MIGUEL BOYAYANEars and microphones: system simulates human hearingMIGUEL BOYAYAN
The musicians are competent, the instruments are top quality, the repertoire is pleasant and the conductor gets the best performance from his orchestra, but the sound does not flow that well. The musicians are not listening to one another, and the audience does not enjoy the finely tuned musical group. This is a problem that arises in new or old concert halls. The architecture, the design and, above all, the material used in finishing of the hall are not always the appropriate for the purpose of obtaining the best sound. To study this kind of phenomenon and its variations , and also to offer solutions and an accessible tool for music professionals, a group of researchers from the University of São Paulo has developed a software to analyze the sound in concert halls, theaters and auditoriums. This software can be downloaded from the Internet for free. By employing this software, researchers can conduct measurements of the environment to be analyzed with the help of a laptop, microphones and amplifiers.
Analysis of the variables encompasses 12-parameter values that provide information on sound characteristics linked to the intelligibility of the sounds, the balance between sharp and low sounds, the distribution of the energy of sound inside the hall and the perception of the position of the sound sources, such as loudspeakers, voices and musical instruments. This also involves reverberation, and the time during which a sound, after being concluded, continues resonating with its multiple reflections on the surfaces comprising the environment.
“The software has functions that make it more encompassing and comprehensive than other products on the market”, says Fernando Iazzetta, professor of the Laboratório de Acústica Musical e Informática/Lami ( Musical Acoustics and Computer Laboratory) at the Music Department of the University of São Paulo’s College of Arts and Communications (ECA). He coordinated a project funded by FAPESP, which resulted in the development of this software. “We study acoustics and the software measures the parameters of sound in the environment to be analyzed.” The software has calculation tools that help conduct the acoustic analysis of environments; the measurements are taken on the basis of special sound signals captured by microphones placed in the positions that are to be analyzed in the facility. The software calculates the impulsive response, which is the reaction of the room to the mechanical impulse of sound, characterized as waves that reverberate in the air between 20 and 20 thousand times per second, or frequencies ranging from 20 to 20 thousand hertz, which correspond to the band perceived by human ears.
The impulsive response provides a timely model for the room’s reflections, registering the behavior of the sound in the environment. “The numbers provided by the software allow us to compare the results obtained with the perception of the musicians, or with the subjectivity of each musician in relation to the room, whether the sound is good or not for the instruments”. This is done through subjective tests in which some musical excerpts were taped in a recording studio where there are no echoes nor reflections of sound waves which are played on a CD recording in the rooms being analyzed. These excerpts are taped in the identical manner in each room and are later submitted to experienced musicians who can compare the recordings and evaluate how a specific piece of music sounds in different rooms. Another possibility in more advanced studies with the software is to use a binaural system for testing recording studios. This binaural system tests human hearing. The equipment is equipped with a dummy which resembles a mannequin. The mannequin’s ears are shaped exactly like human ears, making it easier to capture the sounds through special embedded microphones.
Musical partnership
The software, called AcMus, started being developed in 2002, in partnership with the Computer Sciences Department at the University of São Paulo’s Institute of Mathematics and Statistics/IME. The project was headed by professor Fábio Kon; the research group included researchers from the School of Architecture and Urban Planning/FAU, ,the Polytechnic School/Poli and the Physics Institute, all from the University of São Paulo. The software was validated at the end of 2006 and ready for downloading from the Internet on: http://gsd.ime.usp.br/acmus/ . The software is being used by professionals to help in the renovation and construction of theaters and recording studios, such as for example the theater of the Federal University of the State of Rio de Janeiro/Unirio. “The adaptations were based on measurements produced by the system to provide the environment with superior sound quality”, says professor Iazzetta.
“We have heard that two doctorate students, one from Poli-USP and the other one from the College of Engineering at the State University of Campinas/Unicamp), have already used the software in their doctorate work. We have also heard of master’s theses and a number of scientific initiation projects that have also resorted to the software. But any musician wishing to analyze and adapt a rehearsal room in his home, for example, can use the AcMus, even with home microphones and loudspeakers.” The program can also be used in classrooms and small auditoriums for lectures. An analysis of the results helps improve the acoustics in much smaller auditoriums and consequently improves the level of understanding of a lecture. For example, voice projection problems existed in some of the Centros Integrados de Educação Pública/Cieps (public schools) in Rio de Janeiro. The problem in the Cieps, which were opened up in the eighties, was that the sound coming from one classroom interfered with the sound of another classroom, because the walls separating the two rooms did not go all the way up to the ceiling. As a result, the noise level reduced the students’ concentration capacity and obliged the teachers to speak in loud voices, which in turn damaged the teachers’ vocal cords. The problem was remedied after an acoustic study had been conducted, and the solution was to block the empty spaces with bricks.
It won’t be long before the software will also be able to simulate acoustic behavior in a theater that is still in the architectural design phase. “The same measurement parameters of an actual room are calculated in the design of the construction, in three dimensions. This ensures good quality of sound and eliminates possible problems before construction work is begun”, says Iazzetta. It will be possible to analyze the geometry of the room, simulate the acoustic behavior and prepare a detailed description of the materials to be used. “This simulation takes into consideration not only the geometry, but also the acoustic performance of each surface in the environment. The materials that coat these surfaces have sound absorption coefficients that determine the quantity of sound energy that is absorbed or reflected back to the environment. For example, if a given material absorbs 20% of the waves, it reflects the other 80%.”
Hard, smooth materials such as walls and bricks tend to reflect the sound, while soft, porous materials such as carpets, foam and upholstery, absorb the sound waves. These characteristics have to be kept in mind when choosing the concert hall seats. The seats have to absorb the same amount of sound absorbed by an average person. “The reason is because the sound cannot vary if some of the seats are empty, or even prior to the performance, when the sound rehearsal is conducted in an empty auditorium.” To describe the acoustic measurements in more detail, the AcMus takes other data, such as humidity and room temperature, into account.
To validate the software and test its functionality, the researchers conducted a comparative analysis of the acoustic behavior of six government-owned concert halls and auditoriums. In São Paulo, the researchers took the equipment to five theaters: Camargo Guarnieri amphitheater, with 367 seats, the auditorium of the Memorial da América Latina, with 1.600 seats; the Teatro Municipal de São Paulo theater, with 1.580 seats; the Teatro São Pedro theater, with 636 seats and the Sérgio Cardoso theater, with 864 seats. The sixth facility visited by the team was the 460-seat Teatro Clara Nunes theater, located in the neighboring city of Diadema. The measurements indicated that the theaters with the best sound acoustics for a musical performance were the Municipal de São Paulo and Teatro São Pedro; it is no coincidence that these two facilities are the musicians’ preferred facilities of the six. “The software compares the sound played by the computer with the sound captured by the microphones. Subtracting one sound from the other leaves only the room. In this respect, the most stable facilities are the Teatro Municipal and the São Pedro”, says Iazzetta.
In spite of the good sound stability, the researchers detected some points of sound disturbances in the Teatro Municipal, which was inaugurated in 1911. “Right in the middle of the auditorium, in the central aisle, there is a very unbalanced spot because it is the point focused on by the dome located in the middle of the auditorium. As a result, the sound is concentrated in one place and does not spread out, which jeopardizes the quality of the sound.” The problem disappears in the seats located a little over one meter away on each side of this central point. Another characteristic of the Municipal is the behavior of sound in the niches in the balconies on the sides of the theater. The sound waves are reflected by the walls and by the ceiling, forming resonances that alter the sound balance. “The Municipal, like many other older theaters, has many surfaces with angles and curves and irregular finishing which act as good sound dispersing agents.”
Low frequencies
In the other theaters, structural and material-related problems hinder the performance of sound. At the Sérgio Cardoso theater, a huge camera behind the stage acts as a resonating box which blends the low frequencies and the musicians find it difficult to hear themselves playing. At the Memorial da América Latina complex, the problem lies with the carpeted walls which absorb the sharp frequencies. The solution for theaters like the Sérgio Cardoso and other theaters with acoustics problems is to implement amplification with microphones and a sound system that offsets the sound deficiencies. A classic example was the inauguration of the Credicard Hall concert facility in São Paulo, in 1999 when musician João Gilberto complained loudly about the sound during his show, for which the audience booed him. “The problem was that the sound would reverberate against the back of the stage and resonated with a perceptible delay. It was possible to hear the echo in the first rows of the auditorium”, Iazzetta recalls. “This happened because at the time there was no absorbent covering or diffusing material at the back of the stage.”
The researchers involved in the development of the AcMus want it to be greatly publicized. To this end, in addition to being distributed free of charge, the software is a multi-platform tool. It can be used in Windows, Linux and Macintosh environments. Its structure also allows functions to be added on by means of plug-ins. It is also an open source code tool, which allows for its wide use and collaboration from users in the sense of developing the system.
The Project
Project and simulation of acoustics at musical performances (nº 02/02678-0); Modality Thematic Project; Coordinator Fernando Henrique de Oliveira Iazzetta – USP; Investment R$ 135,425.25 and US$ 3,527.00 (FAPESP)
