Jargon Watch: Tuners per Serving Group
by Leslie Ellis // February 18 2008
Headsup: A lingo shift is happening in cable tech-talk. It’s subtle, but it matters as an indicator of what’s hot in the Department of High Definition.
If you’ve not heard it, or if you’re involved with squeezing more HD channels into cable plant, here it is: A noticeable increase in the use of “tuners per serving group,” simultaneous with a noticeable decrease in the term “homes per node.”
Here’s an example, lifted from a recent batch of verbatim notes: “It’s reasonable to say that most of the industry is going to eight QAMs (quadrature amplitude modulators) and 500-tuner serving groups.”
If you’ve heard this, you were probably talking to someone involved in switched digital video (SDV). On the MSO side, these are the network architects. The keepers of the digital bandwidth, the reclaimers of analog — the people singly focused on clearing shelf space for all the stuff that needs passage. Especially the HD.
On the vendor side, not surprisingly, it’s the people making digital video switches. Their numbers are increasing. There’s Arris, via C-COR, and Big Band, Cisco/Scientific-Atlanta, Harmonic, Motorola, and likely others, over in the tall grasses.
So why are we now talking tuners per serving group, not homes per node? Because when calculating what’s needed for switching video, the math goes further than the house. (This is also true for VOD, broadband data, voice over IP, and all services initiated within the home, not linearly broadcast.)
Think about your house. If you have two dual-tuner digital video recorders, you’re already in command of four tuners. Let’s say all four are working at the same time, meaning you’re watching and recording on both sets.
Before switching, it didn’t matter how many tuners existed. Broadcast television, by its very nature, is load-insensitive. Point to multipoint. It just doesn’t matter how many things are attached to the end of the spigot.
With switching, it does matter how many tuners simultaneously need stuff from the network. That’s because a big part of calculating the efficiency of video switching involves weird, non-linear math, involving “over-subscription rates.” It’s about figuring out how many tuners, per plant segment, are likely to be simultaneously on.
Recall that switching video works best with lightly viewed content. In television, there’s the popular shows, and there’s the less popular shows. The latter category does well on a switch, in terms of helping to manage bandwidth efficiency.
So, as the engineers and their vendors figure out the best ways to add more HD, they’re thinking about putting the lesser viewed programs into the switch. The math of it involves estimating viewership, as a function of tuners, per grouping of homes.
This column originally appeared in the Platforms section of Multichannel News.