Skip to content

Select your region & language

Global

Region

Technical Report: Sound Quality Evaluation 1

In home appliances such as air conditioners and refrigerators, office equipment such as photocopiers, and automobiles, there is a growing demand for comfort through reduced noise levels, in addition to the pursuit of basic performance. Traditionally, physical quantities such as sound pressure level and acoustic power were used to evaluate the noise levels of such equipment. However, since the final judgment of noise is left to human hearing, psychoacoustic evaluation quantities, which are related to physical quantities, are now being used in evaluation. Psychoacoustic evaluation quantities quantify how humans perceive sound, making it easier to obtain clues as to how to improve sound quality from sound quality evaluation quantities. The method of using psychoacoustic technology and feeding it back into the design of equipment is called acoustic design.

1. Introduction

Sound quality evaluation is a technology that uses psychoacoustics to quantitatively analyze sound in accordance with how people perceive it. Unlike conventional mechanical evaluation based on physical quantities, this sound quality evaluation technology, based on human senses, is expected to add a human touch to future living environments. For example, while current air conditioners are much quieter than before, their sound is still quite mechanical. However, by analyzing sound quality with sound quality evaluation technology and advancing sound improvement technology, it may be possible to transform the sound into a pleasant, gentle, and natural sound. Sound quality evaluation technology is also expected to have applications in various fields in the future, such as automobiles and other vehicles, as well as office equipment.

2. What is psychoacoustic technology?

Psychoacoustics, which forms the basis of sound quality evaluation, is a technology that utilizes a wide range of fields such as engineering, medicine, and psychology, and is a technology that is expected to grow even further. For example, there are many parameters for psychoacoustic evaluation, and it is necessary to use them appropriately by selecting and combining them depending on the sound being evaluated. In particular, psychology remains a research topic, as people's sensibilities differ depending on their preferences and social environment. However, it is expected that these issues will be resolved in the future, and that the technology will be put into practical use even further.

3. Popularization of sound quality evaluation

Psychoacoustic technology-based sound quality evaluation was initially used to evaluate the sound inside automobile cabins.

Loudness, a measure of sound intensity, is standardized by ISO. Psychoacoustic evaluation is also gaining attention in industries other than the automotive sector. In 1998, standards for noise labeling were established in the home appliance industry, specifying that loudness should be used in addition to noise level. Furthermore, it is attracting attention as a quantitative evaluation measure for noise reduction in industries such as office equipment.

Psychoacoustics as an academic discipline has been studied since the 1930s, as shown in the table on the next page. Loudness, one of the psychoacoustic evaluation quantities, has been standardized by ISO (for steady-state sounds only), as mentioned earlier, and is beginning to be widely used.

This loudness measurement system was refined into its current form by Zwicker in Germany, based on the accumulated research of various individuals. Many other psychoacoustic evaluation quantities were devised relatively recently, and many of them were proposed by researchers associated with Zwicker's laboratory. Several parameters are still being studied, and this technology is expected to develop even further in the future.

1930 年代

Measurement of Pure Tone Loudness / H. Fletcher, WA Munson

1940 年代

Measuring the temporal effects on loudness / WRGarner

1950 年代

The beginning of physiological research into hearing / G. Bekesy
Re-measurement of pure tone loudness / DW Robinson, RS Dadson
How to calculate loudness / LL Beranek

1960 年代

How to calculate loudness / SSStevens
Loudness and Masking / E. Zwicker
The Influence of Temporal Structure on Loudness / E. Zwicker

1970 年代

How to Calculate Loudness / E. Zwicker
A Proposal for Sharpness / G. Bismarck
A Proposal for Roughness / E. Terhardt

1980 年代

Proposal for Variation Intensity / H. Fastle
A Proposal for Desirability / W. Aures

SWIPE

4. Sound Perception (Auditory Perception) and Sound Quality Evaluation System

Psychoacoustic evaluation (PEP) is derived from studies of the structure of the human ear and the function of the auditory nerve, as well as from auditory experiments conducted on many people. Auditory experiments involve exposing people to sounds and asking them how they perceive them, then examining their responses. A person's response to sound varies slightly depending on their age, past experiences, physical condition at the time, and surrounding environment such as temperature. Therefore, PEP is determined by conducting tests on many people multiple times and statistically processing the results.

So why is psychoacoustic technology necessary? Traditionally, the evaluation and countermeasures for sounds emitted from equipment have focused on physical quantities (sound pressure level, noise level, power level, 1/3 octave spectrum, FFT spectrum, etc.). However, it is humans who actually hear the sound. Furthermore, physical sound countermeasures, which generally aim to reduce physical quantities, have their limitations. Of course, it is not impossible to completely eliminate sound if one is willing to cover the equipment with extremely expensive, large, and heavy materials, but this is not practical. Moreover, eliminating sound does not necessarily create a comfortable environment. Therefore, instead of simply reducing physical quantities such as sound pressure, a quantitative measure of how humans perceive sound is called a psychoacoustic evaluation quantity. Since psychoacoustic evaluation quantities are related to physical quantities, it is easier to obtain clues on how to improve sound quality from sound quality evaluation quantities. This method of feeding back into equipment design is called acoustic design.

Psychoacoustic evaluation quantity
unit
Explanation
loudness

sone

Sound volume
Steady-state sounds are standardized by ISO 532B.

Loudness level

phon

Loudness displayed logarithmically

Sharpness

acum

instep height
This is what you feel when the balance between low and high frequencies is skewed towards the high frequencies.

Roughness

asper

Rough texture, gritty, shaky
This is what you feel when the loudness fluctuates in short cycles.

Fluctuation intensity

vacil

Fluctuation is the opposite of smoothness.
This is what you feel when loudness fluctuates in a slow cycle.

tonality

tu

Pure pitch
This indicates how many pure tone components are contained in a sound.

AI

%

speech intelligibility
Evaluation quantity for assessing speech clarity

SWIPE

To help you understand sound quality evaluation, we will introduce some actual analysis examples. The figure below shows examples of analysis for six machine sounds with different volume levels.

The upper line (green) shows the results of an analysis using A-weighted sound pressure level (noise level), and all six sounds show the same value. In contrast, the lower line (pink) shows the results of an analysis using "loudness," a measure of sound intensity, and reveals the differences between the six sounds. When we actually listen to these sounds, they sound different in intensity, just like the loudness results below. This example shows that the intensity of sound as heard by humans cannot be evaluated solely by an evaluation quantity based on sound pressure, such as A-weighted sound pressure level.

Next is the sound of two motorcycles passing by one after the other.

We performed A-weighted sound pressure level and loudness analysis on this sound as well, similar to the example mentioned earlier.

In the upper line (green), which analyzes using characteristic sound pressure level (noise level), the values for the two motorcycles are the same. In contrast, in the lower line (pink), which analyzes using "loudness," the sound of the second passing motorcycle is louder. When you actually listen to these sounds, the sound of the second motorcycle sounds louder, just as the loudness results suggest.

Now, there are two machine sounds.

The second sound sounds louder than the first. However, when we examine the A-weighted sound pressure level, the value for the second sound is lower. This is an example of how the relative loudness measured by A-weighted sound pressure level can sometimes be the opposite of the relative loudness when actually heard.

So why don't the sound pressure values correspond to the actual perceived loudness of the sound?