Richard D. Godfrey

Far field performance of an anechoic chamber used for speech and hearing research.

The department has a 13×7×8 ft fully anechoic chamber whose commissioning records were misplaced. The graduate acoustics class measured the acoustics of the chamber from 250–2000 Hz in preparation for future research projects. The focus was on the far field and its limits (near field, and approach to wedge tips). Four groups of students made measurements along the same path. The difference in the levels at a location from that at a reference location was calculated. The mean and standard deviations of these differences were then calculated to compare these measurements with the inverse square law theory. Some measurements were also made in the near field, and closer than one‐quarter wavelength from the wedge tips. The average of the means from 12–74 in. was 0.73 dB below theory. The average 95% confidence limits of these data in this range was +/− 1.1 dB. These deviations compare favorably with the acceptance criteria in the ANSI standard. The near/far field guideline also appears to be reasonable. Measur...

The N number, a speech privacy metric for rating walls, revisited

In 1962, Cavanaugh, et al., published the results of research which lead to a single number rating system for speech privacy of walls referred to as the N number [‘‘Speech Privacy in Buildings,’’ J. Acoust. Soc. Am. 34, 475–492 (1962)]. The N number is derived from measured one‐third octave band transmission loss data weighted using factors signifying the relative contribution to the articulation index. When the N number is summed with factors for source room floor area, source room speech use, measured adjacent room background noise rating, and the privacy requirement, the N number correlated subjective reactions observed in case studies better than average wall transmission loss alone. Owens Corning, the sponsor of this research, published the N number along with sound transmission class (STC) for several years. With the emergence of regulation of speech privacy in health care facilities, the Health Insurance Portability and Accountability Act (HIPAA), it seemed reasonable to revisit the N number. This ...

Protocol for measuring dishwasher sound power levels

A leading retailer of home appliances came to the OC acoustics laboratory to evaluate the noise generated by dishwashers sold in their stores. A protocol based on a methodology used in Europe was used. It was modified to account for North American building standards. The standard provides an overall sound power level determination for the complete dishwasher cycle with and without the drying portion of the cycle. As a result customers making a purchasing decision can compare the noise performance of the various products using the same yard stick.

Measured characteristic impedance and propagation constant of some common materials using the two cavity method

The acoustics of porous rigid frame materials is characterized by their characteristic impedance and propagation constant. These properties can be measured using the two cavity method. A sample is placed in an impedance tube with an air space behind the sample and ahead of the rigid termination. Measurements are made in accordance with ASTM E 1050. This procedure is repeated with the same sample, but a different air space. These data are analyzed as outlined by Seybert et al. [J. Acoust. Soc. Am. 86, 637 (1989)] to yield the characteristic impedance and propagation constant. This method was used to characterize fiber glass boards with a range of flow resistivities from 2600 to 56000 mks rayls/m. Repeat measurements were made, and these measurements were compared with the Delaney and Bazley model, [Applied Acoustics (3), 1970]. Measurements were also made for cotton shoddy and cellulose, and power law models were developed. The method produced repeatable results comparable to the Delaney and Bazley model w...

Sound transmission class (STC) is not a good predictor of performance of insulated wood frame gypsum walls used as interior partitions in most North American homes

Home owners say that insulating interior walls improves the acoustic environment. Based on STC alone, no perceptible difference is expected. To define ‘‘improved,’’ ethnographic and laboratory studies were conducted. Ethnographic studies in 33 homes, revealed owners want quieter, less reverberant environments, including rooms where added isolation is desired. Families lives are 24/7, leading to frustration that they cannot use their homes without disturbing others. Laboratory jury studies were conducted where 35 listeners rated the relative isolation of insulated and uninsulated walls. Noise sources included broadband and real home noises. Insulated walls were perceived to perform better than uninsulated walls in all cases. Noise control engineers know that STC is only a quick screening tool (actual sound transmission loss should be used to estimate noise reduction between rooms). This is what the jurors appeared to sense. Jury ratings and the midfrequency average SPL correlated reasonably well. The STC is determined by a structural resonance near 125 Hz. Above this band, insulation has a significant impact on transmission loss (perceptible, 6 dB average). A new rating system is needed that quantifies what actual listeners hear in quiet room environments. A model using some form of room criteria is suggested.

Use of surrogate samples to study variation of diffuse field absorption coefficients of fiberglass with altitude

ASTM C 423 identifies air temperature and relative humidity as significant parameters, but does not address air density effects. In previous papers, normal and diffuse field analysis showed significant changes in predicted absorption coefficients with altitude. These predictions were validated experimentally for normal incidence in a vacuum chamber, and by using surrogate samples, thus showing the feasibility to study altitude effects in a single laboratory. Mechel design charts are normalized by two parameters. One is not dependent on air density. The other (R) is the ratio of flow resistance and the impedance of air. At constant thickness, the effect of lowering air density can be studied by increasing the sample flow resistivity. Samples with flow resistivity ratios of 1.25 and 1.5 were studied in a diffuse field following ASTM C 423 methodology. These values correspond to altitudes of 6000 and 10 700 ft in altitude, respectively. These results followed the predicted trends. At both altitudes, sampl...

Use of surrogate samples to study variation of absorption coefficients of fiberglass with altitude

ASTM C 423 identifies air temperature and relative humidity as significant parameters, but does not address air density effects. At constant temperature, air density decreases approximately 20% from sea level to 5000 ft altitude. In previous papers, normal and diffuse field analysis showed significant changes in predicted absorption coefficients with altitude. These predictions were validated experimentally for normal incidence by making measurements in a vacuum chamber. Reverberation chambers cannot withstand depressurization. They also exhibit significant interlaboratory measurement variability. Another method was soot. The Mechel design charts are normalized by two parameters. One is not dependent on air density. The other is the ratio of flow resistance and the impedance of air. If thickness is held constant, the effect of lowering air density can be studied by increasing the sample flow resistivity of the sample. This surrogate sample should emulate absorptive performance at high elevations in sea le...

A comparison of background noise levels and reverberation times measured in unoccupied elementary classrooms

The key performance criteria listed in ANSI S12.60‐2002, Acoustical Performance Criteria, Design Requirements and Guidelines for Schools, are that the maximum background noise is limited 35 dBA, and that the maximum reverberation time is limited to 0.6 seconds in the most common classroom size. Limits on sound transmission properties of the room envelope are also made. If these performance criteria are met, each student, no matter where he or she sits in the classroom, will be in an environment that affords adequate speech intelligibility, i.e., an adequate opportunity to understand the teacher’s words correctly. During the standard development process, the author had the opportunity to work with his colleges at The Ohio State University. The OSU group made measures in 34 classrooms in the Columbus area (inter‐city, suburbs, and rural; public and private). The author analyzed these data, and compared the results with ANSI S12.60‐2002 performance criteria. It was found that the background noise levels were...

Estimating reverberation time in classrooms by the power method

The Sabin equation relates reverberation time to room volume and absorption. If an acoustic source of known power is placed in a room, the absorption is also related to the power per bandwidth, volume, and pressure. Combining the Sabin equation with the power equation and solving for reverberation time yields a relationship for reverberation time as a function of volume, pressure, and the power of the source. This approach has been reported to deviate from the decay method at low frequency. In a laboratory reverberation room, it yielded reasonable results in the frequency range of interest in classrooms. Reasonable results were also achieved in real classrooms in the 500‐ to 2000‐Hz frequency range. Based on these findings, it may be possible to make acceptable reverberation time measurements in less time with less sophisticated instrumentation than the decay method.

An experimental study of the variation of absorption coefficients for fiberglass with altitude

ASTM C 423 specifies requirements for air temperature and relative humidity during decay rate measurements, but does not address variations in air density even though this parameter decreases with increasing altitude. At constant temperature, air density decreases by approximately 20% from sea level to 5000 feet altitude, resulting in about 20% decrease in its acoustic impedance. In a previous paper, an analysis based on Mechels design charts (normal and diffuse) showed significant changes in predicted absorption coefficients with altitude. Due to the difficulties in depressurizing a reverberation chamber, a normal incidence study was conducted to experimentally investigate this effect by placing an impedance tube in a vacuum chamber. Fiberglass materials were tested at various amounts of vacuum. The measured results followed reasonably well those predicted by the analysis for normal incidence. Based on these results, comparison of absorption coefficients measured at substantial altitude with those measur...