How to loopback Zoom meeting audio for auto-captioning with spf.io

Automatic captioning of your Zoom calls is a convenient way to enhance your meeting with captioning. In this article, we’re going over how to set up your computer and spf.io, so you can provide automatic captioning of your Zoom calls. This is useful where:

  • You are an event organizer supporting a Zoom meeting for a client
  • You have multiple people on the same Zoom call that will be speaking
  • You do not have someone to manually caption your meeting
  • You have a separate computer that can be used for auto-captioning
  • You understand the risk of mistakes that come with auto-captioning

In order to caption multiple speakers on the Zoom call, you will need to loopback the meeting audio so the sound is routed into spf.io to generate captions. Here are the steps to set up and loopback audio of a Zoom call.

Note: If you are also a participant, the computer setup to auto-caption the Zoom meeting must be a separate device from the one that you yourself will use to participate in the Zoom meeting.

For Mac: Computer Setup  |  Within-spf.io Setup  |  Zoom Setup

For Windows: Computer Setup  |  Within-spf.io Setup  |  Zoom Setup


 

For Mac

Computer Setup

1. Download the Source-Nexus Core Audio driver from their website. It is free to download.

2. Double click on Source-Nexus Audio Driver.pkg to begin installing.

3. Go through the installer steps and complete the install.

4. Now, open the Audio MIDI Setup application on your computer. One shortcut to do this is to press Command + Space Bar and type in “Audio MIDI Setup” then press Enter.

5. This opens the application and displays all the audio devices you have available on your computer.

6. Click on the “+” icon in the bottom left hand corner and select “Create Multi-Output Device”.

7. This will create a new audio device. Set the Master Device to “Built-in Output” or your computer’s speakers.

Within-spf.io Setup

8. Create a new session in spf.io then open up the Streamer window.

9. Select the channel of the spoken language in the meeting. Then, click on the “Stream captions” button so it turns red and says “Streaming captions”.

10. Click on the little microphone icon below the “Streaming Captions” button to enable it to begin listening. Then click on the speech bubble icon so it is set to display the raw captions.

11. Next we need to ensure the correct microphone input is selected for spf.io to listen to. Click on the microphone icon in the URL address bar to see which input it has selected. You want “Source-Nexus (Virtual)” to be selected. If you changed the input, you may need to refresh the Streamer window for the change to be picked up.

Zoom Closed Captioning Setup

12. In your Zoom call, click on the arrow next to your microphone button to set the microphone and speaker settings. Select the Speaker as “Multi-Output Device”.

13. Click on “Closed Captions” and copy the API token. (If you do not see this option available, you must be the host of the call and you may need to enable Closed Captions in your Zoom account settings).

14. To set up autocaptioning in spf.io, log in to spf.io, paste in the API token in the Endpoints tab of your session.

15. Now, captions will be generated and appear in the Zoom call when someone speaks. Note: The computer setup to auto-caption the Zoom meeting must be separate from the device that you yourself will participate in the Zoom meeting. If you participate and run auto-captioning on the same computer, your audio will not be captioned in the meeting.

For Windows

Computer Setup

1. Open the Sound settings on your Windows machine.

2. A new window will appear. Click on the “Recording” tab.

3. Right-click on Stereo Mix, and click “Enable”.

4. Then click “OK” to close the Sound settings.

Within-spf.io Setup

5. Now we need to set up spf.io to auto-caption. Log into spf.io and create a new session. Then open the Streamer window.

An annotated screenshot of what users should press to stream audio

6. Click on the microphone icon to enable the microphone and click on the video camera icon in the URL address bar and check which microphone input it is listening to. You want to select “Stereo Mix”. You may need to click on “Manage” to change the microphone input. It will open a separate window where you can change the selected input.

7. Return to the Streamer window. You may need to refresh it so that the change in input is picked up.

 

8. Click on the “Stream Captions” button so it is red and says “Streaming captions”.

9. Click on the speech bubble icon to enable the display of raw captions. A message will appear explaining the potential for mistakes. Read the message then click OK. 

10. Now click on the little microphone icon below the “Streaming captions” button. It is ready to begin auto-generating captions of spoken speech.

 

Zoom Setup

11. Open up your Zoom call and click on the Closed Captions button in the menu. Copy the API token.

12. Within spf.io, click on your session and under Resources > Endpoints tab, paste in the token to the appropriate channel. Then be sure to Save. Now, as the auto-captioning is generated, you should see captions appear within the Zoom call.

 

Didn't see an answer to your question?

en_USEnglish
id_IDIndonesian en_USEnglish
Share This