In module 2, we’ll look at specific patterns relating to consonants and vowels, and apply these patterns to the task of segmenting, annotating and extracting measures from various types of speech sounds. As you start to recognize the patterns of speech acoustics, keep thinking about the link between what you see in a visualisation of speech acoustics (e.g., a spectrogram) and what is going on with the physical articulators when people speak. If you’re starting to be able to recognize acoustics of specific consonants and vowels, start thinking about how you might automated that process. What sort of phonetic transcription would be helpful for this?
This week you should try to watch the videos (in the ‘Videos’ tab for this module) before the lecture on Thursday. You can bring your questions to the lecture, or post them on the speech.zone forum.
Lecture slides
Lecture 2 slides (google slides) [updated 24/9/2024]