Power Up Your Obsidian Notes with Audio Transcription

Unless you’re literally Socrates, you probably believe that writing stuff down is a good way to communicate information and store knowledge. Whether you’re an ad hoc jotter, napkin scribbler, or a prolific diarist, you’ve probably been taking notes for basically forever. Ephemera like shopping lists, reminders that you’re parked on Floor 7 in the orange section, and brief musings about life easily find a home on the various notes apps that come bundled with our phones and computers, but to build a more durable, personal, networked knowledge base, you’re going to need something a little more specialized. 

Obsidian is a very powerful note taking app that can be used as a "second brain" to keep thoughts, plans, general notes, and other information. It also doubles as an internet publishing platform, allowing folks to share their notes and “digital gardens” with the world.

The information is stored in your local folder structure of plain Markdown files. What makes Obsidian really powerful though, is the ability to add plugins to customize and extend the software to suit your particular needs. There are Core plugins that are provided by Obsidian, and there are lots of community built plugins to add even more functionality

One of the core plugins offered by Obsidian is the Audio Recorder plugin. It allows the user to record audio to take a quick voice note. This can be handy when you aren't able to type everything down in the moment.

These audio notes are helpful, but reviewing large swaths of audio can be  difficult. Finding out what the notes are about or what was actually said is only possible by listening to the audio notes again. If you were doing this yourself, by hand, you might scrub through your audio at 3x speed, waiting to find that one nugget of audio gold. And then what? You’re going to type it down, word by word, pausing every few seconds to keep the phrasing fresh in your head? This, dear reader, is no way to live.

Getting a transcript along with audio notes makes them immediately actionable, and makes them legible to good old-fashioned text search. To accomplish this we added Deepgram transcription to the Audio Notes community plugin by Jason Maldonis.

Setting Up Audio Notes with Deepgram in Obsidian

Assuming you already have Obsidian installed, getting set up with the Audio Notes plugin is pretty straightforward. 

Step 1: Open Obsidian Plugins inside of Settings

In the settings, you'll find a tab called "Plugins" on the left-hand side of the screen. Click on it to open the plugin options.

That was easy, right? Next, we’re going to install the plugin we’re going to use.

Step 2: Search for “Audio Notes” by Jason Maldonis plugin in Community Plugins

In the "Community Plugins" tab, search for "Audio Notes" by Jason Maldonis in the search bar. Click on the "Install" button to add the plugin to your Obsidian vault.

Now that we have the plugin installed, we’re going to get a Deepgram API Key in the next step.

Step 3: Go to Deepgram to sign up for an account and to get an API Key

To use the Audio Notes plugin, you need to sign up for an account with Deepgram and get an API Key. Go to the Deepgram website and follow the instructions to sign up and get your API Key.

The next step ought to be exciting: We’re putting the key in the ignition. It’s almost off to the races.

Step 4: Add API Key into the “Deepgram API Key” setting in the Audio Notes plugin settings

Back in Obsidian, go to the "Audio Notes" plugin settings and paste your API Key into the "Deepgram API Key" field. Once you've done that, you're all set to start using the Audio Notes plugin.

And there you have it! You should be all set up to use Deepgram to transcribe your notes into searchable, linkable text.

Using the Audio Notes Plugin with Deepgram

Using the Audio Notes plugin is just as simple. Here’s how it’s done.

Step 1: Activate Audio Notes

With a note open, you can either click the Microphone button in the side ribbon, or you can use the “Audio Notes: Generate quick audio recording with transcription” command.

Up next: Settings.

Step 2: Configure Your Settings

The Audio Note modal is now open and you can select your desired options for the audio note.

Here are the Settings you should be able to modify:

• Folder to save audio files to

• Language the recording will be in

• Model Tier (Enhanced is more accurate)

• Punctuation

• Numbers (Converts “twenty twenty three” to “2023”)

• Profanity Filter

• Keywords

  • You can specify keywords to which the model should pay particular attention to help it understand context; this is known as keyword boosting. Similarly, you can suppress keywords using a positive number.

• Title of audio note

Here's what that looks like.

Next, we’re going to go over how to use the plugin to transcribe your notes into Obsidian.

Step 3: Click the microphone button to start recording

Step 3a: Click the pause button to pause recording if needed. Then click the microphone button to resume recording.

Step 4: Click the Stop Recording button 

This step also saves the audio and triggers the API call to get the transcription from Deepgram

Step 5: Click the “save” button 

This will add the audio note with transcription to the current note.

For some use cases, you could probably stop here, but you'll probably want to edit (and where necessary, correct) your transcript.

Step 6: Review and edit the transcription 

To review and edit your new transcription, click the “</>” button in the top right corner to edit the text if needed. 

The plugin is still in beta and we plan to continuously improve it. If you have suggestions for how we can improve it, or features you'd like to see, let us know in the GitHub repo listed in the plugin.