The Waveform Model of Vowel Perception and Production

The Waveform Model of Vowels is the key to several doors. The organization of the vowel space can be seen in one table. Use this as your introduction to the first working cognitive model that uses this organization to achieve human performance. The WMV PDF

Waveform Communication, LLC

The mission of Waveform Communication, LLC is to promote and develop serviceable applications developed from the Waveform Model of Vowel Perception and Production. The objectives include, but are not limited to identifying concussions from speech productions, speech recognition algorithms, improving speech in noise by 8-10%, talker identification, hearing related tools, audio applications, and the expansion of communication theory.

Medically impaired speech: Concussions (the beginning)

In collaboration, Waveform Communication and Methodist Sports Medicine have collected speech from 45 talkers having concussions. Each subject was recorded on the initial visit they were clinically diagnosed with a concussion, and recorded at each additional office visit until released from care with at least 2 recording sessions and as many as 8. Over 3,800 wav files were recorded across all talkers, and each production was acoustically analyzed every 6 milliseconds providing a dataset of thousands of rows of acoustic data which is unparalleled in detail (healthy or impaired).

Cobweb Sports and Contact Trauma Analytics, LLC

Cobweb employs proven, proprietary waveform speech recognition technology that is being designed to provide an accurate and immediate diagnosis of a concussion. The technology of Cobweb is based on the Waveform Model of Vowels and proven human performance.


CobwebTM development reaches milestone
  • The Waveform Sports Medicine speech database has been created
  • The speech of 45 concussions patients was collected through recovery
  • This is the largest database of impaired speech ever collected
  • Over 3,800 wav files have been collected and acoustically processed
  • Each wav file was processed every 6 milliseconds
  • No other dataset is so detailed (healthy or impaired)


The Waveform Model of Vowels
  • Over 99 percent achieved on multiple data sets (see Blog)
  • Explanation of Vowel Production
  • Explanation of Vowel Perceptual Errors
  • Over 95% accuracy on streaming speech across talkers


Acoustic analysis tool providing real time processing of vowels.

Things to Consider

"It is not possible to look at the waveform of an utterance and say what sounds occurred", (Ladefoged, 1982, A Course in Phonetics, p. 168).
The impossible is now possible.
From speech signal to phonological features - A long way (60 years and counting). Henning Reetz, presented at the 164th Acoustical Society of America meeting in October, 2012.
There should not be a paper discussing 70 years and counting.