Like so many law libraries, we are returning to a mixture of in-person and virtual educational programming. But we still want the virtual output, a video we can post to our public YouTube account for the secondary audience activity. We had a speaker coming in and I decided to play around with creating a video of the in-person activity. It went all right.

If it seems weird for the director to be involved, then perhaps I should explain part of my philosophy. I’ve never worked with really large teams or teams that had a lot of excess resources. When we put on a session, we dedicate the resources we can to it but there’s not a lot of redundancy.

So, when I imagined experimenting with video capture, I didn’t assume or want to impose on staff who were already committed to something else. If anything, this sort of experimentation (if I have the relevant skill set) is a great way for a director to contribute, hands-on. Frankly, I’m better at video editing than at reference questions these days. And many hands make light work.

Capture the Speaker

One obvious benefit of using virtual meeting software for education is the ability to capture the presentation for posterity. The video and audio can be recorded. The speaker is, generally, in a static box. The presentation of slides or other information is also contained within the virtual window that displays the session.

How to do this without a lot of equipment for an in-person session? And what do you do when you are hybrid?

The easiest hybrid situation would be to have your speaker stand at a podium or other fixed location. Then you can put them on a video camera at that location for the virtual presentation. Their slides or other input can be a “screen share” from a device at the podium. So far, so Zoom (or Skype or Teams or Meet).

Harder, and too hard for us at this point, is to allow the speaker to work the room the way they normally would. As soon as you let the speaker loose, you have to accommodate a lot more variables.

This is hindsight. When I was setting up for the session, I didn’t have any notion of what the speaker would do. Our training room is not mic’d and it’s got tables so there’s not a lot of space to move around. We also have a podium, which can anchor a roaming speaker.

We knew that the speaker would have a slide deck and there are 2 screens to project it onto. I set up the camera to capture the podium and one screen and identified a seat at the back of the room (three rows of tables, front to back) from which to run the capture.

Setting the Scene

I used OBS Studio, an open source app that I’ve talked about a lot for video management. I’ve used it for a virtual presentation as well as for green-screen backgrounds. It is also used for video streaming and recording.

One benefit of using something like OBS Studio instead of just a direct video recording is that you can reconfigure your view. OBS Studio allows you to create scenes. You can then toggle between scenes. One might be a close up on the speaker at the podium. One might be a wide angle to capture their movement away from the podium.

I use the Studio Mode in OBS Studio. This allows me to see a preview of a scene before I use it. I can resize and make other changes. So long as the scene I’m working on is not the one in use, live, no one will know. But if you are live and make a change – crop the video, for example – to the live scene, it will change in real-time. It’s also helpful to use the Windowed Multiview to toggle between your scenes (View > Multiview (Windowed)).

I created a single scene, which was one too few. It captured the podium with a bit of extra space, and one slide presentation screen. In the end, the speaker moved around more than this allowed and so, in mid presentation, I needed to create a new scene that showed only the speaker and showed no slides.

Camera Obscured

In a virtual call, the speaker is usually in front of a video camera. They stay there because it’s a static interaction. But with a podium and a room, you don’t know that the speaker will stay in front of a camera at the podium. I wanted to try a couple of cameras to see which would work best with OBS Studio.

I started with my Canon EOS T5 Rebel. It’s an old but reliable DSLR that I use for all of my photographs. Canon released a hurried web cam utility that allowed EOS owners to use their cameras for virtual meetings during the pandemic. The T5 can record directly to the camera. But I couldn’t control anything if I used its own recording system. So I plugged it in to a laptop and recorded using OBS Studio.

I found some latency with the EOS web utility. You don’t really notice in virtual conference calls. But I’d experienced it on live TV interviews with news media. The speaker’s words are not entirely in sync with their lips.

I knew I couldn’t use the integrated web cam on a device so I tried an external Logitech web cam. Normally, it would clip on the laptop I (or the speaker) was using. This time, I taped it to my camera tripod – for stability – and recorded with OBS Studio. It worked great and while the video was a bit grainier than I expected, even on HD, it did a good job of auto-focusing and adjusting as the speaker moved around.

Obviously, if you’re going to try this, test your configuration. I ended up testing two cameras and two microphones and two sitting positions. And even then, I didn’t hit the right combination until about 10 minutes into the presentation.

Audio Inputs

Another thing virtual meeting software captures is the speaker’s audio. A fixed-to-a-podium speaker would be just as easy to capture. A mobile speaker is harder. I ended up using a conference call microphone from Anker that we have. It’s omnidirectional and supports Bluetooth. After trying the onboard laptop microphone, which was too far away from the podium and shielded by the laptop screen, I tried the Anker.

It worked well. Not as well as a lavalier microphone that allows full, wireless mobility for the speaker. But it allowed me to place a wireless microphone near the presenter. As the presenter pivoted, the sound quality changed when their back was to the mic. But it was still powerful enough to pick up the audio, although with the ambient live setting noise.

I was worried the Bluetooth from the microphone to the laptop in the back where I was, about 15 feet, would create latency. But it wasn’t noticeable if it existed. I think that we will probably want to invest in a lavalier microphone that supports Bluetooth for the future.

Pulling the Threads Together

You’ve got your camera. You’ve got your audio working. You’ve created your scene in OBS Studio. Now make sure you’ve create an audio input/capture in OBS Studio for your microphone. And check out your video recording settings.

In my case, it was defaulting to use the MKV video format. I debated using MP4, which is more common an output. But the risk with MP4 is that you have nothing until you stop recording and save. If your production dies mid-recording, you have no capture. MKV format will mean that you’ve always captured something.

You can also increase the video quality rate and make other tweaks to improve your output. Make sure you have plenty of storage space. The video file for 1 hour was over 1 GB. We use SSD hard drives which will create less latency and this may be a wise choice for you too, if you have the option.

Create the CLE Video Output

Your speaker shows up. You power up the camera, the microphone, OBS Studio. You start the OBS Studio camera and you hit Start Recording. Some time later, you hit Stop Recording and you’re done. You’ve got a video file, with audio, of something. Now to cut it together.

At the very least, you are going to want to trim the video. This will cut off any extraneous video at the start or end or at any point in the middle. You can do simple things like this with Microsoft Windows Movie & TV app.

But if you want to do more complicated things with your video, look at OpenShot. It’s another free open source app. I was thrilled to find my staff are already using it. I’ve used Cyberlink Power Director in the past, because I got a free copy with a purchase of some hardware. OpenShot is easily comparable to the version of Power Director I have.

At this point, you can make this as complicated as you want. It can be dead simple – drop the video in, make some minor edits, perhaps add a title screen, and you’re done.

Or you can rework it. I think you should do this, because the secondary audience is not experiencing an in-person presentation. There is no reason to treat the experiences as being the same.

Think about video you see of concerts or other events, even news broadcasts. There are close-ups, there are switches to wide angles, there may be nothing but a voice accompanied by a screen or visual of information. You can do all of this in OpenShot.

This was my approach.

First, I gathered together all of the raw materials I would need. I had the video with audio track of the speaker. But because of the motion, I had lost a visual of the PowerPoint slide deck and the embedded YouTube videos the speaker had used.

Since I had a copy of the PowerPoint, I opened it and exported the entire slide deck as image files (.jpg or .png). This created a folder of images that I could then import into OpenShot.

I then imported the video file into OpenShot. I also found the 3 YouTube videos that were embedded in the PowerPoint, downloaded them from their sources (higher quality, as well as understanding any licensing or copyright issues), and imported them into OpenShot.

Now I was ready to work. At this point, it’s just a matter of dropping your imported library of items onto different tracks. The most important thing to remember about OpenShot is that your tracks are stacked. So you can overlay information by placing it on higher (number) tracks.

In my case, this meant putting the main video on the lowest track. To do that, I just dragged and dropped it on Track 1. Once it is there, you can slide it left and right to the time that you want it to start at. At this point, I had it start at 0:00 minutes. In the end, I slid it right so that a title slide would appear first.

You can also split your audio and video tracks. Right-click on the video, choose to Separate Audio and make a choice. In my case, I chose Single Clip.

Screenshot of OpenShot right click menu for Video, to Separate Audio

I don’t think you normally have to do this. However, I knew I had audio on the three YouTube videos. If I played the high quality YouTube video in the OpenShot editor, it would duplicate and overlap the audio from the presentation, which was recorded by the microphone as it was heard by attendees. By splitting the audio out, I could delete the overlap and just use the higher quality audio from the YouTube video file.

Now, it was just a matter of playing the presenter’s video within OpenShot. As they spoke, I dropped in the static slide images from the PowerPoint slide export I created. This meant that, in the output video, you switch between the speaker on video, then over to a slide with a voice over, and back to the speaker.

You can see what that looks like below. This is OpenShot, showing the library of imported items at the top left, the live video view at top right, and the multiple tracks. Track 5 is just PowerPoint slides dropped in place. You can resize them to last a certain amount of time. The gaps between the slides on Track 5 mean that the lower Track 4 video of the speaker will be displayed. Track 6 was for the YouTube videos.

Screenshot from OpenShot, showing tracks at the bottom, library content at top left, and the live preview at top right.

It took about 2 to 3 hours for me to complete this editing. Part of that is that you need to listen to the entire presentation again. So, for a one-hour CLE, that’s one hour just as table stakes. You could shorten the overall time by:

  • noting when a presenter changes a slide, by writing down a timing during the live presentation;
  • practice with OpenShot. I found that all of my earlier work with both OpenShot and Power Director meant I wasn’t hunting for help or scanning app settings for how to do things;
  • ensuring you’re working on a PC with enough memory and fast hard drives. These will slow your activity and, in my case, crashed the OpenShot app at least once;
  • also on more powerful hardware, finalizing the edited video took about 30-40 minutes, which is totally dependent on the size of your files and your computer’s hardware.

The imported video started at 1 GB. The exported video ended up at 2.4 GB. The upload to YouTube took 35-40 minutes, although that will vary with your internet bandwidth.

What about captions and accessibility?

Another benefit of using a virtual meeting app, you can enable its auto-captioning (Zoom, Skype, Teams, Meet). Otherwise, you can use YouTube’s auto-captioning, created as you upload your video file, and then clean it up. You don’t need to publish it on YouTube.

Once you’ve cleaned it up, in the YouTube Subtitle Editor, you can click a 3-dot menu and download your subtitles/captions with timing in a .sbv file. .SBV is one of a number of captioning formats.

Screenshot of Windows Notepad showing the .sbv captioning file I created for the Miracle Max video I embedded on this site from YouTube.

You won’t find the video on the law library’s YouTube channel. It was an experiment and I think it showed we need work. For the future, I’d like to:

  • investigate a Bluetooth lavalier or have the speaker commit to staying in a fixed location
  • prepare more scenes so that I don’t have to make changes during a presentation / performance
  • anticipate use of copyrighted material that might make our use of a video impossible, perhaps by using speaker waivers

We also would need to create an easy to replicate process, so that whichever staff are available can create the video output. If you are running an education session with just one person, this may be more than they can take on and you should consider adding a second person (even if it’s the director!).

But it makes me realize what a benefit using virtual meeting software for CLE education presentations is. It streamlines a lot of this capture for you. By locking the speaker and the presentation into a screen, it limits the variables. You can either take the raw output from the virtual meeting software or use OBS Studio to create a scene for the slide deck and the speaker, and toggle between them live. This would require less expertise and a lot less hardware testing and maintenance.

It’s a brave new world. Even after two years of pandemic play, I realize I’ve got a lot to learn about virtual education and meeting delivery. But I think open source apps like OpenShot and OBS Studio can make a big difference.