金屬手工具撞擊聲對消費者感性意象之研究-以榔頭為例

Abstract

近幾年，產品聽覺上的設計，大多以使用介面的電子聲音為分析的主要對象，希望可以針對產品本質上的聲音(本體材質的物理碰撞聲、機械聲)，分析其傳遞給使用者怎樣的感覺意象。為產品提升品質形象，有利的操作聲音，作為產品的特色之一。

本研究將針對金屬手工具之榔頭撞擊所產生出來的聲音做物理性與心理性的調查與分析，尋求符合產品意象的聲響，為了使榔頭金屬聲能對應的更多，將收及其他金屬聲作為錨點，將分析的金屬聲範圍擴大。研究內容大致上為：

第一，除了目標物榔頭聲，還廣泛的收集金屬產品撞擊時所產生的聲音，在收錄的過程以高敏感度儀器，要求在相同的環境下進行測錄，錄製了有140個金屬聲音樣本，經由專業錄音軟體的剪修，在對金屬聲音做些物理性(頻譜等等)的分析，調整出各種不同屬性的金屬聲，讓鑑別度提升，使得實驗效果能較為顯著。

接著，以140個聲音樣本數進行兩次的專家前測與訪談，在專家小組的建議與篩選下，留下80個聲音，作為受測者前測的實驗樣本。分析受測者前測實驗的數據後，經篩選保留了40個較有差異性的聲音樣本，作為正式實驗的樣本。

另一方面，針對國內外相關研究中經常使用的聲音形容詞進行收集，對於聲音以心理性與物理性兩族群做形容詞分類，選出30個聲音形容詞。在針對小型前測實驗結果對30個形容詞進行因子分析，再根據各因子解釋變異量篩選出較適合評價金屬聲音的形容詞語彙，最後歸納出10個形容詞作為正式SD調查之量尺。 針對40個聲音樣本數，要求30位受測者，分別以10個聲音形容詞，進行SD評量，形容詞的量表以「非常不強烈」至「非常強烈」由左到右分七等級。選擇有隔音版設備的聲音實驗室進行，以專業的錄音器材播放，實驗的時間，避開白天上班上課時段，以傍晚以後的時間為準，降低環境的噪音干擾。

加以分析時，將30位受測者對每一個聲音在各形容詞量表的評量結果平均，以每個形容詞平均量表值分別為依變項，40個聲音樣本的物理特質數據為自變項，進行迴歸分析，可得每一個形容詞的迴歸方程式，即可找出影響個意象的聲音構成以應用在產品感性需求上。

In recent years, the product design has explored the characteristics of sounds produced by a product, especially the electronic sounds implemented in interface design of a product, and their influence on the users’ feeling, in order to improve the quality of product image.

This research analyzed the physical and psychological characteristics of hitting sounds produced by hammers or other metal hand tools, and investigated the felt images of theses sounds through survey. First, to include the range of studied sound samples as broad as possible, we collected and recorded varied hitting sounds generated by hammers as well as by other metal products. The sound samples were recorded by high sensitive instruments in the real working environment. There were 140 sound samples recorded and analyzed on their physical composition (such as frequency spectrum, etc.) by specialized recording software. Some samples were adjusted to change some attributes in order to make them varied enough to distinguish from other samples.

Experts’ then were invited to screen 80 sound samples out of the pool of 140 sound samples according to the discernible power of the samples. The number of sound samples was further reduced to 40, through analyzing the data in the pilot test, for formal experiment.

On the other hand, we collected the sound adjectives often used in relevant researches. Subjects were asked to group these adjectives according to the similarity in meaning. Cluster analysis was conducted on the classified data to conclude some identical groups which include adjectives describing psychological feeling and physical feature of the sounds. Thirty sound adjectives were selected from each group of adjectives and were verified for their suitability through the pilot test. At last, 10 adjectives were decided to be used as scales in the formal sound image evaluation test through the factor analysis in the pilot test.

Thirty subjects were recruited to evaluate the 40 sound samples on the 10 scales of sound adjectives. The evaluation was rated on 7 levels form 'not feel strong' to 'feel strong' on each adjective. The sound stimuli were presented to subjects by specialized recording apparatus in a sound isolated laboratory during the experiment session which was arranged in the time after dusk to reduce the noise of the environment.

The averaged scores over 30 subjects of the evaluation of each sound sample on each adjective were regarded as dependent variables; whereas the physical compositions of sound samples were treated as independent variables. Accordingly, for each adjective, the corresponding linear regression analysis was conducted to conclude the influence of sound features on each perceived image. This result is expected to assist product designers to design appropriate product sounds with desired images.