[opendtv] Re: 625 video quality is good enough....
- From: Ron Economos <k6mpg@xxxxxxxxxxx>
- To: opendtv@xxxxxxxxxxxxx
- Date: Tue, 19 Oct 2004 03:00:05 -0700
1080@50/60p will have to wait for H.264 or VC-1. In MPEG-2 video,
MP@HL tops out at 62,668,800 luma samples per second or 1920x1088@30
frame per second, progressive or interlaced. Given that nobody is going to
build encoders or decoders to an MPEG-2 level that does not exist (never mind
the bandwidth requirements), you'll never see 1080@50/60p in MPEG-2.
In H.264, level 5.1 (the highest level) tops out at 251,658,240
luma samples per second which gets you to 4096x2048@30i/p or
1080@xxxxxx/p. Level 4.2 (the lowest level required for 1080@50/60p)
tops out at 125,829,120 luma samples per second which gets you to
2048x1024@60i/p or 1080@xxxxx/p
I don't have the VC-1 specification here at home, but I'm pretty sure it
tops out at 1080@60p, so H.264 is more "future proof" in regard to
maximum resolutions/frame rates in it's defined levels.
MPEG-2 Transport Streams will happily carry H.264 or VC-1 video, so
the transport is a non issue for 1080@50/60p. H.264 and VC-1 video only
require that new stream_type identifiers are defined. In fact, 0x1b for H.264
video and 0x88 for VC-1 video have already been assigned. They are
referenced in these ATSC candidate standards:
http://www.atsc.org/standards/cs_documents/cs_t3-608reva.pdf
http://www.atsc.org/standards/cs_documents/cs_t3-609reva.pdf
The 1080@50/60p infrastructure probably approaches zero right now.
If you started development on a 1080@50/60p H.264 encoder tomorrow,
I'd guess you'd have a difficult time even finding any 148.5 MHz HD-SDI
equipment to feed it with. My prediction for the proliferation of 1080p@50/60
equipment, encoders, decoders and content would be at least 5 years from
now.
Ron
Terry Harvey wrote:
> At 10:17 PM 10/18/2004 +0100, Alan Roberts wrote:
> >Indeed, MPEG2 would struggle, but a truly anonymous transmission system
> >could carry MPEG4, WM(whatever version) etc, and by thoroughly future proof.
> >What we need to standardise on is the transport stream, not the scanning
> >format. Let's try to get it right this time.
> >
> >
> > >
> > > I agree, 50/60P should be the ultimate format when technology can
> > > accommodate. However, I don't know that it can be done in a transport
> > > constrained to MPEG2.
>
> What do you mean an MPEG2 transport cannot progressive scan??? The MPEG-2
> transport can carry any data one could wish to convey.
>
> Also, the choice between interlace and progressive scan is a no-brainer.
> Interlace was novel when it was introduced in 1933, but it has no place in
> a modern broadcast system.
>
>
>
> ----------------------------------------------------------------------
> You can UNSUBSCRIBE from the OpenDTV list in two ways:
>
> - Using the UNSUBSCRIBE command in your user configuration settings at
> FreeLists.org
>
> - By sending a message to: opendtv-request@xxxxxxxxxxxxx with the word
> unsubscribe in the subject line.
----------------------------------------------------------------------
You can UNSUBSCRIBE from the OpenDTV list in two ways:
- Using the UNSUBSCRIBE command in your user configuration settings at
FreeLists.org
- By sending a message to: opendtv-request@xxxxxxxxxxxxx with the word
unsubscribe in the subject line.
Other related posts:
