Video Summary:
The video tutorial provides an in-depth guide on documenting a workflow using NotebookLM and Gemini, specifically focusing on transforming a screen-recorded video into a structured slide presentation.
The initial step involves creating a screen recording, which serves as the primary source for the documentation.
It is recommended that the recording be in MP4 format, although it does not need to be perfect; the key requirement is that all actions in the video are clearly synchronized with the accompanying audio narration.
To aid in the documentation process, the MP4 video is converted into a series of PNG images.
This conversion can be efficiently performed using ChatGPT, which can generate the PNG files sequentially from the video.
Users are instructed to request a zip file containing these images, which can then be downloaded for subsequent use.
Following the image extraction, a video transcript is necessary. If users lack a method to generate this transcript, they can utilize NotebookLM to create one.
By uploading the video to NotebookLM, the platform automatically generates an audio transcript, which is essential for providing descriptions in the later stages of the process.
Users are advised to refresh the notebook and upload the PNG frames as sources for the slide presentation.
Once the PNG frames are uploaded, users select these images along with a slide deck prompt.
This prompt requests the generation of presenter slides and a detailed step-by-step process that aligns with the uploaded images and audio transcript.
It is suggested to use the default output length and to test the prompt to ensure it meets the user’s requirements.
After initiating the generation process, the slide deck is created, incorporating the uploaded screenshots and the audio transcript.
The next step involves using Gemini to convert the slide presentation into a visual format that is compatible with Google Slides.
Users activate the canvas feature in Gemini and upload the PDF slide deck.
By pressing enter, Gemini processes the PDF and produces a slide deck in Google Slides, which can then be exported and edited as needed.
An alternative method is also provided, where users can attach the NotebookLM notebook to the Gemini conversation.
This notebook contains the original transcript and frames, enabling Gemini to create a step-by-step guide for Google Slides without needing to reference the slide deck or transcript again.
Users are instructed to activate the canvas and request a Google Slide deck, with the understanding that this process may require some time.
The tutorial highlights the significance of the number of PNG frames generated from the video, as a greater quantity results in a more comprehensive slide deck.
If users face challenges in obtaining an adequate number of frames, they are encouraged to use the free VLC player, which can create an unlimited number of PNG files.
However, users are cautioned to ensure the safety of the software to avoid potential spyware issues.
The entire workflow is designed to be efficient and streamlined, culminating in a final slide presentation that accurately reflects the documented workflow.
Upon completion of the process, users have the option to delete the single-purpose notebook from their account, helping to maintain an organized workspace.
The tutorial concludes by inviting users to explore additional videos for further learning opportunities.
Video:
Video Transcript:
Hello and welcome. In this video, we’re going to document a process inside of NotebookLM. You’re going to start the process with a screen recorded video.
You’re looking at a screen recorded video now that’s playing and if you have an MP4 where you can generate a transcript from that MP4, that’s what you’re going to use as the basis for your documented process.
Your screen recording doesn’t have to be perfect.
Now, the one that you’re looking at is one that was done previously.
You want to make sure that you have all of the steps and you want to have them lined up with the audio as you are narrating what is happening on the screen.
You then need to turn that MP4 into a series of PNG files. One way to do it is to ask ChatGPT to turn your MP4 into a series of files.
That’s what we’re going to do. We’re going to turn this MP4 video, which is short, into 20 PNG files with the file names in sequence.
We’re then going to click enter. ChatGPT is going to give me all the individual files. I’m going to ask for a zip. Have my zip. I’m going to download my zip.
What I need now is a video transcript. If you don’t have a way of generating a video transcript, you can do that inside of NotebookLM.
So when I upload this as a source to NotebookLM, what’s going to happen is that NotebookLM is going to produce an audio transcript.
It has to produce the audio transcript in order to give you the description. So what I’m going to do here is we’re going to click F5 and I’m just going to change the notebook.
If I click on top of this asset in the source area, this is going to be my video transcript.
I don’t need the summary. Now I’m going to upload those video frames as a source. So my frames are now being uploaded here to NotebookLM.
I’m going to select all the sources. What I’m now going to do is I’m going to select the slide deck.
I’m going to give the slide deck the prompt, I’m going to ask for presenter slides, and then I’m going to ask for a detailed step by step process as indicated by the images and the audio.
So again, I want to use the default length. You’re going to need to test this prompt to make sure it gives you the deck that you want.
We’re just going to now click generate Our slide deck is now complete.
We can now enter our slide deck and you’ll see that the AI will then use your screenshots and then give you the presenter element.
So you’ll see that the more PNGs that you can generate and upload, the better your slide deck is going to be.
And you can also tell the AI when you’re creating it that those PNG files match the audio presentation.
You can get better at the prompt, but even with the most basic of prompt, you’ll be able to create a step by step document by uploading your PNG files and the matching audio transcript.
Now, inside of gemiini, we’re going to do is we’re going to turn on the canvas, we’re then going to upload our slide presentation.
We’re going to ask for Gemini to turn this PDF into a full visual slide deck inside of Google Slides.
And what we’re now going to do is we’re going to click Enter and we’re going to allow Gemini to use the canvas in order to turn our PDF slide deck into a slide deck for Google Slides.
And Gemini will then begin to create the slides in Google Slides. We can now take the slide deck.
We can export this now to Google Slides. The slide deck is now completed. We can now open up the new slide deck inside of Google Slides and our slide deck will then be editable.
Another way to think about this process is to again open up the canvas is to upload our slide presentation and we’re going to attach this notebook.
Remember, this notebook that we used is single purpose.
So all that we have inside of this notebook is the original transcript and then the frames. When Gemini uses the notebook, it doesn’t use the slide deck or the transcript.
It’s already got the transcript and it has the frames here. So we’re going to attach this to the conversation.
So we’re going to go here, we’re going to add in the notebook, we’re going to add in the notebook, we’re going to click add, we’re going to ask for Gemini now to create a step by step guide that we can edit in Google Slides.
So we’re going to take now the original slide presentation that we generated.
We’re going to take the actual notebook, we’re going to turn on Canvas and we’re going to ask for a Google Slide deck. Now, the process does take time to work through.
We’re not going to watch the whole process here, of course, again, if you want to see what the AI is doing, you can Open up this element.
Jim and I will then begin to create the slide presentation process is now complete. We can now export this to Google Slides. We can now open up our new slides.
And our slides are now ready inside of Google Slides. The last process we can try is to add in the notebook. We can add in the notebook.
We’re going to add in the notebook where we have the PNGs and the transcript.
We’re then going to turn back on the canvas and we’re going to ask for the Canvas to build a step by step slide deck.
Now, again, the notebook we uploaded has the MP4 which has been turned into a transcript. It has all of the frames that we took at the beginning.
And so we’re now going to ask for Gemini to turn this notebook into a slide presentation. And since Canvas has been activated, we can now submit.
Gemini will then begin creating the new slide presentation. Once the process is complete, we can export to Google Slides. We will then open up our new slides.
You will then choose the process which best represents the workflow that you want to demonstrate.
So again, we have the slide deck where we used NotebookLM to create the slide deck. This slide deck is in PDF format.
This may be the one that you may like best and may fit your workflow best.
We uploaded our presentation direct to Google Gemini and use Canvas and we allowed the Canvas to then recreate this slide presentation.
So it could be that this is a better representation of your workflow.
We added in the Notebook LM Notebook to our gemiini conversation and we added the canvas and we allowed the canvas to create a presentation.
And it could be that this presentation is the best representation of the one that you want to use.
And finally, Again, we uploaded NotebookLM to Gemini and used the Canvas in order to create another slide presentation.
And this may be the best representation of what you want to use.
The key to the entire process is within NotebookLM and the number of PNG frames that you’re able to get from the video to upload into NotebookLM.
If you’re finding it challenging to get a high number of frames from AI, you can use the free VLC player to create an unlimited number of PNG files.
You do need to make sure that this program is going to be safe for your computer because it’s a desktop program.
You of course have to double and triple check to make sure that this is going to be safe for your configuration and that you’re not getting spyware of some kind.
With this kind of program, you’ll want to have as many sequential PNG files inside of the notebook as you can possibly have for this one process.
And because this is going to be a single process notebook, once you’ve completed the process, you can delete this notebook from your account.
Okay. So that. Thanks. And I will see you in another video.