
Whoa, we didn't see this one coming, but in an announcement today, the international Advanced Television Systems Committee approved and published an update (A/72) to the spec we all use for
over-the-air HD that includes the hottest new codec around, H.264. The problem is that the odds of the US adopting the new standard with ATSC 1.0 so wide spread, is slim to none. It makes sense if you think about it,
we aren't even done using NTSC yet, so the last thing every broadcaster in America wants to do now, is to go and replace all those modulators again. And it goes without saying that all the consumers aren't rushing out to replace all the tuners in our brand new HDTVs. So while it may be some time before the we adopt a new standard in the US, the rest of the world that chooses the new standard when they switch digital can at least enjoy more HD, with less bandwidth.
Man, why did we choose ATSC in the first place? ISDB and and DVB-T were around and we had to be different... I know that they claim there are reasons behind it, it seems like a money making thing to me...
What? You couldn't be more wrong. ATSC was up and running HD long before DVB-T started supporting it.
I doubt we'll see anything come of this, at least not in the next 15 years.
Who would broadcast H.264 when the vast majority of TVs can't receive it?
Seriously. They just shipped out the new converter boxes. And now they adopt H.264? Whoops.
H.264 is power... (some) anime (soft subbed) can be encoded perceptually losslessly at 200-300kbps .. potentially a lower bitrate then the AC3 audio...! of course this needs a simple (not too many odd visual effects or crazy motion), clean (minimal-to-no non-digital intermediaries before it hits broadcast or mass produced distribution format (DVD, blu-ray)) anime series. my test case was Azumanga Daioh. other than a few scenes that picked up noise somehow, it was perceptually lossless (compared to the DVD).
(oops, forgot to add)
...it's a pity that considering how long the mandated switchover took, that the gov't didn't mandate ATSC 2.0 compatibility to be in the CECBs.
i'm sure most standard def content could be done in 2-4mbit with ease (it can be done 1-1.5mbit and look perfect in 2pass encoding .. with a good enough encoder).
Okay, I'm all for "perceptually lossless" but your test case is the epitome of horrible, limited motion animation. And since codecs use reference frames and encode differences from that, tons of static imagery is going to give you great results.
But give that a shot on something like Howl's Moving Castle, which looks beautiful in HD (MPEG2, 18mbps). It would not be possible. And your stated 2mbps for regular old anything is way beyond optimistic. Check the rates on DirectTV if you want to know what's what. My nice looking Three Amigos H264 is 15mbps average, and it's one of my "perceptually lossless" caps.
So to put it bluntly, your numbers are way off the mark. Take a look at Lady in the Water on Blu-ray or HD-DVD, VC-1 (claimed by MS to be more efficient than h264) but compressed to smitherenes, to see how bad things get even when the encode is done very very carefully. And that averages near 10mbps.
-Pie
VC-1 is not more efficient than h264. Microsoft is just trying to get people to accept their codec that barely anyone uses.
This will only work if:
Networks are allowed to simulcast in MPEG-2 and MPEG-4 AVC
They'd have to be given double their spectrum allotment.
People could access either OTA signal. Newer tuners would pick up the better picture signal.
Networks would send the cable companies MPEG-2 and MPEG-4 AVC transport streams. Then the cable companies could integrate either into their systems.
It's too late to shut off the original ATSC spec for OTA and move on to 2.0 alone.
As the source documents say, H.264 will most likely be used for newer applications such as ATSC-M/H (mobile/handheld) and ATSC-NRT (non-real-time) since these new applications will use new receiver hardware anyway. We're not likely to see broadcast stations using ATSC. The only real chance I think it has would be on a subchannel carrying additional content. As long as the primary channel remained MPEG-2 and compatible with existing hardware then customers would not *lose* anything, they just wouldn't *gain* the new content. It is like cable MSOs adding new channels on SDV channels and existing CableCARD customers not getting them, or the satellite companies adding new channels using H.264 while those with existing receivers only get their existing channels. Subchannels could be used to carry things like alternate angles, less popular programs, maybe repeats of major content in a different time slot (air the new show on NBC1 on Monday, replay it on NBC2 on Wednesday).
ATSC 2.0 was long overdue, with MPEG-4/AVC codecs being deployed almost exclusively in all new systems on satellite and Cable. If the fact that the HDTV which needed to be carried also had to use MPEG-2 was not enough, the Mobile TV encoded in any format ( MPEG-4, VC-1 or others) could not be carried on the ATSC systems as it did not allow the Hierarchical modulation available in CoFDM based systems.
Hence ATSC 2.0 is a brave initiatve to bring the good old standard in tune
with the new services which it nneds to carry. ATSC-M should lead to a implementations in the next year.
http://www.mobiletvhome.com