Uncategorized

Online video, 10 years in.

By Gavin, 1st April 2006

It’s about 10 years since people started getting serious about online video distribution.

RealVideo launched in ’97 and there was certainly a lot of dot.i.ness about it. There was stuff bubbling away in the year leading up to it, and many people who have fallen by the wayside since.

I remember trying to describe to “them above” that it was going to be big, but it was hard for a 25 year-old upstart to convince TV people that, one day, that tiny pixellated mush would literally blow their TV out of the air. And the internet guys were all “don’t waste our bandwidth dammit!” (now they all have PVRs on their desktop and torrents a-go-go, yes, you know who you are!).

One consistent thing I’ve witnessed (and occassionally attempted to shape) has been the “format war”. No one knew how important it was (or is) going to be (I’d even include Real/M$ in this). Mark Cuban in my opinion was one of the smartest guys I spoke with, because he engaged with the rights owners, built a framework, then sold at the peak of the market. He didn’t care about the format, and Broadcast.com went for about $6bn.

Real lost the format war by creating a bloatware player that pissed people off. More importantly, I think, it immediately cut off its corporate market. You have to bear in mind that the only places with enough bandwidth were offices and by creating the bloatware nonsense that RealPlayer became got them firewalled out of existence. When I started internal corporate webcasting in 1999, Real actually had a lead – but within 2-3 years we didn’t even need to support them, Windows Media was 100% dominant.

Quicktime came into the streaming game very, very late. Their arrogance aournd Quicktime was stunning at the time (I spoke to a number of “senior people” at Apple – it was really silly in a concept-led market in which addressing the “linear-transmission” *need* of broadcasters and licencees was 100% of the game.

So, while everyone else recovered from the dot-nonsense bust, corprorate-sector video streaming grew, and grew (Tornado grew 100% per annum from 1999-2003 doing it and Yahoo!Broadcast ended up as a *client*). Gradually as everyone picked themselves up again, suddenly broadband was there, iTunes had created a digital marketplace for music, people are talking about “Web 2.0” [sic] and the BBC is re-engineering its organisation from the top-down.

So, it’s all happening, and 2006 is the “reboot” of video. But the issues are still there. Real, Quicktime and Windows are all semi-bloatware, they’re still trying to be brands when they should be invisible.

I remember being really excited about a Java video player I saw in ’98, but then the audio support in Java is awful, and it’s inherently going to ask the client machine to do too much decoding work. I’ve been demo’d the next amazing “no-plugin” [sic] Java patented revolution every 6 months for 8 years now, so wake me up if they actually crack it.

Flash video is the new (re)-entrant. This is pretty interesting. Flash is everywhere, we’ve all hated it at some point, and now it seems it might have a use. Google certainly like it.

But, it’s almost like it’s “not bloatware enough” as a user experience (no proper full-screen, too-minimal controls) but definitely a step in the right direction in terms of transparency. I’ve not worked out yet if their timing is perfect, or arriving 10 years too late. There are different issues these days.

Music has always been easier by virtue of the fact that there’s less data, and it’s only 2 dimensional (amplitude vs time). Video is an absolute nightmare. It’s completely interwoven (and more) with the DRM issues, with hardware devices, and with basic *interoperability* issues that make things harder than even if DRM actually worked. Replace someone’s radio and they’re fine (for example, the amount of Radio now listened to via Digital TV, and the Radio industry loves it), replace someones TV and the whole TV industry goes pear-shaped.

Today, the commercial space doesn’t have the same flexibility to experiment as before. Windows Media is *the* dominant force in DRM (yes, more than iTunes). There’s no way iTunes is actually ready to deal with video yet, but it makes for a good story. Everyone has to pander to the awareness of legal issues – so no free playgrounds without immediate lawsuits.

We’re in danger of really damaging an emerging consumer market, not just with DRM, but by the lack of transparency, and the lack of interoperable formats.

In terms of the “social software” elements, we have a really long way to go. The past year has seen ourmedia, youtube, putfile, google, participatory culture and a raft of others get into the game.

It’s all very exciting but there’s a core building block missing, and it’s very, very wrong. Flickr, del.icio.us, blogs, last.fm, RSS, you name it, are all based on open and free data standards.

Audio has its issues, but MP3 was out there and hackable, and now it has Ogg and FLAC.

Video isn’t there and it’s not nearly as hackable, it’s much harder, much more data and automatically tied up in a rights-based industy that will stomp on people way faster than the RIAA ever tried to. And apart from all that, we’ve never shared video on the cultural scale that we have music – there are different social drivers. It’s interesting that the adoption of Theora seems to be going well, but very, very slowly.

In summary, I think it’s really great that there is an “embedded platform” (i.e. Flash player) in nearly every browser, and this will greatly accelerate consumer expectation of video and promote adoption in a way that hasn’t happened before. I’m sure this will stimulate the market, and since everyone also has Windows, Real and Quicktime players (did you notice how iTunes instist on installing QT – a very interesting way to promote Quicktime video!) it’ll help everyone. And extending the iPod would be a good idea.

The BBC’s Dirac is the perfect answer, and in 10 years hence it might be the dominant solution. I wish that VLC would be the dominant player.
I hope Dirac arrives soon as with an end-to-end solution. A Dirac / VLC + streaming/cache server package would be just great…

In the mean time, there’s a huge scope for a video format to get a (consumer-led) hold – in the way that MP3 did. I’m not convinced “Flash” is it. In the current climate, Quicktime seems as good a contender as any, but their server-side is far behind, and Windows Media just keeps growing on a commercial level.

We’re no where near out of the format wars. It’s still a nightmare for all content producers. I’m now recommending people keep a 15, 25 or 50Mbps MPEG2 that can be “transcoded” into the flavour of the day when it arrives.

It may be another 10 years before it settles.


Aside on costs:
As ever, the cost of commercial video distribution is about equal if you’re a provider (unless you use free software). And the main cost will always be the server hardware and bandwidth (roll on re-multicast – another 10 year dialogue!).

M$ still bundle with M$ Server (woo), Real charge per Mbps, Flash is pretty interesting.

A Flash video server is £3,300

However, the Macromedia site quotes…
“Customers who have very large websites that require media delivery to over 1,000 simultaneous users (or 200,000 viewers per month) should consider deploying with Flash Media Server 2 Origin & Edge Servers rather than stacking Professional Editions.

Origin and Edge Servers help achieve scalable deployments by facilitating load balancing, failover, redundancy and clustering. Origin Server deployments start at $45,000. Contact a Sales Representative to talk to you about your deployment needs.”

That is more expensive than RealServer…

Disclaimer