An Evaluation of the Accuracy of MTF-STI Measurements by Comparison to Japanese Three-Syllable Listening Tests

Mochimaru A. An Evaluation of the Accuracy of MTF-STI Measurements by Comparison to Japanese Three-Syllable Listening Tests. In: AES Convention 89. Audio Engineering Society; 1990. Paper 2941. Available from: https://aes.org/publications/elibrary-page/?id=5752

@inproceedings{Mochimaru1990_5752,
  author = {Mochimaru, Akira},
  title = {{An Evaluation of the Accuracy of MTF-STI Measurements by Comparison to Japanese Three-Syllable Listening Tests}},
  booktitle = {AES Convention 89},
  note = {Paper 2941},
  year = {1990},
  month = sep,
  publisher = {Audio Engineering Society},
  url = {https://aes.org/publications/elibrary-page/?id=5752}
}

TY  - CPAPER
TI  - An Evaluation of the Accuracy of MTF-STI Measurements by Comparison to Japanese Three-Syllable Listening Tests
AU  - Mochimaru, Akira
T2  - AES Convention 89
M1  - Paper 2941
PY  - 1990
DA  - 1990/09/06
UR  - https://aes.org/publications/elibrary-page/?id=5752
PB  - Audio Engineering Society
LA  - en
AB  - Acceptable speech intelligibility should be a fundamental and important target for every sound system designer. Speech intelligibility is affected not only by the ratio of early energy to reverberant energy, but also by the properties of each individual discrete sound arrival. Evaluation of subjective listening tests using three-syllable word lists shows clearly the effects of loudspeaker directivity, listener location in the room and the acoustic properties of the room. The comparison of the results of listening tests and measured STI values is illustrated using a best-fit logarithmic regression curve. STI values are calculated from a computer generated impulse response. Listening tests are made using a DSP implementation of the same impulse response. The comparison reveals a relationship between the impulse response shape and the STI values. These results show that speech intelligibility can be accurately predicted by MTF-STI measurements because both the time and the energy characteristics of each sound arrival are considered.
ER  -