Creative Commons License
This work is licensed under a Creative Commons Attribution-Noncommercial-No Derivative Works 4.0 License.
Date of Award
Master of Arts (MA)
Department of Graduate Psychology
Speech sounds can be modeled as a product of two components: the source and the filter. The effects of filter manipulations on speech perception have been studied extensively, while the effects of source manipulations have been largely overlooked. This study was an attempt to assess the impact of source manipulations on vowel identification. To this end, two source manipulations were conducted prior to filtering. First, several harmonics of the source sawtooth wave that were located near formant peaks were mistuned, either towards or away from the peaks. Mistuning towards formant peaks was expected to facilitate vowel identification by helping to convey the position of the formant more clearly; mistuning away was expected to hinder performance. Consistent with this hypothesis, a significant effect of mistuning was observed. However, follow up analyses revealed that this manipulation only had an effect in conditions where harmonics are mistuned away from formant peaks by a large degree (5%). The second manipulation consisted of adding noise to the source signal to “fill in” the acoustic spectrum. Because the addition of noise occurred before filtering, the spectral shape of the noise component was identical to that of the harmonic portion of the tone, and was expected to help convey formant peaks, especially when they were not well conveyed by harmonic information. The results reveal that the addition of noise had no effect on vowel identification. Possible stimulus based explanations for the failure to observe some of the hypothesized effects are discussed.
Becker, Christopher James, "Interaction between source and filter in vowel identification" (2012). Masters Theses. 148.