Commentary: Flinging and Flicking

At this year's European Content Delivery Summit our opening keynote was given by Beat Knecht, the CEO of over-the-top (OTT) provider Zattoo.

Knecht has a view of OTT as an established media type, while many groups (YouView, Apple, and Google come to mind) are still nascent in the territory and try to paint a picture that they are the pioneers. Don't get me wrong: These more familiar organizations are indeed pioneers in some aspects-however Zattoo has simply got on with delivery of hundreds of linear TV simulcasts for 5 years while all the others have been working out if it is even possible! Zattoo now delivers hundreds of millions of streams each month and has quietly become one of the largest users of CDNs in the sector.

While the Content Delivery Summit was not specifically looking at end-user technologies, obviously all the attendees were critically concerned about emerging trends in the access and application layers. During the presentation Knecht made a comment about "flinging." It was new to me, and I subsequently found out to many others in the room.

The idea is that the user selects their choice of video stream on their personal device, and if they wander into their living space and wish to transfer the video from their device to the main screen in the room (i.e., the TV) they swipe their finger across the device screen in the direction of the main screen, and the video output is rendered to the main TV display.

OK, it's a neat trick, and is apparently quite established (Apple had already announced before the Summit that it would be available for iPad/iPhone/iPod Touch users to employ with Apple TV devices). However it really caught my attention, and that of a number of the attendees. Indeed we had several discussions about it over the course of the conference. These broke down into tech discussions about how a fling could be effected, and also discussions about the effect such a technique could have on other trends in the sector.

I think they are interesting enough to break open here and perhaps stimulate some more thoughts from our readers.

Let's start with the tech discussions about how a "fling" could be implemented, and perhaps introduce a new idea we dubbed "flicking."

If you are a PC user you will be familiar with pressing ctrl-f6 (or a similar combination) to toggle the VGA output of your laptop from the LCD to the VGA port (perhaps to a projector, etc.). If this was analogous to a fling then it has an interesting set of ramifications-critically it means that the user's device continues to do the decoding. This has a few upsides, most notably that the rights issues are contained to a single user.

However, if the user then takes their device away it could potentially leave the rest of the family staring at a black screen in the middle of a show. Equally problematic the device itself requires enough power to last the full show duration-something that portable devices still struggle with.

Also of note is the limitation of pairing. Assuming a display would need to be pre-paired with the user's device in order to accept a "fling," it begs the question "What happens when the user gets into a friend's car or house and wants to fling the video stream to their friend's screen?" They would have to interrupt the viewing to sort out the pairing. While this makes a lot of sense on some levels, it's far from seamless. If pairing wasn't needed then it would leave a rights issue hanging, since a user could potentially fling a movie to a display in a public place like a bar's TV display, and this would lead to uncontrolled theatrical showings, breaching many rights restrictions currently tied to most media. Once again a fight between the technical usability of media and the rights lobby rears its head.

So stepping on from this issue the next consideration that went around was a notion of "flicking:" The process would be more or less the same for the user, however the back end works differently. I hasten to note that this idea was notional: Those of us having the conversation (including a number of the Streaming Media writers, and several CDN leaders) came up with the idea of flicking to circumnavigate limits we saw in flinging. Consider it released under Creative Commons, and we'd love to see any implementations!

So in a flick, when the user swipes the video to send it to the TV, instead of redirecting the video output of the device to the TV (away from the devices screen but leaving the device to continue the decoding) the device sends metadata to the TV describing the stream source and position and the TV (connected to the Internet) joins the stream independently.

I personally favour this flicking model to flinging for a number of reasons. First the stream can be formatted for the TV, so for example in today's adaptive bitrate world a Wi-Fi connected device running a 600Kbps stream could flick metadata to the TV which may be wired and able to pull a 2Mbps stream. Indeed the TV may have other codec options that are not available to the lower-power portable device, and so overall the picture quality is independent to the device. That's the first good thing.

The second is that once the flick is completed the user can take their device away, yet the family can continue to watch the show on the TV. In "usability terms" this is great, but it opens a can of worms when it comes to rights issues since the "rights payer" who initiates the stream no longer needs to remain present for other parties to enjoy the stream.

Finally it also opened a larger question: If either of these methods take off, then it places the device firmly in the role of the set top box. If the user can search and find a stream on their iPhone or Galaxy Tab and then pop it onto their TV, what role does a YouView-connected TV specification have? Indeed it takes away the need for an EPG spec (the programming discovery is done on the handheld device), and potentially puts much of the complex intelligence into the device and takes it away from the TV or set top box (which interestingly everyone at the conference agreed was a good thing!) All the TV metadata handling need do is act like a traditional media player and leave all the interactivity in the hands of each of the users (many EPGs for each viewing device). Isn't that where it should be?!

It also neatly joined up the surfing-while-watching experience, since I can see several family members all using thief respective devices to look up content, share it with a flick, and continue to surf and interact in their own way. During X Factor this year I realised something very interesting and relevant to the underlying economics of the media that flinging and flicking present: At the moment the "red button" option presented by my cable TV service allowed one vote per household. This means that each cable TV account was worth at most 35p to the show.

In a flinging/flicking model then, while the whole family could watch the show together, each user could interact separately from their portable device meaning that each viewer could potentially vote through the system That means that (without needing telephone voting systems) each member of the household could generate the show 35p-dramatically increasing the revenue potential, and surely that will be music to everyone's ears (even if most of the X Factor artists are not!)