Published: April 7, 2020 By

Teaching online is a challenging task for most of us.
We all have resources available to provide better quality instruction to our students.

In this blog I will talk about what is TechSmith Relay's Auto Captioning, a feature that will generate captions to your uploaded video, and share my experiences using it.

Recording and Sharing Online Lectures

There are several ways you can share your recorded lectures.
One way is to simply record your lecture and share the file with your students or to upload the video into a YouTube channel.
These solutions can be easily implemented and executed, however, they have a low degree of interaction with the students which may reduce their participation and integration to the online course.

TechSmith Relay, is available to all CU Denver instructors and students and allows instructors to upload recorded videos from files or link YouTube videos, and share them through Canvas.
Moreover, it has a useful set of tools like conversations and quizzes, that will increase the level of interaction instructors and students can develop throughout the course.

My Experience with Auto Captioning

Before we discuss the process of auto captioning, I want to talk about my experience using this tool, and discuss the main drawback and the main gain of using it.
 
Let's start by looking a where this tool falls short. As you may know, this process is not a magical process, which may lead to inaccuracies in the generated captions. In other words, the produced captions may not correspond with what the speaker is actually saying.
There are quite a few conditions that affect the effectiveness of this process, such as

  • the intensity of the audio track,
  • the speaker's voice pitch, and
  • the speaker's accent.

Actually, I remember the first time I used this tool to generate the captions for one of my online lectures and the outcome was really awful.
It was so bad that I basically had to rewrite the entire script which took a lot of time.
The quality of the audio in that recording was not good enough and there were even times that I had to replay portions of the video and carefully listen to understand what I was saying.

I found out later that having a better microphone with a clearer audio output produced a better outcome in the captioning process.
I have started using my iPhone to record the audio instead of my MacBook internal microphone as it has a much better audio input.  This did increase the postprocessing time a little for my videos as it requires merging and synchronizing the different inputs, but the overall quality makes up for that.
To mitigate this, I am considering the acquisition of a professional capacitive microphone to integrate with my computer.

The caption generation process will take some time, but does not require attention.
However, we need to plan for that. As I usually do the recording at night, when the house is quiet, I will leave the captioning process running during night time and then review/edit the captions the next morning.

Besides all these limitations, having an initial script is significantly better than starting from scratch. Even if you need to make edits all way through the recording, as these edits are usually just wording and do not require changing the time where the sentences happen. This saves a lot of time in the process of captioning the videos.

How does Automatic Captioning Work?

In TechSmith Relay, automatic captioning is a process that runs in background and analyses a video by using artificial intelligence to generate closed captions for that video.
What do I mean by being a background process? Well simply that you need to start the captioning process and then just wait until it is completed meanwhile you can keep using the tool (in this case TechSmith Relay) or doing something else.
Once the process is completed, TechSmith Relay will send you an email to your university account as notification of completion of the process so you can later review and/or edit the generated captions.

The overall process is:

  1. Upload your video file.
  2. Start the closed caption generation process.
  3. Review and Edit the generated captions.
  4. Publish the captions to make it available to students.

In the following sections we review this process.

Uploading a Video to TechSmith Relay

Before starting the closed captioning, you need to create (upload) your video source.
To do this, just go to CU Denver TechSmith Relay, and after log in with your UCD credentials, select Upload Media from the Create option in the left panel.
Select and upload the desired video, for example your lecture recording.
Closed captioning will only work with uploaded videos, not with videos that came from YouTube.

Once the video is uploaded, it will be listed in your Library.

Generating the Closed Captioning

Select from your library the video you want to add the closed captions by clicking on it.
A detailed view of your video will appear, and several actions will be available through a menu.
Among those actions, you will find Quizzing, Conversations and Accessibility, as it shown in the image below.
TechSmith Relay Option Menu

Go ahead and click under Accessibility, to access the captioning option menu.
You would then, select from the dropdown menu the option Generate Captions as is being shown in the following image:
TechSmith Relay Accessibility Menu

Once you select Confirm, a status message will be displayed, and a button with the option Stop Generating Captions will be available until the captioning process is completed. Below is a screenshot of this output.

TechSmith Relay Generating Captions Status Message

The generation process usually takes a minimum of half an hour but no less than than the length of your video.
For instance, if your video is about 5 minutes, the captioning will take between 20 and 30 minutes, but if you video is 1 hour 15 minutes long, that is usually how much time it will take complete the process.
Please, note that these values are obtained empirically based on previous experiences as we do not know exactly the internals of the process.

Reviewing and Editing Captions

Once you received the notification email from TechSmith Relay that your captions are ready, you can either using the link on the email or selecting the option Edit Captions in the Accessibility Menu to review and edit the captions.

In this process you can play the video and review the correctness of the captions in both the text as well as the alignment with the image. You can edit misunderstood words or change the start and end point for some text as you review it.
A sample of how your screen may look like is show in the image below.

TechSmith Relay Editing Interface

While you are editing, if you want to make a pause in processing the captions, just click on Save & Exit (top left button) to save your changes and continuing editing afterwards.

Publishing the Captions

After you reviewed the generated captions, you are ready to publish those captions and make them accessible to the students.
To publish the captions while in the editing screen, click on Publish (top right button) which will make the captions available for the students.

You can obtain mode details on the TechSmith Relay Tutorials.


Javier is a PhD Candidate, Research Assistant, and Lecturer for the Computer Science and Engineering department at CU Denver.
Please check out his departmental bio page here, or his personal website here.