改裝商用錄音筆作為噪音劑量量測裝置可行性評估

Abstract

聽力損失所造成職業傷害為在臺灣勞工職業病發生件數的第一名。在國內，約有26.2%的勞工聽力損失介於25至40 分貝(dBA)，甚至有9.6 %的勞工聽力損失達40 dBA以上。為了更瞭解噪音傳入耳部時，特別是中頻音至高頻音對人體的影響，本研究特別開發出一套可以配掛於外耳道外的噪音量測設備，用以測量接近耳道的噪音頻率及暴露劑量，以提供噪音性聽損的相關研究所需。 經過多方的嘗試錯誤，本研究已經成功的將商用量產的MP3錄音筆改裝成可以進行耳外噪音劑量研究的錄音系統。此系統包括ㄧ個商用的麥克風元件、一個改裝過的商用耳機、一個改裝過的商用MP3錄音筆及一套商用的聲音分析軟體。此系統在迴響室中進行的純音校正、混合音校正及劑量校正測試結果顯示，MP3錄音系統所錄下的聲音檔案經過Goldwave聲音轉檔軟體及NoiseLab聲音分析軟體處裡後，可以得到誤差在1 dBA範圍以內的噪音量測結果，並且在測試的70~110 dBA音量範圍內有良好的錄音線性反應，在計算誤差上已經初步達到Type1噪音計的規範。本研究利用一套由王老師團隊自製的錄音檔的分析軟體- Spectrogram2，該軟體可以分析每秒鐘的音量變化與頻譜特性，經過在迴響室內進行的噪音計劑量校正測試結果顯示MP3錄音系統所錄下的聲音檔案經過Spectrogram2分析後其每秒音量與市售的Dosimeter紀錄內容沒有顯著的差異(音量誤差小於1 dBA，劑量誤差0.15%)，在計算誤差上初步已經達到Type 1噪音劑量計的規範。 本研究也成功的開發出一套MP3錄音筆及麥克風的選擇規範，未來的研究人員可以依據這套規範選擇合適的MP3錄音筆及麥克風組合，對於從事類似研究，將有很大的方便性。本研究雖然沒有完成自製MP3錄音設備的開發，但是整體研究過程的經驗，仍然有助於提昇未來類似研究的能量，部分成果也能夠加速個人可攜帶式噪音頻譜分析儀的開發。 目前此系統僅在迴響室內測試，未來盡快安排勞工作業環境的實測，以了解系統的穩定性及對各種不同噪音環境的可用性。此外，強化分析軟體的分析能力並減少計算時間，也是未來需要積極強化的研究課題。

Hearing loss caused by occupational injuries ranked first in the number of occupational diseases in Taiwan. In Taiwan, about 26.2 percent of the labor hearing loss ranged from 25 to 40 dBA, and even 9.6% of the labor hearing loss are greater than 40 dBA. To better understand the noise in the ear, especially the high-frequency sound and its effects on the human body, this project developed a noise measuring equipment that can be hung on the ear canal to measure the noise frequency and exposure doses close to the ear canal and to provide information for noise-induced hearing loss related research needs. After much trial and error, this study has successfully converted the commercial production of MP3 into a recording system for the outer ear noise dose study. This system includes a commercial microphone element, a modified set of commercial headphones, a modified commercial MP3 recording pen and a commercial sound analysis software. The test results of this system in an echo chamber for the pure tone correction, mixed tone correction and dose calibration showed that, the sound in the MP3 recording system, after being converted by Goldwave and analyzed by NoiseLab at the first, can obtain the noise measurement results of error within 1dBA, and the test volume ranging between 70 and 110dBA has a good recording linear response, whose calculation error has preliminarily achieved Type 1 noise meter specification. Meanwhile, the research has also successfully developed a set of MP3 audio file analysis software —Spectrum2 — that can analyze volume changes per second and spectral characteristics. The test results in the echo chamber for noise dosimeter calibration showed that no significant difference existed between the volume per second of the sound in the MP3 recording system analyzed by Spectrum2 and that of the sound in Dosimeter (volume error is less than 1 dBA, and dose error is 0.15%), on preliminary calculation error has been reached Type 1 noise dosimeter specifications. The study also succeeded in developing a set of MP3 and microphone selection criteria, based on which future researchers can choose the suitable combination of MP3 and microphone and benefit from the convenience for similar studies. Although this study did not complete the development of self-made MP3 recording device, the overall experience of the research process still helps to improve the energy of the future similar studies, and accelerate the development of individual noise spectrum analyzer. At present, this system is only tested in the echo chamber, but future tests will be done in the labor operating environment as soon as possible in order to understand the system stability and its practicability in a variety of different noise environments. In addition, to strengthen the analytical capacity of analysis software and to reduce the computation time are both future research topics that need to be addressed.