• 대한후두음성언어의학회지
• /
• 제18권1호
• /
• pp.33-38
• /
• 2007

Background and Objectives: It has been suggested that tonsillectomy possibly causes changes of voice because the morphology of the vocal tract is altered. This may cause serious problems for professional voice users. Materials and Method: Subjects were 26 patients. The oral and nasal sound spectrum of oral vowel /a/, /e/ and /i/ were measured before and after tonsillectomy. The formant frequencies and intensities for oral and nasal spectra were compared. The nasality and fundamental frequencies for oral vowel were measured. Results: The first formant frequencies for oral spectra of all vowels were not changed after surgery, but the second formant frequencies were increased significantly after surgery in the vowel /e/ and /i/. The first and second formant intensities for oral spectra were increased significantly after surgery in the all vowels. The first and second formant frequencies for nasal spectra of all vowels were not changed after surgery, but their intensities for nasal spectra were increased after surgery. The nasalities for oral vowel were not changed after surgery. Conclusion : Tonsillectomy appeared to change the spectral features of oral and nasal components of oral vowel, especially spectral intensities.

• 음성과학
• /
• 제3권
• /
• pp.83-92
• /
• 1998

Vowels are classified by vocal tract shapes. These shapes form constriction points along the tract, which have an influence on such vocal tract resonance as $F_l,\;F_2,\;F_3$, and so on. This study reviews the perturbation theory of the tract and determines the corresponding formant frequencies from modified vocal tracts using vocal tract area function. Then, formant variation is observed from the theory. Finally, each set of $F_l,\;F_2,\;and\;F_3$ frequency is input to a speech synthesis software to make a vowel sound. Auditory impression of each sound without any modification of its vocal tract shape is almost the same as the corresponding phonetic symbol. Formant frequencies of $F_l,\;F_2,\;F_3$ vary according to the perturbation theory. Generally, constriction along the node causes formant values to decrease while constriction along the anti-node cause it to increase. Vocal tracts modified by more than $3\;cm^2$ change vowel qualities of /a/ and /i/ into those of f /v/ and /$\varepsilon$/, respectively. This study will be helpful in simulating sounds from modified vocal tracts before any operation. Further studies are desirable to compare vocal tract shapes of various languages and their sounds together.

• 음성과학
• /
• 제13권3호
• /
• pp.67-82
• /
• 2006

The purpose of this experimental study is to investigate formant characteristics of Korean monophthongs spoken by Japanese learners and to compare the characteristics of vowels produced by the Japanese learners with those of the Korean native speakers. The data consisted of three categories: eight vowels in isolation, words including eight vowels in carrier sentences, and words including eight vowels in natural sentences. In this study, formant frequencies of the vowels were measured by Wave Surfer. It was assumed that the formant frequencies of the Korean vowels produced by the Japanese learners could be different from those of the Korean native speakers due to the influence of their own Japanese vowels. Results of this study showed that the Japanese learners had the difficulties to distinguish between the pairs /-/and /ㅜ/, /ㅓ/and /ㅗ/, and /ㅏ/and /ㅔ/. In Japanese vowels, F2 frequency value of /ㅜ/ was similar to that of the Korean /-/. It means that when the Japanese leaners produced Korean /ㅜ/, they might neutralize /-/ and /ㅜ/. Besides, there were not /ㅓ/and /ㅐ/ in Japanese vowels. Therefore, they tended to pronounce /ㅓ/ similar to /ㅗ/ which has the most similar formant frequency value with that of /ㅓ/, and /ㅐ/ was pronounced similar to /ㅔ/ for the same reason.

• 대한전자공학회논문지
• /
• 제25권4호
• /
• pp.356-365
• /
• 1988

In this paper, we describe an isolated words recognition system for Korean isolated digits based on a voiced -unvoiced decision algorithm and a frequency domain analysis. The algorithm first performs a voiced-unvoiced decision procedure for the begtinning part of each uttered work using the normalized log energy and zero crossing rate as decision parameters. Based on this decision,. each word is assigned to one of two classes. In order to identify the uttered word within each class, a dynamic time warping algorithm is applied using formant frequencies as the basis for the distance measure. We exploit a pole-zero analysis to measure formant frequencies in each frame. We have observed that pole-zero analysis can provide more accurate estimation of formant frequencies than analysis based on poles only. Experimental recognition rates of 97.3% illustrating the performance of the recognition system was achieved.

• 음성과학
• /
• 제9권2호
• /
• pp.1-11
• /
• 2002

This paper aims to provide a guideline for modelling speaker identification and speaker verification by comparing voice similarities between brothers. Five pairs of brothers who are believed to have similar voices participated in this experiment. Before conducted in the experiment, perceptual tests were measured if the voices were similar between brothers. The words were measured in both isolation and context, and the subjects were asked to read five times with about three seconds of interval between readings. Recordings were made at natural speed in a quiet room. The data were analyzed in pitch and formant frequencies using CSL (Computerized Speech Lab), PCQuirer and MDVP (Multi -dimensional Voice Program). It was found that data of the initial vowels are much more similar and homogeneous than those of vowels in other position. The acoustic data showed that voice similarities are strikingly high in both pitch and formant frequencies. It was also found that the correlation coefficient was not significant between parameters above.

• 대한전자공학회논문지
• /
• 제17권3호
• /
• pp.12-19
• /
• 1980

A speech recognition system for separately pronounced Korean digits is described. The system is composed of four stages ; parameter extraction, segmentation by voiced-unovied analysis, formant tracking and pattern matching. Digit speech is segmented into an unvoiced segment and/or a voiced one using ZCR and energy measurements, then to estimate the first three formant frequencies a relatively simple formant tracking scheme is applied to the raw formant data extracted from linear prediction spectra. Finally, pattern matching is made using dynamic programmig method. Recognition experiment is carried out for 150 digit utterences spoken by three male speakers, and recgnition rate 94 % is obtained.

• 말소리와 음성과학
• /
• 제11권4호
• /
• pp.97-108
• /
• 2019

한국인 성인을 대상으로 최적의 포먼트 분석이 가능하도록 자동화된 프랏 스크립트를 구현하였다. 최적의 포먼트 분석이란 프랏에서 포먼트 분석 시 설정하는 2가지 세팅 파라미터(최대 포먼트, 포먼트 개수)를 조합하여 측정된 제1, 제2 포먼트의 편차합이 최소일 때를 가리킨다. 포먼트 분석의 신뢰성을 높이기 위해서는 성별이나 모음의 종류에 따라 LPC 차수를 다르게 설정해야 하는데 프랏 매뉴얼에서는 최대 포먼트 설정 값으로 남성 5,000 Hz, 여성 5,500 Hz, 측정개수는 5개를 권고한다. 그러나 이렇게 권고된 포먼트 세팅 설정이 한국어 모음에 대해서도 타당한지 검증이 필요하다. 본 연구에서 구현한 4가지 스크립트를 적용한 결과, 각 모음별 포먼트 산점도로 확인하였을 때 특히 여성의 경우 스크립트에 따라 측정된 포먼트 변이의 폭이 두드러지는 차이를 보였다. 포먼트 산점도와 통계 결과를 통해 linear_script와 qtone_script가 포먼트 측정에서 더 신뢰성이 높은 것을 알 수 있었다. Linear_script, qtone_script에서 최적의 세팅으로 설정된 최대 포먼트와 포먼트 개수의 데이터 경향성을 살펴보면, 전설 모음 [이, 에]의 경우 권고 설정보다 최대 포먼트 값은 높게, 포먼트 개수의 값은 적게 설정되었다. 반면 후설모음 [오, 우]의 경우, 권고 설정보다 최대 포먼트 값은 낮게, 포먼트 개수의 값은 많게 설정되는 것을 확인할 수 있었다.

• 대한음성학회:학술대회논문집
• /
• 대한음성학회 2003년도 5월 학술대회지
• /
• pp.141-144
• /
• 2003

In this paper, we compared formant frequency extraction algorithms with various conditions, and show their performances. The formant frequency is the resonance frequency which is decided by the vocal tract characteristics. It is related with phonemes, or characteristics of the physical condition of the vocal track. Since the speech signal is influenced by both the sound source and the vocal tract, it is difficult to calculate the exact formant frequencies. Many studies on the formant frequency extraction had been executed already Besides, any new formant frequency extraction algorithm is hardly found recently.

• 음성과학
• /
• 제8권4호
• /
• pp.221-228
• /
• 2001

The purpose of this study is to analyze the problems Korean students, having acoustic structure of Korean vowels, have when they pronounce English vowels by measuring formant frequencies. The experimental results show that the pronunciation of English vowels by Korean students is partially influenced by their Korean vowels. There is little distinction between /i/ and /I/, /U/ and /u/ due to the absence of short and long vowels in Korean pronunciation. Also, as observed in typical Korean vowel pronunciation, there is little difference between the F1 values of /$\varepsilon$/ and /$\{\ae}$/ by Korean speakers, resulting in inaccurate English pronunciation. In addition, compared to English native speakers, Korean speakers show the biggest difference in F1 value of /c/. The fact that they make pronunciation of /c/ covering /e/, /$\Lambda$/ and /c/ positions probably accounts for such phenomenon. The results of this experiment show the interference of Korean that occurred in some English vowels by native Korean speakers.

• 대한전자공학회:학술대회논문집
• /
• 대한전자공학회 2002년도 ITC-CSCC -3
• /
• pp.1799-1802
• /
• 2002

This paper proposes an improved method for formant frequencies estimation based on the complex autocorrelation function of the speech signal. Instead of using the incoming signal as an input fur the LPC analysis, the analytic signal of the autocorrelation function of the speech signal is computed and itself used as an input for the LPC analysis. Due to the properties of the analytic signal, which occupies half of the bandwidth of the original signal, the required model order for the LPC analysis is halved. The accuracy of the proposed method in noisy environments is examined on five natural vowels. The effectiveness of the proposed method is shown by the estimated spectral shapes and the estimation errors of the formant frequencies.