LAB 10: Sound Segment Reduction

Labs: 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 10 | 11 | 12 | 13 | 14 | 15 | 16 | 17

The main purpose of this laboratory is to study the modification of sounds in context using acoustic analysis. A perception test is prepared and carried out during the next lecture.

Here are the resources you will need:

Lab Handout (PDF)

Utterances

  • Spontaneous speech that was recorded in Lab 1.
  • Recordings of clear renditions of modified words, to be made during the lab.

If you did not record a database in Lab 1, or if you do not have access to the tools for recording the new utterances, you can download examples of the utterances for this lab.

The Tools

  • record - the recording software (VAX)
  • maketape - to make listening tests on audiotape (VAX)
  • lspecto - used to make spectrograms (UNIX®/Linux®)
  • xkl - used for spectrum and waveform analysis (UNIX®/Linux®)

Reference Reading

Oshika, B. T., V. W. Zue, R. V. Weeks, H. Neu, and J. Aurbach. "The Role of Phonological Rules in Speech Understanding Research." IEEE Transactions on Acoustics, Speech, and Signal Processing. ASSP-23, no. 1 (1975): 104-112.

Manuel, S. Y., S. Shattuck-Hufnagel, M. Huffman, K. N. Stevens, R. Carlson, and S. Hunnicutt. "Studies of Vowel and Consonant Reduction." Proc International Conference on Spoken Language Processing. Vol. 2. 1992, pp. 943-946.