Vox Humana
Automated Sociophonetic Analysis · BYU Linguistics
Upload
Drop audio here or click to browse
WAV · MP3 · FLAC · M4A  ·  Max 1 GB
Advanced Whisper options

When checked, each chunk is aware of what came before, which may improve performance on interviews. Uncheck for wordlists or short files. Uncheck also if you've tried using Whisper on this file already and the transcript came back with strange repeating loops.

Runs an extra MFA refinement pass after the initial alignment. This improves accuracy, but is slower.

Advanced options

When unchecked, vowels that occur while another speaker is talking are excluded from measurement.

If a vowel is followed by an "L" or "R" within the same word, the two intervals are combined. This, in the case of the word pull, "P + UH1 + L" becomes "P + UHL1".