Dan’s Substack
The Story Machine
My NotebookLM Workflow
0:00
Current time: 0:00 / Total time: -3:49
-3:49

My NotebookLM Workflow

How I made my first Podcast with NotebookLM

Using Google NotebookLM

A generalized workflow for using Google NotebookLM

Generating a basic podcast

Although NotebookLM wasn’t originally intended to be a podcast creation tool, it’s ability to do just that is what made it a very popular app. The features that were overshadowed by the podcast are FAQ, Summary, Timeline, etc. These less known abilities serve useful purposes, but the incredible ability to create an audio program of two people discussing any topic the user presents them with is the icing on the cake.

A podcast, produced by NotebookLM comes out of the mold, so-to-speak as an element that can be played and listened to. In addition the audio can be downloaded as a high quality, mono-track .wav file.

Content Production

Content of various kinds can be used with NotebookLM. Large bodies of multimedia content such as websites and YouTube videos can often be imported directly. However, content may be restricted by DRM. It is up to the individual to decide whether using it is ethical. Scraping content from the internet is a widespread activity and there are many ways to do it. I’ve tried it experimentally, but have never published work derived this way.

In most cases, I use written text with permission of the writer. Content from Google Docs, Google Slides, or copied text from any source can be added using the buttons under “Add Sources”. NotebookLM can accommodate a very large amount of source content.

Once the source content is imported, there is one more step to consider before hitting the generate podcast button. That is, the options. The options feature is actually a way to include a prompt, up to 500 words, that is used to give the app specific instructions about how to generate the podcast.

So far, I’ve used the options prompt a few times, but I’m sure I’ve only scratched the surface of its abilities. Here’s an example of one prompt that worked pretty well:

Maryann Pelland brings us another useful article for writers in her series ”Pen to Profit” published on Substack.

At the beginning of the podcast, refer to the author as Maryann Pelland. For the body of the podcast refer to her as Maryann.

In the first sentence, I told the app that the article is part of a series of related articles. In the second sentence, I told it to mention the author’s name at the beginning of the podcast, then as Maryann throughout the remainder.

That’s it for the basic guidelines. Next we look at how to apply the finishing touches to a complete production ready podcast.

Audio Production

Download the podcast file

Import the podcast file into Audacity

Remove unwanted segments, if any

The following three steps are optional

Create another mono track and add an intro, postscript, or other additional elements

Create another mono track and add background music

Apply audio adjustments such as EQ, compression, fades, etc

Discussion about this podcast

Dan’s Substack
The Story Machine
Artificial intelligence, the written word, and story telling
Listen on
Substack App
RSS Feed
Appears in episode
DW Pelland