Room Acoustics Computer Modeling Research Articles

Room acoustic computer modeling—A case study to compare industry standard software with measured results

This paper extends a previous publication presented at the Inter-Noise 2005 conference [1] and investigates the predictive performance of some industry standard room acoustic software packages compared with measurements. The case study specifically examines the Twyford Theatre, comparing predictions of room acoustic parameters obtained from ODEON, EASE, and Treble software with measurements. To conduct the study, impulse response measurements were performed in the theater using the swept sine method. Architectural drawings and on-site inspection measurements were utilized to construct ODEON, EASE, and Treble models for prediction purposes. The findings of this study provide insights into the predictive performance of these industry standard room acoustic software packages with comparison to measured results. C. Field and S. Shimada, Acoustic Computer Modelling: A Case Study to Compare Predictions by CATT and Odeon with Measured Results, Inter-Noise 2005.

Effects of varied row spacing and adding cushions, carpet and occupants on pew sound absorption

This work examines how the individual variations of row spacing, as well as the presence of cushions, carpet and occupants, combine to influence the absorption characteristics of church pews as a function of sample perimeter-to-area (P/A) ratios. Scale models were used to measure the interactive effects of the four test variables on church pew absorption characteristics. The effects of row spacing on frequency-averaged pew absorption coefficients (500–4kHz) were generally quite small and were smallest for pews without cushions. The variations of absorption increments with varied row spacing tended to be smallest when the pews were less absorptive, i.e. with less cushions, carpet or occupants, similar to the previously studied model theatre chairs. Resonant absorption effects seem to influence the measured absorption of the model pews in the 125 and 250Hz octave bands and these effects are expected to be modified by the height of the pew backs and the spacing between pews. The incremental absorption coefficients of adding absorptive materials, such as carpet or cushions, to the pews were different than the absorption characteristics of cushions or carpet tested independent of the pews, and these differences varied with frequency. Using the resulting absorption coefficients calculated for each block of pews in a church as input to a room acoustics computer model is expected to provide the best currently possible predictions of acoustical conditions.

Experimental investigation of chair type, row spacing, occupants, and carpet on theatre chair absorption.

This paper examines how the individual variations of chair type, row spacing, as well as the presence of occupants and carpet, combine to influence the absorption characteristics of theater chairs as a function of sample perimeter-to-area (P/A) ratios. Scale models were used to measure the interactive effects of the four test variables on the chair absorption characteristics, avoiding the practical difficulties of full scale measurements. All of the test variables led to effects that could lead to important changes to auditorium acoustics conditions. At mid and higher frequencies, the various effects can usually be explained as due to, more or less, porous absorbing material. In the 125 and 250 Hz octave bands, the major changes were attributed to resonant absorbing mechanisms. The results indicate that for accurate predictions of the effective absorption of the chairs in an auditorium, one should use the P/A method and reverberation chamber tests of the chair absorption coefficients to predict the absorption coefficients of each block of chairs and use these results as input in a room acoustics computer model of the auditorium. The application of these results to auditorium acoustics design is described, more approximate approaches are considered, and relations to existing methods are discussed.

Comparing the time variance of orchestral instrument directivities in Mozart’s symphony in G-minor: First movement

Multiple channel anechoic recordings of the musical instrument parts in Mozart’s Symphony in G-Minor (First Movement) are analyzed to study how the directivity patterns of each instrument varies with time. Static directivity patterns are well-documented for many musical instruments, but studies on how their directional patterns vary with time are not as common. Changing directional patterns in time, however, have been found to impact the realism of room acoustic computer modeling simulations. Previous work at the University of Nebraska has suggested a method for studying the time variance of musical instruments across a number of simultaneously recorded channels in an anechoic chamber. The method involves time-windowing each channel and analyzing how the overall directivity index changes across time and frequency. Comparisons of results from some of the fourteen instruments included in this Mozart symphony are presented. [Work supported by a UNL Undergraduate Creative Activities and Research Experience Grant.]

Examining the relationship between total acoustic absorption and late lateral energy (GLL).

Listener envelopment, the sense of being immersed in the sound field, has been shown to correlate with the objective parameter late lateral energy (GLL). Barron proposed that GLL is primarily a function of the ratio of reverberation time to auditorium volume or the total acoustic absorption [Barron, Appl. Acoust. 62, 185–202 (2001)]. An investigation has been conducted to examine this proposed theory with measurement data and predicted values from a room acoustics computer modeling program. Room impulse response measurements have been taken in the Belding Theatre in downtown Hartford, CT. Features of the hall include variable absorption in the form of curtains, both within the hall and also within small adjacent coupled volumes. Measurements were taken in several absorption configurations. A detailed room acoustics computer model of the hall has been created in ODEON v9.22 to examine the accuracy of the model’s prediction of GLL. The model was initially validated using the measured parameters of reverberation time (T30), early decay time (EDT), and clarity index (C80). A comparison of the measured and predicted results will be discussed, along with an analysis of the relationship between GLL and total acoustic absorption. [Work supported by a University of Hartford Greenberg Junior Faculty Grant.]

Investigations of Incorporating Source Directivity Into Room Acoustics Computer Models to Improve Auralizations

Room acoustics computer modeling and auralizations are useful tools when designing or modifying acoustically sensitive spaces. In this dissertation, the input parameter of source directivity has been studied in great detail to determine, first, its effect in room acoustics computer models and, second, how to better incorporate the directional source characteristics to improve auralizations. The room acoustics computer model was initially validated in terms of accurately incorporating the input source directivity. The next study demonstrated that using directional sources over an omnidirectional source in room acoustics computer models produces significant differences both in terms of calculated room acoustics parameters and auralizations. A recently proposed technique for creating auralizations using multichannel anechoic recordings has been examined with numerous subjective studies, applied to both solo instruments and an orchestra. Through these studies, this process was shown to be effective in terms of improving the realism and source width of the auralizations in a number of cases and also modeling different source orientations. In addition, this approach was applied to modeling an entire orchestra with individual sources, in three different configurations, for the first time. This method shows great promise as a new technique for auralizing both solo instruments and an entire orchestra. [Copies may be obtained from Michelle Vigeant, michelle.vigeant@gmail.com]

Evaluations of output from room acoustic computer modeling and auralization due to different sound source directionalities

Evaluations of the predictions and auralizations from the room acoustic modeling program, ODEON, have been run using three directional source types with the same sound power: (a) an omni-directional source; (b) three sources with realistically-directional characteristics based on measurements from real instruments (grand piano, violin, and singing voice); and (c) an artificial, extremely directional beaming source. Objective analyses have been run for nine source/receiver combinations in a simple hall on three acoustic parameters: relative sound pressure level (SPL), reverberation time (T30), and clarity index (C80). Auralizations were subsequently created for two source/receiver combinations and used in subjective testing with 28 subjects. Results show that, objectively, differences in SPL were negligible for the majority of cases. Some differences in T30 and C80 were found between the omni-directional and realistically-directional sources; however, subjects did not perceive any corresponding differences when comparing the auralizations, possibly due to the limited directional octave band data available. Subjects did significantly differentiate between auralizations from the omni-directional source and the extreme beaming source. Subjective results from comparing these two sources in terms of reverberation, clarity and realism were generally consistent with objective data, although source/receiver combination and musical track had some influence on the outcomes.

Concert hall canopy height comparison and computer model resolution efficacy

This line of inquiry is twofold: (1) It uses computer modeling to compare the height of an adjustable concert hall ceiling canopy with established room acoustics metrics, and (2) It does so simultaneously with two computer models, one at a rough resolution, and one at a resolved resolution. The height comparison looks to establish best practices in operation of such a canopy in a hall; how high should it be set? The model resolution comparison looks to establish best practices in room acoustics computer modeling; how detailed must the model be?

Determination of absorption coefficients for audience seating from in situ measurements

Audience seating constitutes the majority of fixed sound absorption in most music performance venues. An understanding of the sound absorption characteristics of audience seating is therefore essential for accurately predicting reverberation times and for evaluating different chair constructions during the design process. Room acoustics measurements were recently conducted in a small concert hall immediately before and after installation of audience seating. These measurements allow the determination of absorption coefficients for audience seating for use in the Sabine equation [L.L. Beranek and T. Hidaka, J. Acoust. Soc. of Am. 104, 3169–3177 (1998)] and in room acoustics computer modeling software. A comparison of these in situ absorption coefficients along with those determined from reverberation room measurements will be discussed.

Practical methods to define scattering coefficients in a room acoustics computer model

To predict acoustics of rooms using computer programs based on geometrical assumptions, it is important that scattering is included in the calculations. Therefore scattering is usually included in terms of scattering coefficients which are assigned to each surface telling the software the ratio between the part of the reflected energy which is not being reflected specularily and the total reflected energy. However the effective scattering coefficient of a surface depends not only on the roughness of the surface material indeed diffraction caused by limited dimensions of the surface as well as edge diffraction also causes scattering. For complex rooms it can be difficult to give a reasonable estimate to the magnitudes of scattering coefficients if these should also include diffraction and even if these frequency dependent coefficients could be obtained in the design phase, the processes of obtaining the data becomes quite time consuming thus increasing the cost of design. In this paper, practical methods to define scattering coefficients, which is based on an approach of modeling surface scattering and scattering caused by limited size of surface as well as edge diffraction are presented. The predicted and measured acoustic parameters in real rooms have been compared in order to verify the practical approaches recommended in the paper.

Investigation of Sound-Absorbing Material Configuration of a Smart Classroom Utilizing Computer Modeling

Advancements in information and instruction technology have led to the evolution of a new type of classroom referred to as “smart classrooms”. These have enhanced audio-visual equipment, computers and seating layouts designed to facilitate interactive learning. Placement of different sound-absorbing finishes in an efficient manner improves the listening conditions within the classroom and reduces the amplification of internally generated noise such as that from computers and instructional equipment. This study investigates the best overall configuration of sound-absorbing material placement and characteristics of surface treatment in an attempt to enhance the listening conditions in smart classrooms. A typical layout of a smart classroom was modeled and simulated using room acoustics computer modeling. Acoustics indicators such as Reverberation Time (RT), Sound Clarity (C50) and Speech Transmission Index (STI) were used for comparing alternative cases in optimising sound-absorbing material characteristics and placement. Additionally, measurements were conducted in similar classrooms to assess the magnitude and characteristics of generated noise. To determine the impact of the resulting background noise simulations were carried out. The resulting configuration of sound-absorbing material for a typical smart classroom can also be utilized by architects and educational institutions to enhance the acoustics of existing conventional classrooms in the process of being converted or upgraded.

Sound-pressure level distribution in a long, narrow hallway: Measurements versus results from a computer model with scattering from surface roughness and diffraction

The sound-pressure level distribution down a long, narrow hallway due to a sound source at one end does not decrease linearly along the length of the hall. This characteristic may be due to the changing behavior of scattering that occurs down the length of the hallway, which is distance- and angle-dependent. A new scattering method that incorporates these effects has been implemented in the room acoustic computer modeling program, odeon [C. L. Christensen and J. H. Rindel, Forum Acusticum, Budapest (2005)]. A comparison of the results from an odeon model with real-world measurements along a long, narrow hallway on the campus of the Technical University of Denmark is provided in this paper. [Work supported by the NSF.]

Predicting and auralizing acoustics in classrooms

Although classrooms have fairly simple geometries, this type of room is known to cause problems when trying to predict their acoustics using room acoustics computer modeling. Some typical features from a room acoustics point of view are: Parallel walls, low ceilings (the rooms are flat), uneven distribution of absorption, and most of the floor being covered with furniture which at long distances act as scattering elements, and at short distance provide strong specular components. The importance of diffraction and scattering is illustrated in numbers and by means of auralization, using ODEON 8 Beta.

Evaluation of room acoustic qualities and defects by use of auralization

Auralizations generated by room acoustic computer modeling programs may be used as a tool for evaluation of acoustic qualities and defects, some of which are not easily detected by objective measures. Examples include reverberance, flutter echoes, perceived room size and distance, apparent source width, listener envelopment, and sound propagation in coupled rooms. In order to reach a sufficiently high level of realism in auralizations for such room acoustic applications, it is necessary that all parts of the chain are modeled with certain accuracy. In particular, it is found that the modeling of source directivity and the late room reflections (the reverberation tail) need careful consideration in order to achieve reliable and realistic-sounding results. However, when implemented in the software the practical use can be simple and quick.

Differences in directional sound source behavior and perception between assorted computer room models

Source directivity is an important input variable when using room acoustic computer modeling programs to generate auralizations. Previous research has shown that using a multichannel anechoic recording can produce a more natural sounding auralization, particularly as the number of channels is increased [J. H. Rindel, F. Otondo, and C. L. Christensen, Proceedings of the International Symposium on Room Acoustics: Design and Science 2004, Paper V01 (2004)]. Further studies evaluating the quality of auralizations using one-channel, four-channel, and 13-channel anechoic recordings have been pursued. The effect of changing the room’s material properties was studied in relation to turning the source around 180 deg and on the range of acoustic parameters from the four- and 13 beams. As the room becomes increasingly diffuse, the importance of the modeled directivity decreases when considering reverberation time. However, for the three other parameters evaluated (sound-pressure level, clarity index, and lateral fraction), the changing diffusivity of the room does not diminish the importance of the directivity. The study therefore shows the importance of considering source directivity when using computer modeling. [Work supported by the National Science Foundation.]

An evaluation of differences due to changing source directivity in room acoustic computer modeling

This project examines the effects of changing source directivity in room acoustic computer models on objective parameters and subjective perception. Acoustic parameters and auralizations calculated from omnidirectional versus directional sources were compared. Three realistic directional sources were used, measured in a limited number of octave bands from a piano, singing voice, and violin. A highly directional source that beams only within a sixteenth-tant of a sphere was also tested. Objectively, there were differences of 5% or more in reverberation time (RT) between the realistic directional and omnidirectional sources. Between the beaming directional and omnidirectional sources, differences in clarity were close to the just-noticeable-difference (jnd) criterion of 1 dB. Subjectively, participants had great difficulty distinguishing between the realistic and omnidirectional sources; very few could discern the differences in RTs. However, a larger percentage (32% vs 20%) could differentiate between the beaming and omnidirectional sources, as well as the respective differences in clarity. Further studies of the objective results from different beaming sources have been pursued. The direction of the beaming source in the room is changed, as well as the beamwidth. The objective results are analyzed to determine if differences fall within the jnd of sound-pressure level, RT, and clarity.

Sensitivity of room acoustic parameters to changes in scattering coefficients

This project uses the room acoustics computer modeling program, ODEON, to investigate the sensitivity of room acoustic parameters to changes in scattering coefficients. Particularly, the study is interested in determining if the results from certain room models are more sensitive to scattering coefficients than from other models, due to their geometry or absorption characteristics. If so, how can one quantify a model’s susceptibility to being sensitive to scattering? Various models of three real spaces in Omaha, Nebraska are tested. The predicted reverberation, clarity, and spaciousness parameters are compared at various receiver locations, while the scattering coefficient of all surfaces is varied from 0 to 0.1, 0.3, 0.5, and 0.8. The resulting data are analyzed by frequency according to the (1) average absorption of the room; (2) magnitude variation of absorption within the room; (3) spatial distribution of absorption within the room; and (4) level of model detail. Initial results indicate that parameters studied may show more sensitivity to scattering coefficients in models that have a wider range of absorption values, more disparate distribution of absorption, and lower detail level. Various schemes that include these aspects are proposed for computing a model’s sensitivity to changes in scattering.

On the importance of early reflections for speech in rooms.

This paper presents the results of new studies based on speech intelligibility tests in simulated sound fields and analyses of impulse response measurements in rooms used for speech communication. The speech intelligibility test results confirm the importance of early reflections for achieving good conditions for speech in rooms. The addition of early reflections increased the effective signal-to-noise ratio and related speech intelligibility scores for both impaired and nonimpaired listeners. The new results also show that for common conditions where the direct sound is reduced, it is only possible to understand speech because of the presence of early reflections. Analyses of measured impulse responses in rooms intended for speech show that early reflections can increase the effective signal-to-noise ratio by up to 9 dB. A room acoustics computer model is used to demonstrate that the relative importance of early reflections can be influenced by the room acoustics design.

Room acoustic computer modeling: The effects of scattering on transient power flow

The CATT© room acoustic modeling program allows the computation of acoustic power projected to three mutualy perpendicular planes at a point in space integrated over a specified time period. By selecting successive time periods it is possible to observe the oscillatory nature of the transient power flow from first arrival to the long-term settled value during diffuse-field reverberant decay. The computation is normalized, and displayed as a fraction of the total power. Introducing scattering on the room surfaces results in a significantly smaller peak amplitude of power flow oscillation. Several cases are shown involving different room sizes and shapes and different orientations of source and receiver. Conclusions are made on the number of mean-free path intervals required to ‘‘settle’’ the computer-modeled acoustic field.