Auto-transcribe attached audio

Use case or problem

I want to be able to create a note, record audio of me talking, and then have it auto-transcribe into markdown text that I can edit from there.

Proposed solution

If you insert an audio file, it transcribes it

2 Likes

For now, you could check out Otter, which would do all you need except for create a note in your vault.

1 Like

I simply use my Android phone for that … talk directly into Markor (a Markdown editor for Android). The files get auto-synced into my +Inbox folder and I can work from there on with Obsidian on my laptop or desktop.

2 Likes

You can try this speech to text converter and then convert your text with any md converter. This software has an in-built editor and supports many audio formats. Hope it helps.

Hey there! I use Maestra, an automatic speech to text platform.

There are several pretty good tools out there. I’ve had pretty good luck w/ hyperia for transcribing notes. its pretty cool – labels speakers if several ppl are talking and captures + OCRs screen data. otter is also pretty good if you just need regular transcription (no screen stuff). i know the mozilla project is also working on some open source speech stuff, which is great

If I need to transcribe audio files to text I usually use professional transcription services by Professional Transcription - Audext. It also has many helpful features.

The problem I have with these often is

  • they transcribe only your own audio recordings, and only once you put them in a note.
  • they deliver a full transcription (more than I mostly need, it clutters my notes with unnecessary text).

I want to transcribe on-the-go when listening to podcasts/clips, and then only what interests me (the last x seconds).

On Android you can use Momento for Podcast clip-to-text transcriptions.

It will transcribe the last 20 seconds right away by pressing a button. The transcription can then be shared with the Obsidian app to create a note from it.
As a Podcast player it’s a bit restricted I find, but it has a Spotify interface. Also one for readwise.io.

2 Likes

I can recommend you a really good service called Strok.ai https://stork.ai/

Or talk directly into the Obsidian app via text to speech from the keyboard…