Early access to version 0.18.3 standalone

Last week we pushed version 0.18.3 on the private repo, so here come the links for the standalone version too. To see a full list of the updates, see the CHANGELOG from the public GitHub page.

More export options and time codes!

We've added the ability to export to Avid DS files and into Blackmagic Fusion Text+ nodes and more flexibility to use time codes when dealing with transcripts or transcriptions.

Group Questions

This is another feature we've added which basically lets AI detect and group questions from transcripts - this is available both in the Transcription Settings window before transcribing and in the Transcription window  - check it out and let us know how it works for you!

This feature might be quite useful for anyone who has to quickly scan through lines and lines of interviews where there's a person asking questions. The cool thing about it is that it's not only using question marks or symbols for question detection, but it's also able to figure out if the intention of the speaker is to ask something. 

Anyways, this is still AI and an early implementation, so do expect some false positives or negatives along the way. Also, keep in mind that it will take a bit to process the transcript and also that it takes longer the first time you use the feature since it will download a classification model from the Internet.

We're excited about adding this feature since it helped us figure out how content classification works using models that can be kept on your local machine. This means that we'll soon add more features and more flexibility for you to ask AI to label and classify your transcripts based on topics and so on.

Download links

For macOS arm64 (M1, M2 etc.) use this top-secret link.
For macOS Intel use this top-secret link.
For Windows (CUDA only) use this other top-secret link.

Remember, you're among the first to try these out outside our editing room, so please let us know if something doesn't work, preferably via the GitHub issues page.

Have fun, thanks again for supporting this work
and don't forget to spread the word!