Review: Panopto Lecture Capture and Webcasting System

The increasing acceleration of worker turnover makes knowledge capture and management a critical function for most organizations, large and small. Even organizations blessed with low turnover can benefit from capturing information detailed in presentations, webinars, and less formal meetings for archiving, sharing, and training.

Like most video-related functions, however, knowledge capture has been vastly decentralized during the last 5–10 years. In 2000, for example, the corporate video department had to shoot and capture an executive presentation or training session. Today, anyone with a webcam can host a webinar or present some form of training. For this reason, organizations considering a knowledge capture and management system should move products and services that don’t require a proprietary capture station to the top of the list. Panopto is one such company.

As an introduction, Panopto (company and product name) is a lecture capture/webcasting system that creates a library of videos for public or private viewing. You perform all capture on standard computers or mobile devices, so there’s no proprietary capture stations to buy. Capture options are unusually flexible; for example, you can capture multiple inputs during the live event, such as multiple video cameras, PowerPoints, and screencasts, and maintain them as separate objects for later integration into on-demand presentations. PowerPoint slides are input as text, making it simple to index presentations by slide and to search within a presentation for relevant information.

Once captured, viewers can access video in the library via computers and mobile devices, and logins and video views can be integrated with different learning management systems and Active Directory-compatible portals. On-demand and live videos can also be made available for general web viewing on computers, though notably, mobile devices can’t view live webcasts at this time.

In terms of pricing, Panopto is offered as a SaaS, with per-user pricing starting at around $50 and dropping significantly for very large numbers of customers. In the corporate market, you pay for both viewers and producers; if you have 10 producers and 500 viewers, you pay for 510 licenses. In the educational market, you only pay for producers; students can view for free.

Overall, Panopto was very well-conceived, and in general, is very easy to use for both presenter and viewer. One key differentiating feature, however, is the ability to input video from multiple sources, such as a document camera or extra webcam. I found this feature challenging to use both during the presentation and when preparing content for distribution; I explain why near the end of this review.

From a high-level technology perspective, Panopto chose Silverlight as its recording, management, and playback platform, which generally worked well in my tests, but may force potential viewers to download the Silverlight player. While acceptable within the enterprise, it’s a negative for those seeking to use the system for external viewers, many of whom won’t have the Silverlight player installed. In addition, recording capabilities vary in both functionality and interface on Windows and the Mac, with the Mac the red-headed stepchild. As you’ll read, if the Mac is your primary platform for live webcasting, Panopto is not for you, at least in the short term.

Panopto is transiting to HLS and DASH this year, which means the end of Silverlight, and also hopes to provide identical form and function on its Mac and Windows clients by the end of 2014, as well as mobile viewing of live events. So the company obviously has a lot of critical development tasks on its plate, and much of what you see in this review will look a little different by the end of 2014. With this as prologue, I’ll walk you through the various ways you can create, deploy, search for, and playback content from the Panopto system.

Creating Content on Panopto

As with most SaaS offerings, you enter the service by logging in with a browser. You start all content creation activities on Panopto by clicking the Create button in the main interface. As you can see in Figure 1, there are five different ways to create content, which I’ll run through in the following sections.

Article9a.jpg

Figure 1. Creating a recording in Panopto

You can record a new session on both Mac and Windows clients, though you can only produce a live session (or webcast) on a Windows client. Click Record a new session, and you’ll be shown download links for either the Mac or Windows Recorder; once that’s downloaded and installed, you’re ready to record.

The Windows Recorder (Figure 2) is more visually similar to the actual presentation than the Mac Recorder (Figure 3), which makes it more intuitive. On the Windows client, you configure the Primary presentation source, typically a webcam, separately from the secondary sources, which can be PowerPoint, your desktop, or a document camera or other video source. Webcam options range from 320×240 to 720p at a maximum data rate for that component of 1.5Mbps. Depending upon which setting you choose, Panopto will encode the video file into two (Standard quality) or three (High and Ultra) streams for adaptive delivery. Your secondary screen can stretch as large as you care to make it, though Panopto recommends the settings shown in Figure 2, 1280×800, 500Kbps and about 10 frames per second.

Article9b.jpg

Figure 2. Here’s the Windows Panopto Recorder.

Article9c.jpg

Figure 3. The Mac client isn’t as logical or functional as the Windows client.

Once you’ve got all of your sources ready and configured, you press the Record button (Windows) or Start Recording button (Mac) to start broadcasting. Then you start talking, and use normal commands to view your PowerPoint slides or an application that you’re demonstrating (Figure 4). At this point, the Recorder is not in view, though any questions asked by the live audience will pop up so you know to address them. 

Article9j.jpg

Figure 4. Here’s what a presentation looks like during viewing

During the live event, all video sources, including screen capture, are streamed to the Panopto servers, so viewers can switch between them during the event. If you have multiple input sources, such as a webcam, screen capture, and one or more document cameras, you have to stream them all out separately, so outbound bandwidth requirements bear watching.

Panopto also stores higher-quality versions of the streams to hard disk during the event. These are later uploaded to the Panopto servers to serve as the basis for the on-demand presentations. More on this later.

Not surprisingly given the reliance on Silverlight, Panopto captures in the Windows Media format using the VC-1 codec, converting the streams to H.264 for delivery to mobile and for all the desktop streams that I could check. I’m sure that Panopto will change over to H.264 during its switch to HLS/DASH.

More Content Sources

Working down the list of ways to create content in Figure 1, the next technique is to upload a video file, which is as simple as it sounds. Note that you can drag and drop multiple files simultaneously, a nice convenience for those transitioning into the Panopto system or who otherwise need to upload lots of content. As with all content in Panopto, once uploaded, you can add descriptive information or captions, or elect to have Panopto transcribe the video at a cost of $2.75/minute for 2-day turnaround. Captions added to the file also become searchable.

To create a live webcast, you choose this option in the Create menu, which opens a screen for the name and description of the event, with an “Anyone on the web can see this” check box for making the event public. Once checked, the system hands you a public URL you can distribute for general internet viewing. You can also grab a code to embed the Panopto player into your website for viewing there.

When it’s time for the event, you run the Panopto Recorder (Windows only), and choose that event in a drop-down list — no, you can’t produce a live webcast on the Mac version, though you can record a presentation. Then click Record on the Windows Panopto Recorder to start the event as normal, and the incoming streams are fed through this event.

For scheduled recordings on other systems, you would need to first configure the remote recorder into the system. Then setting up the event is similar to creating a recording as discussed above, except that you insert a future date and time into the event creation screen.

Combining Video and PowerPoint

Even after you start using Panopto, there will probably be times when you can’t use the system to record an event, such as informal events where you can bring a video camera but not a notebook or desktop or videos captured with a mobile device. For this, Panopto provides an editor for integrating uploaded video with a PowerPoint deck. I tested this with a presentation from the Streaming Media Europe conference in 2012 and found it unintuitive and painful.

The browser-based editor (Figure 5) works by downloading a very low-quality video stream and low-quality slide images into the editor. You play the video, and then click an Add button next to each slide when you want to add one to the presentation. Rather than simply inserting the slide at that point, and moving on, you have to click the playhead in the timeline to insert the slide. Not a huge burden, but it is definitely an unnecessary step, multiplied by 60 for each slide in the 1-hour presentation.

Article9d.jpg

Figure 5. The Panopto editor was clunky when integrating video and slides.

Even worse, on the two machines on which I tried this procedure, one a brand-new HP Z1 all-in-one workstation running Windows 8, the other a Mac Pro, after adding a slide or two, the video window disappeared into a white screen. That is, the video window shown on the upper left in Figure 5 would turn white, and you had to save the edits, refresh the browser window, and pick up where you left off. This was complicated by an overly twitchy video playhead that made it difficult to scan to a specific spot in the video.

I was testing with a 60-minute video file, which perhaps was the source of my problem because the editor was much more responsive in the shorter webcasts that I edited. Still, if you’re integrating long-form videos with PowerPoint slides in the editor, you may be in for a frustrating time.

Mobile Recording

Panopto offers a free iOS app that lets you record videos and then upload them to the server, after which you can present them stand-alone or integrate them with other presentation elements. This function worked well on my iPhone 4S. Note that Panopto has an Android app in the works for later 2014 delivery that will provide similar functionality for that platform.

Viewing Panopto

So far, we’ve concentrated on the creation side. Now let’s turn our attention to viewing. Overall, searching for and finding live and on-demand content to watch in the video library is straightforward on both desktop/notebooks and mobile clients. Actually watching the events is where things can get a little hairy, particularly if you add a third source such as a document camera into the mix. But let’s start with the default experience that most viewers will have. 

The live experience is shown in Figure 6. Beneath the webcam on the left, you can add private notes or public comments, or read information about the webcast. You ask questions by typing them into the Ask a question field beneath the slides. Note that there’s only one tab in the upper left of the larger player window on the right, labeled Screen Capture.

Article9e.jpg

Figure 6. Here’s the live experience.

The corresponding on-demand presentation is shown in Figure 7. During the presentation, Panopto monitors when you change slides, creating index points that appear in the Contents tab beneath the player and also the thumbnails beneath the main player window on the right. All text, also pulled from the slide, is searchable in the Search tab, along with captions or any other metadata.

Article9f.jpg 

Figure 7. Here’s the on-demand experience of a live webinar (or recorded presentation).

Again, all this is pretty straightforward. The only concern is in the main viewing window on the right, specifically the two tabs on the upper left for Slides and Screen Capture. The Slides tab shows the PowerPoint slides, while Screen Capture shows what was shown in the same window during the live webinar. If a viewer clicks the Slides tab for some reason, they may miss action taking place in the Screen Capture tab. Since the Slides tab performs no useful purpose that I can see, Panopto should consider removing it or at least allowing users and producers the option to remove it. Either way, it’s probably not a big deal. 

Paradoxically, the situation for mobile viewers of on-demand content is much cleaner. That is, on tablets, all they see is the Screen Capture content with the webcam on the upper left, similar to what you see in Figure 6, except there are no tabs, just the content. On smartphones, the webcam window gets embedded on the lower right, outside the Screen Capture content, so there’s no obscured content.

The potential for problems and confusion gets much more serious when you add a document camera or other source into the picture (Figure 8). During the event itself, Panopto displays another tab for Object Video in the main window, which contains the document camera. As presenter, you have no simple way to direct which tab the viewer watches; if you’re switching back and forth between your desktop and the document viewer, you have to instruct viewers to do the same. Or, within the Panopto interface, you can add the document camera to the presentation when you want to show documents, and delete it from the presentation when you’re done with it. This would open the Object Video tab while the document camera was configured in, and delete the tab and return to the Screen Capture tab when it’s removed.

Article9g.jpg

Figure 8. A live webcast — you have to tell the viewer to switch tabs to watch the correct stream.

Both of these solutions are awkward and error-prone; Panopto needs a simpler mechanism to allow the presenter to control what the viewer actually sees. For example, webcast systems such as Go To Webinar or Google Plus Hangouts typically record and transmit what’s showing on the screen, a much simpler approach. While you don’t maintain the screen and document camera as separate inputs for subsequent editing, I’m not sure this is useful to most presenters, who typically want to edit as little as possible. 

On-demand viewing of a webinar recorded with a second video source is even more confusing, as shown in Figure 9. Unless you edit the second video in and out after the presentation, the viewers will be forced to manually change the tabs themselves to view the right stream. On mobile devices, where the tabs aren’t available, they’ll only see the document camera unless and until you edit the event.

Article9h.jpg 

Figure 9. Until you edit the event, the viewer has multiple viewing choices that won’t automatically switch. 

Editing would have been easier to manage if the Panopto editor was straightforward, but I found the tool very unintuitive and lacking expected niceties such as an undo button (Figure 10). In theory, editing should be simple; you edit out what you don’t want on the respective timelines and you’re done. But that wasn’t the case in my use, and the multiple tabs in the main player window (Screen, Object, Object, Slides) made it very difficult to understand what the viewer will actually see and when. While operations such as deleting heads and tails from the session are simple, I found steps beyond that very challenging.

Article9i.jpg

Figure 10. I found the Panopto editor very challenging. 

Fortunately, if you need to show a document camera or other video source, there’s a simple workaround: Don’t configure it as a separate source within Panopto; just show the input from your desktop in a separate viewing application, where it will be recorded and streamed as part of the screen capture. While this won’t save the input as a separate editable stream, it will vastly simplify both the live event and any postevent editing. 

Again, if you’re not interested in adding a document camera or similar source to the mix, the last 500 words of criticism are largely irrelevant; your viewers will see what you want them to see during the live webcast and when viewing the on-demand version, with no editing required. But if you are interested in this feature, keep in mind that it may be challenging to use both during and after the webcast.

Overall, if you’re considering a learning capture system, you should check Panopto out; the economics of a system that doesn’t require proprietary hardware capture stations is simply too compelling. However, be sure to go into the evaluation with a very well-formed view of the types of presentations you’ll be producing, and determine the precise workflow necessary both during and after the presentation to achieve the desired result.

About Jan Ozer

I help companies train new technical hires in streaming media-related positions; I also help companies optimize their codec selections and encoding stacks, and evaluate new encoders and codecs.

Check Also

Tuning for Metrics: What About VMAF and VP9?

If you’re comparing codecs with video quality metrics, you should consider tuning for that metric. …

Leave a Reply

Your email address will not be published. Required fields are marked *