Stephen Dance

Computer prediction of sound distribution in enclosed spaces using an interference pressure model

Abstract Current computer models for the prediction of sound distribution in enclosed spaces use intensity to represent the sound radiating from various sound sources. This approach reduces the amount of information to be calculated and makes the modelling system simple enough to be represented on a desktop computer. With the advent of new computer technology it has now become practical to model sound distribution using pressure, including phase information, in empty enclosed spaces. Measurements taken have shown that in empty enclosed spaces sound interference can be observed both close to and, to a lesser extent, far from the sound source. Independent measurements have also shown the measured sound pressure level at the same floor position to fluctuate significantly with receiver height. A comparison of predictions by both an intensity and a pressure based model showed that the general shape and accuracy of the propagation curves predicted by the pressure model were similar to those measured, whereas the intensity based model gave curves that did not predict interference effects.

Modelling of sound fields in enclosed spaces with absorbent room surfaces. Part I: performance spaces

This paper introduces a three-part report describing research into the use of Millington absorption coeAcients in the computer modelling of sound fields in enclosed spaces with absorbent room surfaces. The historical background to the prediction of reverberation time is presented together with three types of computer models used in the investigation. In part one, the computer models are described, the Millington reverberation time formula is validated, Millington absorption coeAcients are derived and the sound field in a concert hall is predicted. This enables the accuracy of the three types of computer models to be compared and the eAect of applying diAerent absorption coeAcients to be studied. Part two of the report consists of an extensive investigation into the prediction of reverberation time in multiple configurations of an experimental room with absorbent material partially covering the room surfaces. This determined the accuracy of reverberation time formulae and the computer models using both standard and Millington absorption coeAcients under controlled conditions. The final part contributes a verification of the accuracy of the predictions using Millington absorption coefficients in a factory space with a barrier installed, and a refined diAraction model based on a ray-tracing model. # 1998 Elsevier Science Ltd. All rights reserved.

Modelling of sound fields in enclosed spaces with absorbent room surfaces Part II. Absorptive panels

Abstract This is the second part of a report on the modelling of sound fields in enclosed spaces with absorbent room surfaces. Both Sabine and Millington absorption coefficients are used to further investigate which approach produces the most accurate predictions. Classical formulae and geometric acoustic models, previously detailed, were used to predict the reverberation time in a test room with highly absorbent room surfaces. In addition the test room was configured with a partially absorptive surface. This configuration has previously only been predicted inaccurately. It was shown that the Millington formula predicted more accurately than either the Sabine or the Eyring formulae in the configurations of the room investigated. Verification of the previous results were confirmed when two of the computer models gave consistently more accurate predictions using Millington absorption coefficients than Sabine absorption coefficients.

Computer Prediction of Insertion Loss Due to a Single Barrier in a Non-Diffuse Empty Enclosed Space

This paper describes the development of an efficient barrier model for the prediction of sound distribution in non-diffuse empty enclosed spaces. Diffraction is modelled using an extended version of the Ondet and Barbry computer model, RAYCUB, which is a proven model used to predict sound distribution in empty and fitted non-diffuse enclosed spaces. As RAYCUB is based on geometric acoustics, it is not possible to directly model diffraction around a barrier. Diffraction around barriers is known to cause only localised, frequency dependent effects on sound distribution in acoustically complex environments such as factories. It was intended that a barrier model should impose a minimum of additional computation time on the factory noise prediction program. Three methods of approximating barriers were developed, all based upon a simplified implementation of the geometric theory of diffraction. The barrier models were tested in two configurations of an empty test space. The model REDIR gave more accurate results than the original RAYCUB model, especially at lower frequencies.

Minimal input models for sound level prediction in fitted enclosed spaces

Abstract Highly advanced computer models for the prediction of sound fields in rooms are now available. However, these tools are complex and require a skilled acoustician to use effectively and hence there is a need for more simpler models. A simple model needs to be accurate and quick to use, but most importantly should require a minimum amount of input data to construct the model. This is only achievable if the scope of the model is reduced to one or two acoustic parameters. Three simple models were investigated two empirical based formulae and a geometric acoustic model. The models were validated in six configurations of an experimental room simulating a textile workshop and two real engineering workrooms. It was found that all the models executed near instantaneously, but the obtainable prediction accuracy and consistency was proportional to the amount of input data. The models are now available on the Web, running directly inside Netscape or Internet Explorer.

Modelling of sound fields in enclosed spaces with absorbent room surfaces Part III. Barriers

Abstract In the third paper on the modelling of sound fields in enclosed spaces with absorbent room surfaces, both Sabine and Millington absorption coefficients have been used for an acoustic barrier to investigate which approach produces the most accurate insertion loss predictions in a factory. The accuracy of three computer models in predicting the sound field on both sides of the barrier, using the two types of absorption coefficients is presented. In addition, a refinement of the diffraction approximation used in one of the models is described and validated. The results lend support, as do the previous results of this investigation, to the premise that Millington based absorption coefficients for absorbent material give consistently more accurate predictions than the standard absorption coefficients, when used in geometric computer models for the modelling of sound.

Straw bale sound insulation: Blowing away the chaff

Popular opinion states that straw bale walls are good at isolating sound. Cheap load bearing straw bale houses could contribute substantially to low carbon sustainable construction. However, literature on the subject was found to be highly anecdotal. The paper presents a summary of nine laboratory and field sound insulation test reports and two especially commissioned tests. Data were compared to European party wall sound insulation criteria, and it was found that straw walls could perform as well as, but sometimes worse than, conventional constructions, due to poor performance at low frequencies. Better performance could help to promote the use of straw bales in multi-unit housing. It was found that by adding a plasterboard layer on studs to just one side of a plastered straw bale wall would allow the construction to pass all of the criteria reviewed.

IPods listening levels on London Underground

In the last 5 years the prevalence of the iPod/MP3 players has grown exponentially. The use of such in‐earphones under urban conditions has been reported widely in the press at the anecdotal level. This study compared listening levels under quiet conditions and that representative of a London Underground train journey. Calibrated recordings of underground trains running in tunnels were played through loudspeakers in an anechoic chamber, whilst pop music or speech podcasts were played through the iBuds. Thirty‐three participates listened to the iPod whilst a Binaural Head and Torso measured the noise levels through the in‐earphones. The participants had time to adjust the volume setting on each occasion. A small audiometric study was undertaken for those participates with particularly high and low volume settings a week after the tests were completed. Results show very high volume settings were used when speech was played with the train noise.

Modeling of sound fields in enclosed spaces with absorbent room surfaces

Commercial computer models are now generally available for the prediction of sound levels in industrial workshops. Four models have been used in this investigation; including raynoise v3.0, ramsete v1.3, rayscad+v3.3, and fame v1.0. These models were validated in rooms with highly absorbent room surfaces using standard and Millington‐based absorption coefficients. The rooms included hypothetical, performance, experimental, and industrial spaces. Measurement data were available for the six central octave bands. The results are presented for sound levels and reverberation time, in addition to which the execution time for each model is detailed. It was found that Millington‐based absorption coefficients provided consistently more accurate predictions for each type of space than standard absorption coefficients.

Teachers' voice parameters and classroom acoustics—A field study and online survey

Many studies have suggested that teachers have a significantly higher rate of voice problems than the general population. In order to better understand the possible influences of room acoustics on different voice parameters, a study has been carried out by London South Bank University which involved measurements of voice parameters for teachers working in classrooms with varying acoustic conditions. Data relating to the voice, including the average speech sound level, fundamental frequency, and phonation percentage, were captured using an Ambulatory Phonation Monitor (APM) which measured directly from skin vibrations in the neck, thereby excluding the effects of other noise sources in the environment. The measured voice parameters were compared with the room acoustic data for the classrooms involved, which were surveyed separately from the voice measurements. In addition to the field measurements, an online questionnaire was undertaken with the support of two UK teacher trade unions. This was designed to ...