determined: average daylight factor, spatial daylight autonomy, spatial useful daylight illuminance, simplified daylight glare probability on selected viewpoints, and annual lighting energy demand. Results suggest that variation of façade thickness yields different impacts, which is largely significant on the average daylight factor, but is subtler on the climate-based daylight metrics and lighting energy demand.
BuildSim

CFD Simulation in Naturally Ventilated Multi-Purpose Sports Hall: Effects to the Basketball and Badminton Matches

M. Z. Afif, R. F. Fela, S. S. Utami, and 2 more authors

In Proceedings of Building Simulation 2019: 16th Conference of IBPSA, Rome, Italy. More Information can be found here , Sep 2019

Abs DOI HTML PDF

Minimizing the operational cost of a building while maintaining the thermal comfort is one of the criteria required in a sport facility building. However, the standard to obtain this condition in Indonesia is not yet available. The thermal comfort condition of a multipurpose sport facility in a public university in Yogyakarta, Indonesia was investigated by observing the air movement through natural ventilation openings. Performance metrics for the thermal comfort are the temperature distribution, relative humidity, air flow rate, and outdoor climate. Simulation was done using SolidWorks 2013 and the thermal condition was known. The results were compared to the Indonesia National Standard (SNI) and Design Guidance Note from Sport England. The average temperature near the arena is 30.09oC with relative humidity of 67%, which meets the category of a comfort warm condition. The spectator seats at the tribune are considered comfortably warm. The sport hall being observed has an air change rate (ACH) of 6.14 which exceeds the standard minimum of 1.5 and therefore, it is in no need of a mechanical ventilation system to support the air circulation.
AIP

Study of Partition Walls to Minimize Leaking Sound in Counselling Rooms

S. S. Utami, R. F. Fela, J. Sarwono, and 1 more author

In AIP Conference Proceedings. More Information can be found here , 2019

Abs DOI HTML PDF

Counselling rooms are needed by psychologists to perform psychotherapy, an activity that requires private conversation between the psychologist and the client. Complains of sound leakage through the counselling rooms of Psychology Counselling Unit in UGM Indonesia were reported. This study provides recommendation to optimize the sound insulation in the partition walls. The method relies on improvement of the STC using room impulse response (RIR) and modeling of the composite partitions with INSUL 8.0 software. The RIR provides reverberation time using omnidirectional speakers and 2 omnidirectional microphones positioned 1.2 meters above the floor. The RT estimates the total absorption of the space. The noise reduction is the difference between the SPL measured with a microphone in the source’s room and the SPL measured with another microphone in the receiver’s room. The source was a pink noise signal generated through a direct speaker. The simulation result of the existing partition materials, consists of gypsum board ½ inch double panel without absorbent material, has an STC value of 36. Meanwhile, from the on-site measurement, for room D.204 and room D.205 the STC is 31, for room D.206 the STC is 28, and STC of 27 for room D.205 with the source in room D.206. This result does not meet the standard STC for partition walls in counselling rooms which is 45dB. The optimization of sound insulation performances was done through simulations in INSULS 8.0 with 2 scenarios that are most reasonable to occur, given the existing room conditions. The best partition wall design attained an STC of up to 45 using 16 mm double gypsum board panel with 68 mm cavity containing 60 kg/m3 rock wool. Furthermore, the result also indicated that the noise leaking was caused by the lack of sound insulation capability of the partition walls and occurrences of flanking noise.
AIP

Prediction of Sound Insulation for Middle Class Apartment Based on Noise Site Analysis

H. A. G. Barabah, S. S. Utami, H. R. Hapsari, and 3 more authors

In AIP Conference Proceedings. More Information can be found here , 2019

Abs DOI HTML PDF

As a middle-class apartment in Indonesia, Rusunami are commonly found in urban area that are exposed by noise pollution from traffic road and railways. Sound insulation for building envelope is therefore needed to avoid the building occupants from acoustical comfort problems (i.e. health and psychological effects) due to the exposure of high sound intensity. By using result of the noise-site analysis, this study aims to propose the Sound Transmission Class (STC) for building envelope. Materials and its layers were modeled using Insul to obtain Transmission Loss (TL) value of which further calculation to result the STC value. Investigation of sound propagation were modeled and simulated using acoustics pressure module in finite element based-software. The noise reduction inside Rusunami was calculated based on EN 12354-3: 2000 using the outdoor to indoor calculation. The indoor SPL was further compared to the noise standard on the indoor residential area of 45 dBA. The result revealed that the wall with 110 mm thick concrete material with a layers of 15 mm cement sand and 5 mm plaster could provide STC 53 and indoor SPL 37.72 dBA while lightweight concrete material with same thickness and layers could produce STC 49 and indoor SPL 41.39 dBA. The use of both materials as building envelope for Rusunami have been in accordance with the standard of noise level. They have been identified as requisite to protect public health and welfare with an adequate margin of safety for indoor residential area.
AIP

Soundscape Study of an Urban Campus Park

D. A. Puspagarini, S. S. Utami, A. S. Sudarsono, and 1 more author

In AIP Conference Proceedings. More Information can be found here , 2019

Abs DOI HTML PDF

Aural experience of a space, or soundscape, is one of the factors that influence visitor’s comfort in a park. An ideal park should able to provide a tranquil environment and noiseless condition, dominated by the presence of natural sounds. The acoustical condition of the park’s environment is therefore an indicator of a good park. This study aims to evaluate the acoustical condition of an urban campus park in Yogyakarta using soundscape method. This park located aside to the campus main road and residential areas, creating a lot of noise exposure to the park that reduces the expected soundscape quality. Soundscape analysis is based on the environmental acoustic parameters of LAeq, L10, L50, L90, and soundscape ecology indices of NDSI, AEI, and H. These parameters were processed to characterize the sound objects within the park. An on-site subjective measurement was conducted to evaluate visitor preferences of the acoustic environment using questionnaires. The results indicate that the value of certain parameters in the park varies depend on the zone and time measurement. The environment sound level is quite high, which is above the permitted SPL of 50 dBA. In addition to that, the sound compositions inside the park consist of natural and artificial sound sources, which are common entity found in urban parks. However, the designed zoning of the landscape in this area did not create any unique soundmark for each zone. Meanwhile, utilizing the principal component analysis (PCA) of the subjective measurement results, shows that major dimension with the most significant effect on respondents’ perceptions is calmness. In overall, even with the high sound level inside the park, the acoustical condition is still perceived positively by the visitors.
AIP

Evaluation of Sonic Environment in Prospective Middle-class Apartment Building in Demak Regency

H. R. Hapsari, S. S. Utami, H. A. G. Barabah, and 2 more authors

In AIP Conference Proceedings. More Information can be found here , 2019

Abs DOI HTML PDF

The location of prospective middle-class apartment building in Demak regency is adjacent with roadway and double track railway. The high mobility of the various modes of transportation in these locations could potentially for excessive noise exposure above the noise level standard. By conducting noise measurement and subjective survey, this study aims to investigate the sonic environment of the site and its effect on surrounding communities. Measurement of noise was carried out in thirty-six positions by using calibrated H6 Zoom and Sound Level Meter. To determine the distribution of LAeq, the measurement results are visualized using QGIS, Geographic Information System based software. In addition, subjective survey was conducted to investigate the perception of annoyance from people living nearby the location. The result is to provide a suggestion toward the design guideline to create a good acoustics quality that meets the standard for building.

2018

Book

Menelusur Jejak Implementasi Konsep Bangunan Hijau dan Pintar di Kampus Biru

S. S. Utami, R. F. Fela, R. J. Yanti, and 1 more author

In Menelusur Jejak Implementasi Konsep Bangunan Hijau dan Pintar di Kampus Biru, 2018

Abs PDF

This book chapter traces the implementation of green and smart building concepts at the "Blue Campus" (Kampus Biru) of Universitas Gadjah Mada. It documents sustainable design strategies, energy efficiency measures, and intelligent building systems deployed across the campus.
ICSV

Hospital Soundscapes: Soundscapes Interventions in Intensive Care Unit

S. S. Azzahra, I. R. N. Prasetiyo, J. Prasetiyo, and 2 more authors

In 25th International Congress on Sound and Vibration 2018, Hiroshima, Japan, Jul 2018

Abs PDF

Acoustic environment of intensive care unit (ICU) is known to be noisy and stressful. The sounds from medical equipment and staffs’ activity is found out to give the most contribution in the overall sound level inside the ICU. Many study has shown its negative effect for patients and staffs, both psychologically and physiologically. However, several studies also shown that the information content in ICU soundscape is consedired as important, especially for the nurses. Thus, the following question needs to be answered:“How to improve occupant’s perception without eliminating the information content inside an ICU?” Several studies have shown the positive impact of soundscape intervention to the perception of occupants in healthcare facilities. However, previous studies were mostly done from the patient’s perspective. In this study, the effect of sound intervention to ICU soundscape will be studied from the nurse’s perspective, specifically to find out several sounds that mostly preferred by the nurses. The sound was added to ICU soundscape through soundscape simulator, and the nurses were asked to rate the soundscape. It has been found that certain sound with certain level can improve the nurse’s perception. This result can be used by practitioners to improve the soundscape while still maintaining the information content in an ICU.
ICSV

A Pilot Study: How to Develop Auditory Virtual Reality for Pre-registered Nurses Education?

Y. W. Fela, R. F. Fela, J. Sarwono, and 2 more authors

In 25th International Congress on Sound and Vibration 2018, Hiroshima, Japan, Jul 2018

Abs

This pilot study explores the development of auditory virtual reality systems for nursing education. The research investigates how spatial audio can enhance immersive training scenarios for pre-registered nurses, particularly for practicing clinical communication and emergency response in realistic soundscapes.
ICSV

Acoustics Survey to Predict Occupants Preferences of A Library

S. S. Utami, R. F. Fela, J. Sarwono, and 1 more author

In 25th International Congress on Sound and Vibration 2018, ICSV 2018, Hiroshima, Japan, Jul 2018

Abs

Objective measurement only may not be able to characterize the acoustic condition of a library and therefore, occupants’ expectation is necessarily needed. This study aims to survey the occupants’ preference of the acoustic condition in a university library through objective and subjective measurement. Room acoustics simulation using 2 sources and 4 receivers with on-site background noise data was carried out to measure the T30, RASTI and C50. Meanwhile, for the subjective survey, sixty visitors (36 males, 24 females) with an age range between 17–28 yo (Mean= 20.43, SD= 2.18) participated in answering the questionnaire about the soundscape and occupants’ preferences of the sonic environment. A semantic differential model using 6 rating scales was used in the questionnaire and Principal Component Analysis (PCA) was used to describe the data. The analysis reveals 4 significant labelled dimensions from PCA (KMO Test= 0.694), which are Pleasantness (32.03%), Room Effects (25.6%) and Privacy (10.85%). This result proved the consistency of the dimension from the previous study and verified that the soundscape dimension is possibly correlated with two main aspects involving room function and room effects. The perception of the recognized sound, pleasant–unpleasant sound, and expected condition is also provided.

2017

INTER-NOISE

Hospital Soundscape: Perception Analysis of Acoustics Environment in Intensive Care Unit

I. R. N. Azzahra, R. F. Fela, J. Sarwono, and 2 more authors

In INTER-NOISE and NOISE-CON Congress and Conference Proceedings, Hong Kong. More Information can be found here , Aug 2017

Abs HTML PDF

The quality of acoustic environment can be measured according to soundscape dimensions. The study of soundscape dimension has been conducted mainly in an urban area and resulting soundscape dimensions for soundscape rating. In an urban area, the main noise sources (such as traffic noise or construction noise) usually do not contain valuable information. However, in Intensive Care Unit (ICU) of a hospital, the characteristic of main noise sources are different. Our previous study shows that some main noise sources (medical equipment) are expected to be heard and important for the nurse because the noise sources contain valuable information regarding the patient’s condition. Due to the characteristic difference, the soundscape dimensions in ICU might be different from the urban soundscape The aim of this study is to define soundscape dimension for ICU from the nurse’s point of view. A questionnaire using semantic differential scale was distributed to the nurses to find out their perception to acoustics environment in ICU. The result is then analyzed using principal component analysis to determine the soundscape dimensions for ICU. The soundscape dimension obtained can further be used as a tool in evaluating the soundscape of ICU. Initial study shows that the soundscape of ICU can be described in terms of calmness, dynamics, and information.
INTER-NOISE

Data Recording of Audio Stimulus for Continuing Nurse Training Using Auditory Virtual Environment

R. F. Fela, J. Sarwono, S. S. Utami, and 2 more authors

In INTER-NOISE and NOISE-CON Congress and Conference Proceedings, Hong Kong. More Information can be found here , Aug 2017

Abs HTML PDF

A healthcare environment noise may reduce the nurse’s ability and therefore, evaluating nurse’s perception in noise environment and knowing their response in medical provision is being studied. During the VR era, auralisation has been used as one of the main technique to provide audio files which represent the real acoustic phenomenon within the room model. It is however very challenging to use auralisation technique in spaces where activities are continuously happening such as in the healthcare units. Therefore, the conventional RIR method is not suitable for this purpose of simulation. This paper proposed a combined technique in subjective evaluation and sound level measurement to determine the recording data scenario of the audio stimulus for auditory virtual environment. Questionnaire from 20 respondents (nurses) with 3 working shifts in 24 hours were collected in an ICU of National Hospital in Indonesia, a national benchmark ICU with less noise environment. Data collected demonstrates that nursing care is conducted mostly near the patient-beds, medical observations are being reported near patient’s bed and inside the nurse station, and getting drugs from the pharmacy are the most daily work task done by nurses. The graph also demonstrates that medical equipment including mechanical ventilator and EKG monitor were the most influential sound sources which necessarily exist in ICU. These highlighted equipment respectively generate 75.5 dBA and 69.6 dBA then similarly decrease over the stepped distance about -2 dBA. Based on questionnaire result, 38 recorded files in binaural and ambisonic format have been collected to represent nurse’s movement within ICU rooms due to their daily tasks and can be further considered for VR design for training purpose.
ICSV

Design Considerations for Improvement of Speech Intelligibility and Privacy in A Hospital Intensive Care Unit

S. S. Utami, R. J. Yanti, J. Sarwono, and 2 more authors

In 25th International Congress on Sound and Vibration 2017, London, UK. More Information can be found here , Jul 2017

Abs HTML PDF

Acoustics problems that mainly occur in an ICU are noises from medical staff activities and equipment. Excessive sound from human conversation is considered as unwanted noise. Activities within the ICU requires the staff movements, which also creates non-stationary conversation area especially near the patient’s bed. Despite of this condition, nurse station is the location where the noise produced from conversation will raise significantly and where lack of speech privacy becomes a concern and higher chance of less confidentiality of medical information. The research is based on computer modeling that enables inclusion of scattering effects. By principle, the acoustical treatment in a nurse station is to avoid speech leakage from this area as well as to improve the speech intelligibility. Several improvement ideas is by installing hanging abfussor at the nurse station, hanging abfussor near the patient bed, and a glass partition at the nurse station. The speech privacy distance (rP) and distraction distance (rD) can be predicted using ISO 3328-3:2012 method and %Alcons formula based on the STI values. To evaluate the effectiveness of the improvement, an auralization scenario of the activities within the ICU is used in a survey. All the improvement design proposed did not create significant impact to the acoustic quality based on the rP and rD values. However, this paper demonstrates a good technique to predict the range of easiness of nurses’ communication in their workspace while they continuously move. It is given by the distraction distance that is used here to predict speech intelligibil-ity of certain source to receiver distance. The result is a guideline for nurse station design optimization using three design approaches above by utilizing abfussor (absorber and diffuser) as part of the building interior.

2016

JNTETI

Sinkronisasi Sensasi Pendengaran dan Visual di Dalam Dunia Simulasi Komputer

S. Utami, M. Navvab, and R. Fela

Jurnal Nasional Teknik Elektro dan Teknologi Informasi. More Information can be found here , 2016

Abs HTML PDF

Current application of virtual environment synthesized by computer simulation still focuses on generating visual cues, while auditory cues has not yet been optimized to create individual movements within the virtual space. Auditory cueing is limited to the ability in detecting the presence of a sound source without further perception. Auralization of variety of spaces with the impact of a variety of architectural elements were synthesized. The auralization data were produced using current computer-simulation technique where both visual and acoustics properties of the room boundaries are considered. Noticeable differences in the auditory cognition of the virtual spaces are measured through the use of subjective evaluation. The acoustical quality are also evaluated utilizing several objective acoustic measures. Diffuser panels were able to increased ability of sound localization up to 12.5% out of the 40 respondents. These panels also increased the speech intelligibility quality, indicated both by the C50 which increased by 89% with 97.5% of respondents approved this condition. Analysis of the subjective and objective measures indicates to what extent the alteration of the simulation is required to create a noticeable auditory cues that enhance the auditory and visual sensation of the virtual space.

2014

Conference

Analysis of Dome Geometry Effect on Acoustic Conditions of AR Fachruddin Mosque UMM Using CATT-Acoustic

R. Ikhwanuddin, S. S. Utami, and R. F. Fela

In Conference paper. More Information can be found here , 2014

Abs HTML PDF

Mosque is central building for worship activities in Islamic community. There are similarities in terms of acoustic perspective, that the whole activities within will produce sound speeches. The existence of dome in the acoustic mosque space could potentially produce concentrated distribution of the sound energy. Acoustical comfort for each activity in mosque requires speech intelligibility as well as tranquil atmosphere for concentration in worship. The acoustical quality prediction is produced by using computer modelling CATT-Acoustic. Numerical calculations were performed based on the hybrid method. The ring structure is part of mosque ceiling. Analytical calculations are performed to investigate sound ray reflections due to variations of the dome geometry. The model analysis is based on the investigation of/ratio. Dome geometry produces focusing effect by/ratio 0.74, 0.83, and 0.92, respectively for models 1, 2, and 3. Means of T30 models 1, 2, and 3 for congregational prayers are 2.6, 2.4, and 0 seconds. Means of T30 models 4, 5, and 6 for Quran recitation are 2.7, 2.3, and 0 seconds. Areas near mihrab are 14× 6, includes 4–5 first row produces poor quality of clarity. Clarity for congregational prayers, C50 are-4.7,-4.6, and-4.1 dB, respectively for models 1, 2, and 3. Clarity for Quran recitation, C80 values are-1.3,-0.7, and-0.8 dB, respectively for models 4, 5, and 6. The results demonstrate that increasing the height of the dome structure will produced better clarity and speech intelligibility, given the smaller T30 and the higher both of C50 and C80 values.