Nanogenerator

The evaluation of sound from fabrics is a sound evaluation process where the noise of fabrics can be linked to emotions. This was first attempted by research conducted in 2002.

Background

The experience of wearing clothes is a necessity in life that people need in order to protect their bodies from conditions in the world around them. As part of this experience, the need to feel comfortable and at ease in one’s clothes is required by some people, whether it is for health or personal reasons. The levels of comfort from wearing clothes can be determined under three of the five methods of perception.

Vision is obviously important for style and appearance in most cultures, as is touch when determining the materials of clothing for specific target audiences, such as children or the elderly. Surprisingly, the sound of clothes is becoming an interesting topic of study for researchers and clothing manufacturers alike, as it has been previously implied that sounds fabrics can make from the daily wear of clothing may contribute to the sensory experience. In order to determine the levels of comfort that can be achieved from various fabrics, a system that can evaluate the sound levels omitted by these materials would have to be considered and applied.

Evaluating Sound from Fabrics

In order to deliver the sound evaluation system that has been discussed in this research, the following sound parameters need to be obtained:

• The level pressure of total sound (LPT)
• Sound colour factors (∆L and ∆f)
• Constants of autoregressive models that are based on the fast Fourier transform algorithm (FFT). These coefficients are ARC, ARF and ARE.
• The loudness (z) and sharpness (z) parameters from psychoacoustic models
• Mechanical properties from the Kawabata Evaluation System (KES).

To generate these properties, the fabric sounds need to be created in a consistent process. A device was constructed to assist in this effort, known as a measuring apparatus for fabric noise (MAFN). The MAFN creates the sound by rubbing two pieces of fabric together on a belt-pulley. The fabrics are attached to a moveable plate which is connected to two weights suspended on a wire rope. The weights counterbalance to create a pulling effect causing the fabrics to move, creating friction and audible noise levels. A microphone is positioned at the point where the two pieces of fabric make contact in order to capture the best possible sound.

In order to record how the sound of fabrics affects human sensations, an evaluation of the sounds created is required from a number of different subjects. In the research phase, thirty university students were asked for their assessment in the form of a questionnaire. The different sounds made by seven different fabrics were to be judged on by seven different attributes, clearness, highness, loudness, pleasantness, roughness, sharpness and softness, in the format of free-modulus magnitude estimation (FMME). A FMME is a method of magnitude estimation. In this example, after listening to the different sounds, participants were asked to give a numerical value to the first sound they heard. Then, based on that value given, they were asked to give a value to the next value. Once all of the values have been given, an average of multiple participants’ can be collated. The end result is the data being presented in a rational scale.

The psychological characteristics that are categorised in this process are softness, loudness, sharpness, clearness, roughness, highness and pleasantness. In order to establish a connection between these characteristics and the objective measurements that have been recorded from the FMME, stepwise block linear regression is employed as a modified regression model. This is achieved by following the steps that Kawabata proposed in the KES to establish prediction models for primary hand values. The value for each sensation is shown in a table of values in the report with the model applied being shown below:

$Y=C_0+\sum _{i=1}^k{C}_i\frac{X_i-\overline{X_i}}{{\theta }^i}$

• $Y$ = sound sensation
• $X_i$ = $i$th objective measurement
• $\overline{X_i}$ = mean value of the $i$th objective measurement or its logarithm
• ${\theta }^i$ = standard deviation of the $i$th objective measurement or its logarithm
• $C_0$ = the constant
• $C_i$ = the coefficient

Conclusive Results

The conclusion of the study illustrates that thicker fabrics, such as Melton, showed higher ∆f than thinner ones. Thinner fabrics, such as tropical wool, which are resistant to shearing showed the lowest values for loudness and the highest for ∆L. Most wool is perceived as loud but tropical wool, however, is perceived as clear and pleasant.

Similar Methods

There have been similar lines of research to this method. One attempt studied in particular how the specific sound of fabric being rustled can be unpleasant. This method of evaluation was measured through Multidimensional Scaling analysis, or MSA. Multiple linear regression was also adapted here, which shows that it is a method commonly applied in this field. This method was only applied to one emotion, unpleasantness, whereas the study from Yi, Gilsoo, Youngjoo and Casali in 2002 was able to determine the participant’s response from a choice of seven different emotions.

A similar study measured the same seven emotions and identical sound parameters but adopted a Semantic Differential Scale (SDS) for the participants. Although the ranking system differed slightly for the subjects, the results of the two procedures were quite similar.