Friday, January 13, 2012

The Next Big Thing for Video

It's the carrier's dirty little secret. More demand means more program flows. Problem is that there's only so much bandwidth. Something suffers. Can you think of what it is? It's video quality. It will get worse before it gets better. And the problem is that everyone, MSOs, TELCOs, CDNs, OTTs, and IPTVs will all suffer. What's worse is that we will be paying for intermittently degraded TV for a while.

At this point we can measure and analyze real-time video quality as it leaves the programmer and as it leaves the headend. There's talk about including a quality probe in every set-top box, but the messages that will flow back in the upstream will bring it to its knees. There's a cost for measuring too much.

So, where does that leave us? How about spatial analysis and reconstruction? It's a technique that seems a bit out of Roswell. It can  rebuild an SD image into an HD, even a super HD image. If you're a member of the IPTV and OTT groups on Linkedin, you've seen samples. They are quite impressive. It is a game changer coupled with MPEG-4. This new compression technique actually improves quality and it's only going to get better.

But, with such a large embedded base of only MPEG-2 decoders, it may be a while before we see any real improvement. So, as we see more and more badly constructed compression, we better get used to more macroblocking, jerkiness, mosquito noise, and smearing.

Next, a look at the biggest contributor to poor video quality - Customer premise.