Keywords

1 Introduction

Operation feedback sounds are used on a lot of electric devices. When you operate a button on a touchscreen, these sounds are heard. Several studies reported these operation feedback sounds can improve the operational feeling and reduce a number of operating errors [1]. We carried out experiment to develop the sound design guideline for improving the feeling of button press [2].

Almost all feedback sounds used on contemporary electronic devices are “beeps.” Regarding these sounds, we propose the use of various timbres on feedback sounds. We think they can express product’s own unique images and make it easier for users to distinguish between products, allowing users to feel that certain products are better to operate than others.

The purpose of this study is to build a timbre image scale for designing feedback sounds for operation. It can be useful for adapting sound image to product design, which means it can be a tool to share the timbre image between a product designer and a sound designer. We build the image scale based on an impression evaluation experiment.

2 Impression Evaluation Experiment for Building Image Scale

2.1 What Is an Image Scale?

An image scale [3, 4] is a visual representation of the relationship between stimulus and its impression. The procedure of building an image scale begins with the participants evaluating each of stimulus (e.g., color, sound) image using the semantic differential method. Then, a factor analysis is conducted on the obtained data to extract some factors of the underlying image. And then, rectangular coordinate system could be made using these factors, and the stimulus are plotted on the system.

As an example, Kobayashi et al. built the color image scale [3]. They made connections between colors and adjectives and mapped them on a rectangular coordinate system. The vertical axis of the color image scale is “SOFT-HARD” and the horizontal axis is “WARM-COOL”. The scale was built using 180 pairs of adjectives for color. The scale works as a color design support tool for designers who deals with color design.

We think that operation feedback sounds are part of product design, and the timbre image scale could work for the product designers.

2.2 Stimuli

In general, sounds consist of loudness, pitch, and timbre. In this research, we focused on timbre and prepared 72 feedback sounds that were different in timbre.

Sounds Designed by Sound Designers.

We asked three sound designers to design operation feedback sounds using various timbres.

We showed the designers the acoustic guideline [2] (see Fig. 1). The guideline indicates the conditions of amplitude envelope: attack time is between 0 ms and 20 ms, decay time is between 50 ms and 100 ms, and total time is between 60 ms and 110 ms. When you hear the operation feedback sound which meets the conditions, you can get the good operation feeling.

Fig. 1.
figure 1

Guideline for designing feedback sounds

Full size image

Each of the designers had his own plan for designing sounds. One designer designed simple and plain sounds, another designer designed the sounds like sound-effects used in video games, and the third designer designed sounds using recorded material, such as beating or knocking on something. They designed 45 feedback sounds in total.

Sounds Designed by Authors.

We designed feedback sounds based on sin wave given the following our rules.

  1. 1.

    Fundamental Frequency: 500 Hz, 1000 Hz, 1500 Hz

  2. 2.

    Harmonics:

    1. (a)

      Adding 2nd harmonic tone with 1/2 amplitude to the fundamental frequency

    2. (b)

      Adding 3rd harmonic tone with 1/3 amplitude to the fundamental frequency

    3. (c)

      Adding 4th harmonic tone with 1/4 amplitude to the fundamental frequency

  3. 3.

    Attack time and decay time of the amplitude envelope:

    1. (a)

      Attack time: 20 ms, Decay time: 50 ms

    2. (b)

      Attack time: 5 ms, Decay time: 70 ms

    3. (c)

      Attack time: 0 ms, Decay time: 90 ms

We designed 27 feedback sounds in total by combining these parameters.

2.3 Bipolar Adjective Pairs for Evaluation

As already mentioned, we think that operation feedback sounds are part of product design. Therefore, we decided to use the adjectives, which are used in design process, for semantic differential method. Saito et al. told that there were 12 clusters of adjectives related to design: “functional beauty,” “sense of accuracy,” “kindness, cuteness,” “sense of quality, soft feeling,” “sense of color,” “elegance,” “strength, familiarity,” “traditional beauty,” “creation and satisfaction,” “modernity, fresh beauty,” “luxury,” and “noticeability” [5]. We selected 15 adjectives from these 12 clusters, at least one adjective from one cluster, and added another 5 adjectives. Then, we established pairs of adjectives by adding an antonym adjective to each. Table 1 shows the bipolar adjective pairs for evaluation.

Table 1. Selected bipolar adjective pairs
Full size table

2.4 Participants and Procedure

We used semantic differential method for the evaluation. 26 participants (17 men and 9 women) took part in the experiment. Participants evaluated the sound stimuli by just listening to them (they did not operate the button.) 72 stimuli were presented randomly, and the time for evaluation was one minute per stimulus. While a stimulus sound was being replayed 20 times, the participants evaluated it with 20 bipolar adjectives on a 7-point scale using a semantic differential method.

3 Result and Discussion

We conducted a factor analysis to confirm the impression structure. When we conducted the last factor analysis, the items “appropriate-inappropriate,” “not tire of-tire of,” and “likable-disagreeable,” were excluded. Principal factor analysis and varimax rotation were carried out as the method of factor analysis.

Table 2 shows the calculated factor loadings. Specifically, adjectives that had high loadings for factor 1 were “wild-refined,” “rough-accurate,” “matte-glossy,” “masculine-feminine,” “classical-futuristic,” “modest-noticeable,” “cheap-expensive,” “natural-artificial,” and “decorative-functional.” Those with a high loading for factor 2 were “mature-childish,” “calm-lively,” and “formal-casual.” Those with a high loading for factor 3 were “plain-gorgeous,” “complex-simple,” and “serious-airy,” and those for factor 4 were “cold-mild” and “bold-delicate.” We named these four factors; artificiality, liveliness, gorgeousness, and gentleness.

Table 2. Factor loadings of each of 17 bipolar adjective pairs in the semantic differential
Full size table

4 Building Timbre Image Scales

We built two timbre image scales using the four factors we had gotten from the analysis. The first (and the main) scale is shown in Fig. 2. We used the artificiality factor as a horizontal “NATURL - ARTIFICIAL” axis, and the liveliness factor as a vertical “LIVELY - CALM” axis. We call this scale the “artificiality-liveliness scale” (see Fig. 2). We also made the second scale which is “gorgeousness-gentleness scale” by using the third and fourth factors as the “SIMPLICITY - GORGEOUSNESS” axis and the “MILDNESS - COLDNESS” axis.

Fig. 2.
figure 2

Artificiality-liveliness scale

Full size image

We mapped the sounds on the scales using the factor scores as the coordinate values, and also mapped the adjectives using the factor loadings.

5 Conclusion and Future

In conclusion, we built two timbre image scales. These could be useful for selecting or designing operation feedback sounds, the sound images of which are adapted to the product design.

In the future, we will integrate these image scales in the iPad application “Sign Sound Selector (SSS)” [6]. SSS is a tool to share the timbre image between a client (ex. a product developer) and a sound designer. It could be used for designing sound sign and UI sounds.