WebRTC Expert Feature

October 13, 2014

Highfive: Almost WebRTC, but H.264 for Today

Highfive is a new video system out from some former Google and Apple employees that I was asked to look at from a WebRTC perspective.  Last week, I met with Shan Sinha, CEO and Founder, and Dave King, vice president of Product Marketing . They detailed why they believe Highfive is poised to change video. 

The team at Highfive comes from Google and Apple, and the investors in Highfive are a fairly impressive group and with $17.5M invested so far, it shows just how much they believe in the concept.

As this is WebRTC World and I know you are interested in WebRTC, first off, Highfive is not a WebRTC system.  The company uses a WebRTC derived stack with H.264 coding and a switched/routed (no MCU mixing) infrastructure in the cloud with variable transmission bandwidth that is not SVC based.  Shan said that they started the company 2.5 years ago and that they started with H.264.  However, he said they hope to move to WebRTC when it becomes a bit more mature.  He was critical of the lack of support from IE and Apple and of the lack of codec resolution, issues that are key to getting WebRTC moving.  I think there are other plans, more on that a bit later, first, a bit about the system.

The Highfive system seems much evolved compared to the clunky video systems of a few years ago and looks a lot more like the WebRTC solutions I have been using this year.  The experience of the Highfive team in Apple FaceTime is clear in the solution.  I would not call the Highfive solution transformational or unique, because what Highfive has done is not really profoundly different than what came before, it is just better in some significant ways.  An analogy is that Highfive is to video conferencing as the original iPhone was to the other smart devices available back then -- limited in some key ways, but with a resulting simplicity that seems very similar to the WebRTC simplicity we have seen.

The Highfive solution starts with a simple room camera unit.  On the surface it is somewhat similar to existing devices such as the Tely Labs telyHD Pro, however, there are some simple and significant differences.  Other room systems have remotes and can be used stand-alone, Highfive does not  -  more on how that changes use in a minute.  The long side bars on the Highfive unit have four microphones that use beam shaping and other techniques to optimize the sound without an external microphone.  This eliminates those pesky cables for a separate microphone unit.

At the core, Highfive is a cloud based video conferencing/meeting system.  The result of the WebRTC like architecture is high video quality, relatively low core/cloud processing cost, low delay, and compatibility with the video acceleration hardware in Apple devices, now available through an API in iOS8 (note that the Highfive release with this acceleration is in Beta).  With Apple exposing the acceleration, I think this is even more reason to include h.264 as a mandatory codec in WebRTC.

In the Highfive system, a user merely creates a named meeting and it becomes a URL.  Highfive uses the domain name of the customer company as the basis for these names, assuring that all company URLs are unique.  So, if I call my meeting “Phil” it is unique to my company/domain, but there can be other “Phil” URLs.  Perhaps we should start calling the URL for a web meeting the UML (Universal Meeting Locator), as this paradigm seems to be the new norm in communications. Send your friends a UML and you or they can join at will.  In Highfive, a created UML can be sent to anyone and they can join using their device.  The UML can be sent in an email, in a text, or can be a standing event.  If you have a company account (based on your email having the right domain name for security), you go right into the meeting, if you are a guest, there is a notice to all in the meeting and anyone can admit you.  Meetings are joined from virtually any device.  For PCs and Macs, a plug-in is added to the browser once (not a separate app), for Smart devices, n app has to be downloaded.  As I said earlier, the Highfive room system never joins a meeting directly.  The model is that a user joins the meeting using a device or PC and then ”throws” the video stream to the room.  This is a simple process. Each room has a Bluetooth and Highfive uses Bluetooth, without pairing, for the signaling.  The result is a simple and continuous transition from the device to the room.  So, I start a video meeting on my device, and with a swipe, send it to the room, or with a swipe can bring it back and leave the room and continue the meeting.  The room system never talks to the device other than the basic signaling of association, all of the video goes to the cloud.  Highfive uses the same capability to send a projection display from the device to the room, or to the entire conference.  This allows an easy use of the room system for display without a video conference.

Shan started the demo by showing how easy it is to install and use a Highfive system.  There are three plugs: power, Ethernet and HDMI (wireless is in the unit, but not yet operational).  After plugging in the unit into the monitor and placing on top using the bracket, the screen shows a number for identification.  The unit has already connected to the cloud and now that number is used to give the unit a name, again within the domain name of the account holder.  And that is it, the unit is configured and operational.  Now you can do a video call and the unit will be available to take over the video in the room from your device.  If there is more than one unit in Bluetooth range you get a named list to select the one you want to use.  The key to the simplicity of the Highfive concept is in the lack of a remote and making the room system a peripheral to a user in the room, it is not a video conferencing system per see, it is a personal shared use peripheral.  This concept, which may be more akin to Steve Jobs insisting on one button on the iPhone, is the key to simplicity.  By eliminating the room system as a stand-alone device, Highfive has truly simplified video.  While I did not have time to do an extended analysis, the video quality was good and seems to be true HD.

This brings us to the final piece of the story, the Highfive room unit is $799, and the price of a user account is, as far as I can tell, free.  As Highfive uses the domain name as the company registration, once a company has bought one unit and registered it to the domain name, any employee in that company can get an account at the Highfive cloud and do video conferencing, with or without a room system.  While Highfive has a future paid account option on their website, the current business model is that you buy one Highfive unit and can use the rest of the system for free, assumedly forever.  While I assume that Highfive expects a ratio of employees to rooms, that does not seem to be a system purchase requirement.  For a 500 person company with 20 conference rooms, a complete Highfive system can be purchased for a retail price of about $16K for the 20 room units, and everyone in that company domain name can use video for all of their meetings.  If we assume a 3 year amortization of that investment, that works out to less than a dollar per month for video conferencing per employee.   It is easy to see that the value of video should exceed that, especially as it can easily extend outside the company as well.

While Highfive is not WebRTC compatible today, talking with Shan left me with the distinct impression that the core technology is close and that WebRTC is in the future.  Highfive used the WebRTC stack to build the core of the platform and WebRTC could easily be integrated.  Shan said they would prefer not to have browser plug-ins.  We also talked a bit about how the system could be used in a non-Highfive WebRTC based meeting if it supported WebRTC and I find the concept intriguing.  If I join a WebRTC session from my device, can I; throw it’ to the room.  This concept is very interesting and should be something for all of us to consider.  The concept of the room as a peripheral is not new, it is something you can do with Chromecast, but simply changing the end point of the video stream from the base device to an alternative video display seems to be an interesting capability.  Something to talk about in November in San Jose.

Edited by Stefania Viscusi

Comments powered by Disqus


Free WebRTC eNewsletter

Sign up now to recieve your free WebRTC eNewsletter for all up to date news and conference details. Its free! what are you waiting for.

Featured Videos