[Feature] Upload video/audio + reference transcript → Get subtitles/lyrics with timecodes

[porg]

Use Case

• You upload audio/video plus a plain text file containing a perfect or near perfect transcript without timecodes.
• whishper's added value in this use case:
  • It creates the necessary timecodes for me.
  • Possibly with very high precision, see proposal #82
• All the other benefits of whishper in conjunction
  • Editor UI
  • AI Translation services
  • Self Hosting / Local Processing

Best Practises

• Noraebang for Everyone by Korean company Gaudi Lab allows that exact use case.
• But with the downsides:
  • No possibility to download/export lyrics/subtitle files
  • Server side processing
  • Beta only, later will become a SaaS I assume.