Closed Cmdv closed 4 years ago
This one is tricky. Current classification uses SOX, which doesn't give the classified timestamps of each "sentence", but merely outputs a lot of audio files for each of them. I've been contemplating writing a custom classifier for audio but I've not had the time.
Yeah after digging in the code I spotted that, seems like quite the task!
Indeed. :( Could possibly use Sox as a library and retrieve the timestamps somehow, but haven't looked into that.
Den ons 23 jan. 2019 14:28 skrev Vincent Orr notifications@github.com:
Yeah after digging in the code I spotted that, seems like quite the task!
— You are receiving this because you commented. Reply to this email directly, view it on GitHub https://github.com/owickstrom/komposition/issues/80#issuecomment-456821042, or mute the thread https://github.com/notifications/unsubscribe-auth/ABZYCIDsNBoA16Kk0IgSaxABFoet2qSFks5vGHF6gaJpZM4aJKj0 .
Closing this for now. Thanks!
no worries need to revisit this project, it was fun to work on last time 😃
Me, too. I've been quite inactive the past 6 months.
When Audio clip is selected add a
Start/End
to the sidebar menu to be able to edit the start and end of an audio clip much like you currently can with video.