Research optimizing media file format for transcriptions #33
Labels
spike
Research for features with lots of unknowns
voice note
Issue directly related to Voice Note feature
User Story
As a voice note user, I may want the fastest voice note transcription possible instead of a slower one with more confidence.
Overview
Research which, if any, media file formats affect transcription job times in Amazon Transcribe. Currently, WAV with PCM 16-bit encoding is a larger file format than most and may be overkill for simple notes. Ultimately, this could result in toggle between faster, lower quality notes and slower, higher quality ones. If not prior research exists, we can implement 2-3 other file formats (research length of time to do so) and measure/benchmark the transcription times in a repeatable environment/way.
The text was updated successfully, but these errors were encountered: