The sound spectrograph, as a wave analyzer, produces a permanent visual record showing the distribution of energy in both frequency and times. Since early-nineteenth century, the production of sound spectrograph, as originally proposed by Potter(1934), made the voice analysis to develop rapidly and many authors, such as Fischer(1958), Peterson and Barney(1952), Kamamodo(1956), Fant(1960), and Dalston(1975) etc, had studied the physicoacoustic characteristics of the vowels and consonants respectively and many acoustical cues had been verified in their own languages. Thus, internationally, there has been much scientific research in this field, but in Korea we have still remained unexplored. So I present this paper of research on the medico-sonagraphic analysis of normal Korean vowels and consonants and hope this results to be aid for correct diagnosis and treatment of speech disorder as well as basic datas for the speech word listing in the clinic.

Experimental Subjects and Methods : As experimental subjects, the author selected each 20 healthy Seoul Dialect speaking men and women(21~30 old age) who were born in Seoul, via the following preliminary tests ; 1. History taking-No past and present illness of voice production organs. 2. Physical examination 1) Normal structure of nasal cavity and paranasal sinuses. 2) Normal structure of oral cavity and pharynx. 3) Normal structure of larynx and chest. 4) Normal structure of hearing organs. 3. Stroboscopy with Timcke stroboscope(Type KS 3)-Symmetrical amplitude, complete closure, and normal wave-movement of vocal cords. 4. Vocal function test with Collins Respirometer(Type 9L) 1) Normal mean flow rate. 2) Normal maximum phonation volume. 3) Normal maximum, phonation time. 4) Normal vocal velocity index. The author used the following instruments as experimental materials which have set up at Vocal Dynamics Laboratory in Otolaryngological Department of Yonsei University College of Medicine. 1. KAY Sonagraph-Type 6061B, 2. Ampex Tape Recorder-Type AG300-2, 3. Miniform Contact Microphone-Type ITT 106, 4. Sound Proof Recording Booth-Type IAC. 40-1A. With the above subjects and instruments, the author has performed the medico-sonagraphic analysis of normal Korean vowels and constants and the following results were obtained. 

Results and Conclusions : 1. In front vowels(ㅣ, ㅔ, ㅐ) the 2nd formant is distributed far apart from the 1st formant and close to the 3rd formant. On the other hand, in back vowels(ㅜ, ㅗ, ㅏ), the 2nd formant is distributed close to the 1st formant and far apart from 3rd formant and in other vowels(ㅚ, ㅡ, ㅓ), formants are approximately equidistant from each other ones. 2. The widest band-width of the formants of vowels is observed in the 1st formant and the strongest intensity of formants of vowels also observed in the 1st formant. On one hand, front vowels have more numbers of formants than other vowels. 3. In the high pitched sounds, the formants are distributed in higher frequency especially 1st, 2nd and 3rd formants than in the low pitched sounds and many numbers of the formants are observed in the high pitch sounds, probably due to the branching of the formants. 4. In disyllable words(VCV pattern), prepositioned vowels have longer phonation time than postpositioned vowels and the intrinsic phonation time is more longer in the low vowels than in the high ones. 5. The 2nd formant of the diphthongs shows the remarkable formant transition and in CV monosyllable diphthongs, the formant transition is indistinguishable or disturbed. 6. The plosive consonants are marked by the closure(the period of silence), expolsion, and noise-like waves just before the 1st formant of vowels. On the other hand, the aspirated plosive consonants show the noise-like waves just before the each formants of vowels, followed by the closure and expolsion waves. 7. The fricative consonants show the high frequency noise-like waves and ㅎ is marked by low frequency noise-like waves and the noise-like waves just before the each formants of vowels. 8. The africative consonants show the mixed characteristics of the plosive and fricative consonants. 9. The nasal and lateral consonants are marked by the formant-like low frequency energy waves. 10. In disyllable words(VCV pattern), the postpositioned high vowels make the phonation time of consonants to be longer than the low vowels, with exceptions of bilabial and laryngeal consonants. On the other hands, the postpositioned low vowels make the phonation time of consonants to be shorter than the high vowels. In summing up the above results, the author thinks the important acoustic cues of simple vowels and diphthongs are the relative frequency distribution and the formant transitions of the 1st and 2nd formants, respectively, and it is probably due to the time duration of the lingual movement that the low vowels show longer phonation time than the high vowels. On one hand, the remarkable acoustic cues are the low frequency noise-like waves just before the 1st formant in the plosive consonant, the high frequency noise-like waves in the fricative consonants, the high and low frequency noise-like waves in the africative consonants, the formant-like low frequency anergy waves in nasal and lateral consonants. It is probably due to no involvement of the lingual movement that, in the biliabial and laryngeal consonants, the phonation time is not affected by the kinds of prepositioned or postpositioned vowels.