[Posted Feb 21st 2008 12:39PM]

Another thing to consider is AVAILABLE effective bandwidth. The bandwidth that a provider has available will drive their compression choices. Effective bandwidth refers to any efficiencies a provider can gain by multiplexing different channels into a single signal, allowing a statisitica gain (not all programs need bandwidth at the same instant, and clever algorithms can pack and fill in combination, gaining back otherwise wasted bandwidth).

For example, when DirecTV was restricted to their legacy satellites, using MPEG2 encoding only, they had to struggle to get all their HDTV offerings with only a few available transponders scattered over 3 satellites. Som they compressed a LOT and it showed, gaining them the HD Lite label.

Another example is unswitched digital cable (and FiOs for that matter), which has a 1GHz limit for all channels. Right now they have to compress or drop something else to add an HD channel. They are moving to local switches so that unwatched channels aren't on the local part of the cable, and this will help some.

FiOs is moving to a packet-switched model which is better yet, as it allows them to use the full fiber, essentially ending bandwidth limits.

DirecTV now has a dedicated national satellite and another on the pad, and has more of a "need more content" problem than a "too much content" problem. THe need to FILL bandwidth, not find it. So their new MPEG4 channels are generally compressed only so far as it doesn't degrade. They have a bigger issue in that a lot of what is called "HD" is just upconverted SD, some of it by broadcasters that stretch and crop to get there.