Comcast Puts the Screws To HDTV

Todd Spangler writes "Comcast, like every video distributor, compresses its digital video signals. But to fit in more HDTV channels, Comcast is squeezing some signals more than others. The cable operator claims it is using improved compression techniques, so that most subscribers won't see any drop-off in picture quality. But A/V buff Ken Fowler claims the differences between some of Comcast's more highly compressed channels and Verizon's FiOS TV are indeed noticeable. He's posted his comparative test results on AVSForum.com — and the results are not pretty."

Comments

Re:Not suprising at all

[drspliff]

All Americans suck because their high-def cable TV is riddled with compression artifacts.

Re:Screws to HDTV? Not exactly

[evilviper]

H.264 or even XVID would be multiple times as efficient,

I guess you qualify for the idiotic post of the day award...

Xvid isn't multiple times as efficient as MPEG-2. At high bitrates, the difference is fairly minor, what it does well is looking acceptable at low bitrates. And those 600kbps Xvid DVD rips? They don't look nearly as good as the original, and are typically downscaled by quite a bit.

H.264, even according to it's loudest and most biased proponents, is at most 2X as efficient as MPEG-2. That's a nice improvement, and companies like DirecTV are jumping onto that bandwagon, but it's not exactly free... H.264 decoders need a LOT more horsepower than they would to decode MPEG-2. Not to mention that re-encoding is always going to produce some quality losses, so you won't be getting H.264 at half the bitrate, looking as good as the original.

and the latter is free so you don't have to deal with this crap..

Neither Xvid nor H.264 is remotely free. Xvid is MPEG-4 Part 2 (SP/ASP), and H.264 is MPEG-4 Part 10 (AVC). Both are licensed by the MPEG-LA (aka. "this crap"). http://www.mpegla.com/m4v/m4v-agreement.cfm

If you had bothered spend 60 seconds clicking-through and reading those links (MPEG-4 Part 2 and H.264/AVC) in the Wikipedia article you'd linked to, you'd know that.