Park, Mun Hum

Preferred name

Park, Mun Hum

Alternative Name

Park, Munhum

Main Affiliation

Music Science and Engineering

munhum.pa@kmitl.ac.th

Now showing 1 - 5 of 5

Some of the metrics are blocked by your
consent settings
Measurement of oral-binaural room impulse response by singing scales
(2019-01-01)
Park, Munhum
Oral-binaural room impulse responses (OBRIRs) are the transfer functions from mouth to ears measured in a room. Modulated by many factors, OBRIRs contain information for the study of stage acoustics from the performer’s perspective and can be used for the auralization. Measuring OBRIRs on human is, however, a cumbersome and time-consuming process. In the current study, some issues of the OBRIR measurement on human were addressed in a series of measurement. With in-ear and mouth microphones, volunteers sang scales, and a simple post-processing scheme was used to refine the transfer functions. The results suggest that OBRIRs may be measured consistently by using the proposed protocol, where only 4~8 diatonic scales need to be sung depending on the target signal-to-noise ratio.
Some of the metrics are blocked by your
consent settings
Noise in the intensive care unit and its influence on sleep quality: A multicenter observational study in Dutch intensive care units
(2018-10-05)
Simons, Koen S.
;
Verweij, Eva
;
Lemmens, Paul M.C.
;
Jelfs, Sam
;
Park, Munhum
Background: High noise levels in the intensive care unit (ICU) are a well-known problem. Little is known about the effect of noise on sleep quality in ICU patients. The study aim is to determine the effect of noise on subjective sleep quality. Methods: This was a multicenter observational study in six Dutch ICUs. Noise recording equipment was installed in 2-4 rooms per ICU. Adult patients were eligible for the study 48 h after ICU admission and were followed up to maximum of five nights in the ICU. Exclusion criteria were presence of delirium and/or inability to be assessed for sleep quality. Sleep was evaluated using the Richards Campbell Sleep Questionnaire (range 0-100 mm). Noise recordings were used for analysis of various auditory parameters, including the number and duration of restorative periods. Hierarchical mixed model regression analysis was used to determine associations between noise and sleep. Results: In total, 64 patients (68% male), mean age 63.9 (± 11.7) years and mean Acute Physiology And Chronic Health Evaluation (APACHE) II score 21.1 (± 7.1) were included. Average sleep quality score was 56 ± 24 mm. The mean of the 24-h average sound pressure levels (L<inf>Aeq, 24h</inf>) was 54.0 dBA (± 2.4). Mixed-effects regression analyses showed that background noise (β =-0.51, p < 0.05) had a negative impact on sleep quality, whereas number of restorative periods (β = 0.53, p < 0.01) and female sex (β = 1.25, p < 0.01) were weakly but significantly correlated with sleep. Conclusions: Noise levels are negatively associated and restorative periods and female gender are positively associated with subjective sleep quality in ICU patients. Trial registration: Www.ClinicalTrials.gov, NCT01826799. Registered on 9 April 2013.
Some of the metrics are blocked by your
consent settings
A preliminary analysis of the sources of noise in an open-plan neonatal intensive care unit
(2017-01-01)
Park, Munhum
;
Vermeulen, Ronald
;
Laroche, Sabine
;
Gillies, Murray
The acoustic environment in neonatal intensive care units (NICUs) is known to be hardly conducive to the treatment and development of preterm infants. Many studies report the results of acoustic surveys in NICUs, where some studies also investigated the sources of noise. These studies employed a human observer to annotate the noise sources on site, which unfortunately makes the results vulnerable to the Hawthorn effect. Furthermore, most noise source surveys were carried out for a few hours at particular times of day, thus revealing only part of the complete situation. To overcome the shortcomings of the previous studies, the noise sources in an open-plan NICU were investigated in the current study by an offline annotation of 24-hour continuous audio recordings made inside and above an incubator. The results showed that a total of 17 noise sources could be identified in 30159 intervals that occupied 22.1/24 hours, where up to 7 concurrent sound events were present in each interval. The events produced by staff and non-patients’ activities (excluding speech) were counted most frequently, 12878 times, and the alarms from life-supporting/-monitoring devices were present for the longest accumulated duration, 19.5/24 hours. Noise level data were also extracted from the calibrated recordings and analyzed in synchrony with the annotation results. The results showed that the strongest noise source was the patient’s own noise (e.g., crying) inside the incubator and the other patients’ noise above the incubator. Excluding the patient’s own noise, the alarms and the non-patients’ activity noises contributed to the total acoustic energy inside the incubator by 43% and 42% and above the incubator by 53% and 26%, respectively. It is suggested that a better management of alarm systems and an appropriate staff education may improve the acoustic environment in NICUs.
Some of the metrics are blocked by your
consent settings
Analysis of warning sounds used in public transportation systems
(2017-01-01)
Chuprasert, Pavarit
;
Phupantachsee, Nawasch
;
Park, Munhum
For the safety of passengers, warning sounds are used in public transportation services, for example, when the train doors are closed. Although the audibility of these sounds is normally required and must be regulated by transportation authorities, the associated noise levels are often unnecessarily high, which frequently lead to the passengers’ discomfort and complaints. In this study, we investigated the door-closing sounds used in the Bangkok Mass Transit System (BTS) Skytrain by evaluating the noise levels and spectral characteristics. The maximum A-weighted sound pressure level (L<inf>AFmax</inf>) was repeatedly measured at 7 representative positions within each car for two types of train, which differed in terms of the location of loudspeakers. The range of L<inf>AFmax</inf> was found to be 67.4 dBA to 97.2 dBA with the median at 77.1 dBA, and the train type was found to significantly influence the noise level (p=0.006), and so was the location in train car (p<0.001). The spectral analysis revealed that the door-closing warning sounds consisted of strong ~3-kHz tone bursts (with a few harmonics), which were found to be 15~20 dBA above the ambient level. The outcome of the study suggests that the warning sounds used in the BTS Skytrain are likely to distress the passengers due to the excessive playback level and the tonality. Further discussions are made on methods to reduce the psychological impact of the warning sounds on passengers while maintaining a good audibility.
Some of the metrics are blocked by your
consent settings
The effect of "Twinkle Twinkle Little Star" on short-term memory
(2018-01-01)
Park, Munhum
;
Chuprasert, Pavarit
;
Kloemwilai, Achcharaphan
;
Fahkrajang, Napat
;
Sawetratanastien, Pruch
Irrelevant speech effect (ISE) refers to the decrease in cognitive performance under the influence of a sound which is not related to the task. Typically, ISE has been quantified for speech noise by the rate of errors in recalling visually presented items (serial-recall task). Despite the limited effect size, non-speech sounds have also been shown to disrupt the short-term memory in serial-recall task, of which the spectro-temporal characteristics may easily be adjusted to investigate some aspects of ISE. In the current study, a well-known tune ‘Twinkle Twinkle Little Star’ was used to create a set of stimuli that differed in three factors: 1) Instrument type (played either on a MIDI piano or in pure tones); 2) number of notes per second (two or four); 3) playing order (original, reverse or random). The results showed some trends that the piano sound disrupts the task slightly more than the pure tone, and so does the four-note version than the two-note. When played in a random order, the error rate was higher than the original-/reverse-order versions, whereas the latter two conditions were almost equal. These trends may be explained in relation to the temporal and spectral variations between sound tokens.

Park, Mun Hum

Filters

Settings

Sort By

Results per page