[opendtv] Re: Superbowl XL(1080)I
- From: "John Willkie" <johnwillkie@xxxxxxxxxxxxx>
- To: <opendtv@xxxxxxxxxxxxx>
- Date: Wed, 7 Feb 2007 14:44:19 -0800
Let me try to explain it another way. When one leaves a will, the decedent
can allocate that person x gets the Ferrari, that person y gets $10,000, and
the remainder goes to person z. That remainder might be $1,000,000 or more,
or if the recently departed spent his/her fortune on (insert your least
favorite vice here) in the final days, it might be less than $1.
Video and audio can and usually do have variable bit rates. One second,
they might be putting out 17.3 mbps, and another second, they might be
putting out 18.3 mbps. There's no way of knowing in advance.
Since the channel has to put out 19.39 mbps, something has to take up the
slack, to handle the "remainder." That's null packets. The only time that
one 'sets' the rate of null packets is to establish a minimum, for the
purposes of reserving bandwidth downstream. One doesn't set the maximum;
the encoder inserts null packets to pad out to the constant bit rate that is
ATSC transport streams.
This is real-time. If you were doing multi-pass encoding with a DVD as the
final medium, null packets could be a sign or quantizer / rate control
settings weren't optimal. At the encoding point in a broadcast station, you
could test that. You have no way of testing that at home, since null
packets are inserted into the stream as part of the routine of creating real
time transport streams.
Somehow, a clip showing one frame of a three hour broadcast (without audio)
doesn't sound like much of an infringement. But, that's up to the NFL.
However, the copy of TS_Reader that you used was for "non-commercial use."
That might be another matter, but that's up to Rod Hewett.
John Willkie
_____
From: opendtv-bounce@xxxxxxxxxxxxx [mailto:opendtv-bounce@xxxxxxxxxxxxx] On
Behalf Of Meehan, John
Sent: Wednesday, February 07, 2007 2:20 PM
To: opendtv@xxxxxxxxxxxxx
Subject: [opendtv] Re: Superbowl XL(1080)I
Uh, Oh! I've never seen that RC indicator on other streams.. I guess I
just broke the law by sending that Superbowl thumbnail over the Web, huh?
I was just wondering if the null rate could be lowered at all.. (I've seen
other stations that are multicasting use rates in the neighborhood of 1 Mbs
for additional channels.)
It'd be nice to reduce the occurrence of null packets as much as possible,
no? (But I can't really say how many is "too many", though - is there any
sort of guideline?)
Too many nulls would mean that the quantizer / rate control settings weren't
optimal, right?
John Meehan
JVC
Other related posts:
