As presented in Banff, Alberta, Canada:

The keynote speech from the 24th AES International Conference
by George Massenburg
|
My esteemed associates...brothers & sisters in fine audio...above all,
friends. Welcome. Let me personally thank all of you for coming to Banff.
How ironic, and maybe appropriate, that we are holding our Multichannel
conferences in front of such imposing mountains. Maybe moving
Multichannel ahead is perhaps easier than moving these mountains.
Today I stand before you as a producer of music, and I am humbled and
especially proud to be here with some true masters in the field.
The headline for the 24th AES International Conference on Multichannel
Audio is "The New Reality". Before we delve headfirst into what promises
to be the comprehensive, diverse and challenging proceedings I'd like to
take a few minutes to mark where we've been, where we are and
where we
seem to be going in Multichannel.
Many of us gathered two years ago for the Surround Sound Conference in
Elmau, Germany and discussed what we thought was needed to further
Multichannel Audio as an art and as a basis for new business. There have
been many advances since Elmau and there have been some interesting
impediments.
In the last few weeks I've been mentioning to a lot of folks in the recording
industry that we were having a "Multichannel Conference" in Banff. I heard
no lukewarm responses. More than a few in the industry were skeptical.
Enough so to at least cast a bit of a pall over the claim that
Multichannel is a "reality" at this point in time.
Just one example. If you "Google" for the word "Multichannel" on the Web
you will be surprised to find that the highest-ranked hits directes us to any
number of Multichannel-Marketing sites-one "delivery channel" being, for
instance, brick and mortar retail sales stores, another "channel" being "direct
Web sales" or "mail-order" & etc. It shouldn't surprise us that
Multichannel Retailing is very clearly a reality..
But let me try to synthesize where Multichannel Audio is today, quickly
summarize the different areas, present examples, and give you some brief
thoughts.
You know after doing this list and looking back on it, a pattern emerged that
I'll point out...eventually.
|
1. As regards Classical Music recording and production
a. The good news from major labels in the US is that given
reasonable budgets nearly everything is recorded with the
resources (mics and tracks) sufficient to at least eventually re-mix
for surround. I'd like to be able to tell you that
Multichannel Audio is a given for classical music releases but,
unfortunately, efforts are mixed.
b. At Telarc over 75% of releases go out SACD with both Stereo
& Multichannel streams.
c. Clearly Philips is very actively supporting core classical and is
doing so in SACD Multichannel. They have recorded over 50
titles in surround.
d. On the other hand, most of the majors, including BMG &
Warner's worldwide, have abandoned core classical music
almost completely, save for reissues or "special projects". Even
at Sony core classical is much more likely to be released in 2-
channel stereo; only perhaps 5-10 of their 100 or so releases in
a given year have Multichannel streams.
e. It's hard to ignore that the purely classical music business in
any format is still struggling in the marketplace, although music
on the margins of classical - world music, Jazz, avant-garde,
contemporary- are faring much better.
f. As you look further down the scale to the independent labels,
and to more mid-level classical productions, increasingly one is
likely to find them recording only in stereo, even when the
added expense is modest. I've heard more than once that there
is no confidence in the current implementations of the center
channel and there's quite a bit of work being done in 4.0. And,
at the bottom of these mid-level budgets, there are more than a
few projects recorded to 16-bit DAT.
g. Nonetheless, multichannel classical music presentations make
the most impressive demonstrations. The best of these are often
utilized in high-end audio sales rooms to invoke "the John
Eargle effect", which states that the better the demonstration is,
the faster checkbooks will fly out of your pockets.
2. Concerning Multichannel in the Cinema.
a. The success of multichannel in the Western world (now thought
to include India as well as Europe) is in the range of 56 to
58,000 screens. Of those about 35,000 are in the states. Both
Dolby and DTS have penetration in the 90% range. THX has
about 1,000 of these screens, mostly on the West Coast.
b. Film scores are for the most part being produced multichannel.
The music mixers that I've spoken to in Hollywood are
recording and delivering ALL of their scores in 5.1 or in a few
cases, 7.1. In most cases the final mix makes use of that
Multichannel score for the final theatrical release.
Multichannel in cinemas is growing, US and worldwide.
c. With the release of the Star Wars Episode 1, Dolby EX (acting
on an initiative from Lucasfilm) provided an additional array in
the rear. I'll ask us to make a note here that what was desired is
a more defined localization in the rear, especially for pinpoint
sources at or near the center rear. It's still desired.
d. Here's a hopeful note: if a film deal is made for pop music and
multichannel stems are provided, it generally finds its way into
the theater in 5.1.
e. Unfortunately, it is not clear that there is any support on the
horizon for extended resolution or bandwidth, although there is
always room for more speakers (more about this later). And
when the music and effects mixers that I spoke to were asked
what they'd most like to see in the future in cinemas, their first
response was, "Make all of the speakers the same as the full-range
speakers in the front."
f. Theater owners, a stressed market segment if there ever was
one, will no doubt weigh the investment in any new technology,
be it direct digital delivery or full-range rear speakers, only
against increased popcorn sales.
3. DVD-V and Multichannel
a. The success of DVD-V is well documented and I won't go into
it here except to say that it was last year's success story and
there's no doubt that it will continue to grow.
b. There are somewhere between (as few as) 45 million and as
many as 125 million DVD-V equipped players out there in the
world, depending on who you ask. It's extremely hard to tell
how many home theater systems are actually set up and playing
multichannel, but conservative estimates are less than a million.
c. It's been confirmed with several houses that remix for DVD-V,
that these releases almost always use the original Multichannel
stems, including the 5.1 underscore as well as Multichannel
dialog and effects. Also, if somewhat rarely and only if the
budget supports it, new 5.1 content is implemented.
d. Also, successes abound in music on DVD-V. Let me give you
some figures for Canada alone:
i. Norah Jones live DVD at 30,000
ii. Diana Krall "Live in Paris" DVD-V 50,000.
iii. Paul McCartney DVD is over 50,000; interestingly, sales
of the CD of that concert are only 71,000.
e. I found it difficult to get exact sales figures for the Eagles
DVD-V of their "Hell Freezes Over" live tour recording, but
rumors are that it's significant.
2. There have been notable experiments in Multichannel Broadcast.
a. In February our associates at Swedish Radio, who are
represented here, announced that they would begin Digital Video
Broadcasting on Sirius 2 satellite, available to the Nordic and
European regions, broadcasting DVB along with 96Khz 24bit
audio in DTS format. The trials were to end in April and we're
anxious to find out what they've learned.
b. Secondly, they are offering a free 5.1 demonstration download in
DTS format, both 44.1 and 96k from their website. I
downloaded (legally) a sample, burned a DTS format CD-R,
and had myself a quick and easy Multichannel demo. In 9
months they reportedly saw 800,000 downloads, 80% of these
to users outside Sweden.
c. Broadcast television in the US is a different story. The
2002/2003 television season marks the fourth consecutive year
that CBS broadcast the majority of its primetime schedule in
HD, and the second season that it has offered all of its scripted
entertainment series in the HD digital format. By the end of
2002, CBS expected to be transmitting digital programming
across more than 100 owned and affiliated stations, reaching
over 83 percent of the country.
d. Unfortunately, to the best of our knowledge, until this year's
Grammy presentation CBS had never presented anything in 5.1
and HD; Phil Ramone and his team basically had to direct a
seminal installation.
e. The future is a bit more promising: the US government has
mandated benchmarks for the implementation of HD television
over the next decade. And the government is quietly driving
resources to PBS stations to fund Multichannel production and
broadcast. There are plans for some shows, such as "Live from
Lincoln Center", to go out this fall in HD & 5.1. "Austin City
Limits" is reportedly generating material, and is almost ready to
do 5.1 broadcasts.
f. Something to think about: at the most recent NAB show,
surround music packages for station ID tags started to became
available.
3. Popular Music Recording & Releases.
a. Quite a few record companies are ordering 5.1 pop mixes along
with 2-channel work even when their immediate plans only call
for 2-channel music releases. Most often, 5.1 mixes are
initially used in DVD-V's. Occasionally, there's a
simultaneous multichannel release.
b. Notable among multichannel record companies are…
i. DTS Entertainment, represented here, one of the first to
make a concerted effort to merchandize Multichannel and
now providing both DTS and DVD-A formats
ii. Sony, with arguably the Multichannel hit of the year in
"Dark Side of the Moon", among many other releases in
SACD.
iii. Telarc (which produces not only core classical but pop,
jazz & blues as well). Telarc releases in SACD
iv. WEA, who is aggressively repurposing it's deep pop
catalog for Multichannel as well as doing new releases,
exclusively in DVD-A.
v. Universal, who is currently releasing both SACD and
DVD-A.
vi. EMI Music/Canada, represented here, pursuing an
enhanced package to include DVD-A and DVD-V and
Red Book when this technology is feasible.
c. I'm going to say as little as possible about formats save this:
DTS & AC-3 decoders are available in the majority of existing
one hundred million DVD-V players. As far as SACD &
DVD-A capable players are concerned, there are far fewer
units, and the numbers for truly universal players like the
Denon and the Pioneer are even farther below that. It's hard to
know what the unit numbers really are between SACD &
DVD-A; both manufacturing camps make claims of between 4
and 5 million units, and outside estimates range from just below
2 million units. I'm just going to move on.
d. Popular music is perhaps one area where we've been over-anxious
for results, using traditional Pop CD sales numbers as
reference points. For instance, currently the top sales bracket
for a multichannel release seems to be in the range of 50,000 -
100, 000 disks, compared to where we try to focus on a pop hit,
which is more around 1-3 million units, depending on the
format. And caveats abound. For instance, "Dark Side Of The
Moon" has sold as many as 100,000 disks, but it's also a hybrid
release with a remastered Red Book layer accompanying the 5.1
and 2-channel SACD streams. By contrast, the Queen DVD-A
that Elliot Scheiner remixed last year has sold around 45,000
copies. Right now this seems to be the average number for a
modest Multichannel success.
4. Professional Electronics - in short...
a. Digital audio workstations are more and more multi-channel
capable.
b. Since Elmau several high-end Multichannel production tools
have become available including reverbs from TC Electronic,
Lexicon, Sony, and effects processors from Eventide among
others.
c. Unfortunately, new development seems to have slowed down
considerably. I've been told that sales of certain of these tools
have been less than encouraging, and there's a rumor that the
777 has been discontinued.
5. The Automobile
a. For some time now the mantra of Multichannel, such as it is,
has been, "We will see the breakthrough when Multichannel
systems get into cars." I'm sure many of us remember hearing
the Harman demonstration system in Elmau, and since then
several models have become available with surround sound
systems, notably the Volvo, and this year, even a Lincoln with a
THX sound system. Speaking for a group of music engineers
currently very active in 5.1 Multichannel mixing I'd like to say
that we're more than a little unhappy that so many current
offerings aren't discrete 5.1. We are told that it is because the
design cycle was begun before discrete hardware was available,
but one wonders why, when other major manufacturing
industries can turn around completely new designs in 2 years or
less, automobile manufacturers, another deeply challenged
industry, can't do better.
b. We are thrilled that Banff has a number of promising car demos
this year. We're anxious to hear them.
c. But the automobile has turned out to be a very special case;
there are significant challenges to delivering great audio in the
car, SUV or van, notably how to do deliver the best "Surround
Experience" to all of the passenger positions. There has been
considerable serious work by many in this area. And speaking
for that same group of 5.1 mix engineers I'd further like to add
that we take offense at the notion that automotive sound
designers insist that they are more qualified to judge what
constitutes an effective "Surround Experience" than we are,
especially as it ignores or corrupts the musical intention.
d. I'm told that most automotive manufactures are loathe to
implement a matched, full-range center speaker; we're told that
there's no room for it. We're told that automobile
manufacturers simply won't give a millimeter of dashboard
space or volume. I say we get them together with Don Henley,
or Graham Nash, or Sheryl Crow or whoever their most
venerated artist is to shame them into what they should be
doing.
e. Beyond that, the question is: whether we as mixers should
support or avoid that weak center speaker in our mixes. The
thinking I hear most often is that special cases should be
accommodated by special processing at the delivery end, and
that we shouldn't lower our standards.
6. May we take a moment to review multichannel in Video Games?
a. From the largest manufacturer of software in the US, I can tell
you that there is considerable progress in Multichannel, but I
must inform you of the following:
a. In video games, picture is everything. The audio stream gets,
on average, 5 to 6% of the player's bitstream capacity.
Anything that disturbs the video performance gets knocked
back. One Electonic Arts game, SSX Tricky, a motorcycle
game, used up to 10% of the bandwidth for sound, and was
widely panned for a "slow picture".
b. Nevertheless, many game boxes have Multichannel capabilities.
The X-Box can utilize Dolby AC-3 and ProLogic. The
PlayStation II does ProLogic and more recently DTS. Most
often 4.0 is implemented, sometimes 5.0.
c. Some PC native games implement 5 channel surround for
sound cards, notably the Creative Labs Soundblaster.
d. Most games utilize surround for dramatic effects. For the
recent "Lord Of The Rings" game, Electronic Arts were given
the 5.1 Music stems, but these were truncated to stereo.
Nevertheless, a few Playstation games have music going into
surrounds.
e. Finally, most often the sound channel is implemented as
22.05k, with compression of around 3.5:1.
|
|
So, looking over this, you could say we have a mixed bag of successes.
Incidentally, looking over this list of disciplines and the revenues associated
with each I'm drawn to make the informal, loose conclusion that, "the lower
the quality, the higher the revenue" and vice-versa. I mean, one video game
title alone, the John Madden Football franchise, is perhaps a 2 billion dollar
industry in and of itself.
Nonetheless, across all implementations of multichannel, there is a common
thread, which is the excitement of content creation. Forgive me for stating
the obvious once again: for those of us who make music - for those of us
who make mixes day-by-day for Multichannel - a door is opened by
multichannel, and it's a big door to new possibilities. And it's one that we
hate to have to pass back through. After you've had that immediate, visceral
response to the size and the spatial texture and the animation that's available,
after you hear that sense of time and place and story, there's simply no
going back. Not happily.
So, where to go technologically?
We certainly need to do more of what we've always done, which is:
continue to experiment with what sounds good, staying tuned into what else
is going on out there, and then to grow techniques to do even better work.
Oh, and make every effort to do great demos.
At the same time it would help greatly - and let me put it very simply - to not
repeat the same mistakes with the technology.
For instance: we know from our old experience in Quad that one of the
biggest problems was in the delivery format or formats. I believe that one of
the villains there was the matrixing that was forced on us - and we were so
easily persuaded to accept the compromise. Did we learn to work with it?
Yes, over the years a 4-2-4 matrix was used with success in cinema
surround. Was it right, assuming "right" means "sounds good"? Well, no, it
wasn't very good, but it moved us ahead.
I hope we're beyond that now. Simply put: discrete delivery channels best
preserve the intention of the artist, producer & mixer.
And time and time again you'll hear the same refrain from us about sizes of
the speakers. Not to sound like a skipping CD, but please let them all have
the same full-range performance. And about that pesky center speaker:
we're either going to use the center speaker and call this technology 5.1 or
we're not. I really like to use it, and others do, too, in a properly
implemented playback system. Maybe we should mix as if we'll eventually
solve site or system-specific problems downstream where they come up.
So, O.K, O.K., it comes to a go/no-go situation based on cost and we have to
compromise - and we always do - could we all at least compromise with
some reluctance? And when the productizers are deciding on our
compromises, could we please hear a little bit of what you have in mind?
And maybe even talk about it?
Now as far as the software, we have more work to do, because as good as
the classical multichannel demos are, we're still struggling to make pop
music as universally effective..
Here's what we content providers will do for you while we sort things out:
we'll make better recordings for less money. We'll be way less self-indulgent…
and we promise to use less vile, opportunistic compression and
fewer hideous-sounding plug-ins.
Maybe we'll even try to stop sneering at making accessible records.
But may we ask you to please understand what we do?
I would point you to Henry James, who affirmed, on behalf of his own art,
"We work in the dark." Multichannel, especially for pop, is still a new
paradigm. It has new rules. It's a new art. We don't know what it can be
yet. We just know it can be better.
Producers and mixers must truly think like innovators. We're going to have
to work with tools new to many of us and we're going to have use them to
charm artists into our dance. We have no choice; to craft the compelling
presentation of the future we will have to seriously woodshed with artists &
musicians to find a story to tell in a Multichannel context. And until it's
compelling, we have to keep trying.
Now, let me speak to equipment manufacturers about those "tools"; yes, we
know that we have to implement what you've already made for us. But in
pop, in particular, we are still lacking the ability to do some very basic, and
important, things.
O.K, we're trying to take more tracks live. But we can only go so far in a
business that's for many years been defined by the flexibility of working
with very many source tracks & time-displaced performances. What is so
convincing in the big ambiences - the big reverbs, the long, diffuse tails - in
classical presentations in Multichannel are for the most part utterly
unsuitable for pop work. It's time to move beyond the Schroeder reverb
work. It is more important than ever to be able to model small, real spaces,
however complex the processing.
Also notable is the lack of available tools to easily, convincingly animate
sound in 5.1 - to move a source around throughout a 2 dimensional space.
Guys, our fine researchers from Japan have demonstrated time and time
again that moving sources better define localization. You know, the
multichannel panner is the very first feature that the ingénue remixer reaches
for on his or her first Multichannel mix.
I'm happy to report that since Elmau Moore's Law is still applicable, so,
please don't whine about processing power or cost. Build us a useable
virtual room. Please.
We have great hope that in the future this can be done extremely effectively
in surround because we've now heard it done with Wave Front Synthesis. If
you haven't heard it, the demo is here in Banff and I urge you to take it in.
Let me take a giant step and say that in my mind technology such as this
(and you know, it's not new) is most likely to be the 15 - 30 year future of
Multichannel presentations, and hopefully much sooner. And also in
my
mind this clear vision of the future validates all of the other innovators who
are experimenting and commissioning systems with more and more
speakers, notably Tom Holman.
So we know where we're going, now let's be geniuses and try to figure out
how to get just a bit of this in 5.1.
Now, I know that someone out there has to buy this stuff...and that there's a
risk involved.
Speaking only for myself, I currently live in a region in the US where
producers are more or less loathe to spend money on recording technology,
and do so only when threatened. Even in the face of a business model that's
quickly and demonstrably crumbling, the priorities of most main-stream
Nashville producers are: 1> new cars (or new/old cars), 2> new houses, 3>
new motorcycles, 4> new boats and 5> time-share jets...more or less in that
order.
I've none of the above, and have instead invested in the technology. And to
know why I think it'll eventually be successful you only have to listenJ. For
instance, you have only to listen and compare the noisy and unpleasant 2-
channel mix of Faith Hill's last record to Elliot Scheiner's organic and
detailed Multichannel mix.
I have faith because, you know, lurking in the counter-culture where I live
are some of the finest musicians, writers & innovators in music today. They
are anxious and motivated to get to work making music again. And there is
no better time to experiment than now, while the record business is in a
meltdown and even at that are still making execrably bad records. We have
a bit of a breather - some room to experiment - better to find our own way
in the new paradigm. Recall that it was engineers like Bill Putnam who
were the pioneers in credibly advancing stereo in the early 60's, and that in
many instances he made his own, private stereo recordings during what were
to have been mono sessions. He did this at his own expense -and he did it
because it sounded good. Oh yeah, eventually record companies came back
to him to negotiate a fee for these stereo masters.
To summarize: pop music mixers are going to have to work harder,
cheaper...and more effectively. I would suggest that we consider in the very
near future making multichannel mixes a component part of the mixing
phase of each new project. Oh, by the way, keep in mind that no less than
Bob Clearmountain does this now - includes 5.1 mixes as a standard part of
his deliverables.
For Multichannel to move ahead in our individual careers if not our lives, we
have to have - and I'm sorry to have to return to this year after year - we
have to have faith that if we do great work we'll win the non-believers over
one-by-one. There are going to be delivery channels for what we do in
Multichannel Audio. We have to have truly impressive work ready to go by
the time the opportunity comes along. Not vice-versa. Let's make
that the
reality.
I certainly have many people to thank for coming forward with ideas and numbers:
among these the contributions of Goeff Martin, Jason Corey, Bob Woods,
Elliot Scheiner, Stuart Bruce of Real World, Fraser Hill of EMI Music
Canada, Jeff Levison & Jeff Skilling of DTS, James Guthrie, Gary Rydstrom
& Leslie Ann Jones of Skywalker Sound, Laurie Fincham of THX, David
Smith and David Kawakami of Sony, Murray Allen of Electronic Arts, Phil
Ramone, Ed Cherney, Bruce Botnick, Dennis Sands, Frank Wolf, Tony
Faulkner and Mike Novitch of Warner Hollywood. Special thanks as well to
Theresa Leonard and John Sorensen for assembling great demonstration
systems.
|
148870
website visitors to this page.
|