fantastic_futures_2024_whisper

Tools for the fantastic futures 2024 whisper presentation

Tools/Steps:

load_content

Grabs all of the content for the selected items from HCP/SCP listed in the spreadsheets

normalize_content_media

Strip the trailing silence from the end of the media files. The algorithm boils down to:

• Find all of the silence segments
• starting at the end and working backward look at each silence segment:
  • Get the length of the non-silence duration between the end of the silence segment and the computed duration
  • If the length is less than the duration of the silence segment, then it's inconsequential noise and reset the computed duration to the start of the silence segment (i.e. shorten the file)
  • If it isn't, then it's real audio and we stop recomputing the duration
• cut the file at the computed duration and replace the original file.
• The metadata