[originaltext] If we want to measure voice features very accurately, we can

If we want to measure voice features very accurately, we can use a voice analyzer. A voice analyzer can show four characteristics of a speaker’ s voice. No two speakers’ voices are alike. To get a voice sample, you have to speak into the voice analyzer. The voice analyzer is connected to a computer. From just a few sentences of normal speech, the computer can show four types of information about your voice. It will show nasalization, loudness, frequency and length of articulation. The first element, nasalization, refers to how much air normally goes through your nose when you talk. The second feature of voice difference is loudness. Loudness is measured in decibels. The number of decibels in speaking is determined by the force of air that comes from the lungs. The third feature of voice variation is frequency. By frequency we mean the highness or lowness of sounds. The frequency of sound waves is measured in cycles per second. Each sound of a language will produce a different frequency. The final point of voice analysis concerns the length of articulation for each sound. This time length is measured in small fractions for each second. From all four of these voice features—length of articulation, frequency, loudness and nasalization—the voice analyzer can give an exact picture of a person’ s voice.
Questions 16 to 18 are based on the passage you have just heard.
16. What is the voice analyzer connected to?
17. What does the first element nasalization refer to?
18. How can the frequency of sound waves be measured?