It depends on what your "real world" experience is. The average
consumer (and I'll have to lump you into that category Bert) just
doesn't have the tools to evaluate codecs in any meaningful kind
of way. You don't have access to the original uncompressed source
and you don't have access to the same uncompressed source coded
with different standards and more importantly, different implementations
of those standards.
However, engineers at companies like DirecTV and Dish Network
do have access to such comparisons. And they have decided that
H.264 was worth spending tons of cash on new encoders and STB's.
Clearly, they were convinced by their own experience and testing
(and not just reading stuff on the Internet) that H.264 has a significant
advantage over MPEG-2.
By "significant", were talking about a 50% advantage (that is, H.264 takes
approximately 1/2 the bitrate of MPEG-2 for the same material). Remember
that these satellite companies were shown MPEG-4 Part 2 (with it's 10 to
20% advantage over MPEG-2) and rejected it. In fact, there's no major
broadcast service using MPEG-4 Part 2 today.
H.264 is most definitely the "real deal" and the mere fact that it's
replacing MPEG-2
in some services confirms it's viability. However, MPEG-2 is still very much
alive and at the core of many systems. There still mucho money to be made
selling MPEG-2 products and to call it anything but one of the most
successful
standards ever devised would show no understanding of how it's changed
our lives in a very tangible way.
When I started in the compression industry in 1993, I used to say that
"someday, /all/ the TV you see will be digital". 15 years later, we're
almost
there. That's a fairly short time for such a technological paradigm shift.
I compare it with the replacement of steam locomotives by diesel locomotives
on American railroads. After the production limits were lifted on diesels
after WWII, it only took about 10-12 years before steam locomotives were
essentially gone from the landscape.
The change to diesel locomotives by railroads was purely driven by
economics. Diesels were undeniably cheaper to operate. The same
thing is happening with H.264. It's undeniably takes less bandwidth,
and bandwidth still equals $ today.
BTW, the H.264 de-blocking filter is inside of the prediction loop. You
can't
do that with MPEG-2, so any de-blocking is just post-processing, and not
directly adding to compression efficiency.
Ron
Manfredi, Albert E wrote:
Craig Birkmaier wrote:
One thing is clear from the presentations, h.264 with the FRE
is significantly more efficient than MPEG-2 ( 8 mbps for h.264
versus 24 mbps for h.262 for essentially the same level of
perceived quality.)
Honestly, though, isn't there some bone in your body that questions
these assertions, even after real world experience has shown that things
aren't quite so cut and dried?
I saw that same presentation, and all I could conclude is that the
comparison was probably valid for A PARTICULAR MPEG-2 encoding. Or
perhaps, for some odd reason, the result only applied to a very specific
case of compressing 1080 at 24p material. So I did a little searching.
One interesting tid bit I found was that MPEG-2 encoding can be much
more effective if the composition of the GOP is varied in real time,
according to the subject matter. And that decoders can handle this
variation quite nicely.
Another interesting discovery was an explanation of the AVC deblocking
filter. Seems like a completely separate algorithm within AVC, which can
just as easily be applied to MPEG-2, or any other block-based algorithm,
as it can be to AVC. (Very likely what Algolith did.)
Another interesting and not-too-recent improvement was this Digigami
system:
http://www.engadgethd.com/2005/12/23/rumors-of-mpeg-2s-death-greatly-exa
ggerated-digigami-does-hd-o/
from end of 2005. Which claims that 720p can be compressed in MPEG-2
down to 3 - 7 Mb/s. I assume this means 720 at 24p, presumably DVD
movies. Still, that easily meets the 50 percent improvement over the
normal MPEG-2 rates we have seen quoted to us over the past decade.
(Wasn't 720 at 24p supposed to require an average of 9 to 10 Mb/s in
MPEG-2? Isn't an average of ~5 Mb/s a 50 percent improvement?)
So if the 1080 at 24p file in the presentation was encoded using this
process, would it still require 24 Mb/s for equal quality to the AVC
file at 8 Mb/s? I'll bet you it would not. My bet is that the comparison
would have been far less dramatic if state of the art MPEG-2 techniques
had been used.
There are plenty of articles online that repeat the party line. There
are other articles which present a much more balanced picture of the
codec wars, such as this recent one:
http://www.studiodaily.com/studiomonthly/currentissue/9262.html
Unless we own IP in AVC, I'm not sure what motivates the constant
exaggeration of the facts. I don't think MPEG-2 encoding is hopeless
just yet. I do think that the comparisons made against MPEG-2 are VERY
often biased, and for obvious reasons. Meanwhile, for systems where
migration to AVC is painful, there seems to be NO EXCUSE for this
pretense that we can't move on without it.
Bert
