IBC
As live production moves to the cloud, can SMPTE Standard 2110 evolve to retain broadcast quality standards or should the industry wholeheartedly embrace web technologies?
https://www.ibc.org/trends/smpte-2110-is-it-fit-for-the-future-of-broadcast/7301.article
Broadcast production is at a crossroads and CTOs have a decision to make: should their new studio or mobile facility be built using SMPTE standard 2110 or something that might be more suitable for the cloud computing age?
It’s a problem that has surfaced in recent months as
large-scale live production — the area of premium broadcast programming for
which 2110 was principally designed — has shut down or reverted to using less
conventional technologies to keep on air.
For many, ST 2110 still represents the bedrock of
professional production and a relatively risk-free way to segue the industry’s
legacy SDI base into IP. Others see an existential crisis in which broadcast
engineering-based standards are a cul-de-sac and that if traditional players
are ever to innovate on par with internet-first streamers they need to change
the narrative.
“Broadcast TV must adapt to online entertainment formats
faster than the online entertainment formats make broadcast irrelevant,” says
Johan Bolin, Edgeware’s chief product and technology officer.
“A growing number of broadcasters are asking themselves do
they really need to continue building the broadcast stack along conventional
lines or is now the time to embrace web and other technologies more generally
in production.”
For many the issue boils down to the engineering mindset. If
your starting point is to build a perfect pipeline where all the important
performance indicators like frame sync are under full control and can be
guaranteed, then this will inevitably fail when working in the cloud.
A STOP2110 website is blunt. It lambasts the standard as a
“train wreck”, “worse than SDI” and “old school hardware engineering combined
with design by committee.” The site doesn’t suggest any genuine alternative and
its author lacks the courage to go public, but it begs the question – why has a
dry international specification drawn such ire?
The value of ST 2110
From the analogue era through SDI the industry has used baseband video simply
because it’s the highest quality. Without processing, uncompressed media also
offers the lowest latency because in live you want to interact with your studio
and your performers.
When it came to devising a means to migrate the industry
into IP, these fundamentals were sensibly maintained. Standard 2110, for which
SMPTE and co-developers VSF, EBU and AMWA have been awarded a Technical Emmy,
reinvents SDI by providing for uncompressed video and precision timing.
It differs significantly from SDI in splitting audio, video
and metadata into separate streams (or essences). Instead of having to worry
about running the correct type of cable and signal to various locations,
broadcasters have far greater versatility to be a responsive studio business.
“The idea that you could separate the flexibility that you
needed from the cabling you laid down was considered a goal worth achieving,”
says Bruce Devlin SMPTE’s Standards Vice President.
Since standardisation in 2017, ST 2110 interfaces have been
added to core equipment from cameras to multiviewers enabling broadcasters to
build new facilities entirely in IP. BBC Wales’ new facility in Cardiff, is one
example.
It achieved its aim of unifying the industry around a common
suite of IP specs and allows broadcasters to migrate as fast as their
investment allows by keeping one foot in the SDI camp.
However, the rocketing rise of OTT streaming and the advance
of cloud computing exacerbated by Covid-19, has put the future of 2110 under
scrutiny —even at SMPTE itself.
“It is not really that 2110 is the wrong standard, it’s that
the means of content consumption has started to change rapidly,” Devlin says.
“The global pandemic accelerated this when live sports and stage events, all
the stuff that 2110 is dedicated to, almost vanished overnight.”
Cloud-based workstations using PCoIP, and low-cost low
bandwidth video transmission has become the norm. Business teleconferencing
tools, smartphone cameras and webcams are in routine use in at home scenarios
for both production crew and on-air talent. ST 2110 was not designed for this.
What’s more, the audience has begun to accept what the IABM
calls ‘Covid Quality.’
“The use of off-the-shelf collaboration tools may not be
ideal, but it keeps the media factories running,” it finds. “Audiences started
to accept glitches, streaming issues and for that matter more often than not
poor video and audio quality; our expectations for more 4K UHD in 2020 turned
into Covid Quality.”
PTP meets floppy timing
It’s not as if things will go back to normal when the pandemic passes. Remote
production links contributed over the internet were advancing anyway. Now they
are entrenched. Cloud computing and cloud services are becoming ubiquitous.
“We’re having to find ways to use the 2110 ecosystem to
connect nano-second accurate studio environments with remote operations over
the internet or in the cloud where floppy timing exists,” Devlin says.
The Joint Taskforce on Network Media (JT-NM) which
coordinates SMPTE 2110 and the wider development of a packet-based network
infrastructure, is investigating ways to connect the Wide Area Network of a
production plant with tools, applications and facilities outside of the studio.
However, current cloud connections are not up to the quality
standards required for low latency live streaming media. Therefore, SMPTE says
research into quality-enhancing technologies, such as retransmission or
Automatic Repeat reQuest (ARQ), is crucial to improving the network
infrastructure required to deliver broadcast-quality transmissions.
“The JTNM say we still need 2110 accuracy within a facility
but we don’t necessarily need 2110 perfection between two facilities or between
an OB truck and a facility,” says Devlin.
“It’s finding a way to take the gold-plated excellence of
2110 together with parts of the ecosystem which are less gold plated and using
them both to produce better in a Covid world.”
One option is to compress media to get it to and from the
main production plant or into and out of cloud. The leading scheme is ISO
standard JPEG XS, a mezzanine compression that squeezes the bits sufficiently
to save on bandwidth but not hard enough to destroy the quality needed for
manipulation, like chroma keying, in production. Crucially for live production,
JPEG XS exhibits extremely low latency. It is already mapped into the 2110-22
ecosystem and products are launching with JPEG XS capability.
The BBC also expects ‘hybrid’ architectures to evolve, and
its R&D team is looking at how to ensure interoperability. In a blogpost,
it says: “ST 2110 isn’t naturally suited to deployment in a cloud, so we expect
‘hybrid’ architectures to evolve, and will be looking at how to ensure
interoperability in these. This is likely to include work on ensuring that
media identity and timing information is preserved, including where we need to
go through compressed channels, such as for contribution from remote studio.”
ProAV interoperability
Also in the works is a proposal to standardise the interoperability of products
within the Pro AV sphere. The Internet Protocol Media Experience (IPMX) would
encompass many technologies being used by at-home productions such as robotic
PTZ cams and web-cams as well video conference codecs.
IPMX is based on 2110 and promoted by the Alliance for IP
Media Solutions (Aims), which is chief cheerleader for ST 2110 in broadcast.
This makes sense since, according to Aims, a quarter of its members sell into
both broadcast and AV markets.
The move also recognises both that AV and broadcast are
undergoing a transition to IP. The benefits are similar for both industries
such as bi-directional cables and reduced space. The gear used to produce and
distribute content for giant screens at music venues, for digital signage or
esports events is also sold into broadcast. And in many cases the quality of AV
content exceeds that of broadcast.
The elephant in the room when talking to SMPTE, Aims and
ST2110 supporting vendors like Imagine and Sony is the widely used video over
IP transport scheme NDI. Developed by NewTek and owned by Vizrt, NDI is a live
production protocol considered a non-starter by backers of 2110 because its
heavy compression is considered unsuitable for broadcast and its proprietary
nature incompatible with open standards.
Innovating production
However, these arguments are precursors to the wider challenge of evolving
production to deliver truly personalised, interactive media.
This is generally considered the future of ‘TV’ and is
tantalisingly in reach thanks to high-speed high bandwidth technologies like
5G. In comparison, the production of content itself remains in the dark ages
and ST 2110 is considered by some to be part of the problem.
“Fundamentally, if TV is to transform it must overcome the
brick wall between production and distribution,” says Bolin. “These two domains
are separated and 2110 is not the solution.”
He argues the while the upstream process in TV all about
creating content the downstream process attempts to reach as wide an audience
as possible, whether through satellite, cable or DTT and now the internet.
“Upstream has worked with the same production processes and
tech stacks for five decades but the growth of the internet has forced
broadcasters to increasingly work with internet-based technologies downstream.
“Yet it is incredibly difficult today for viewers to
contribute video upstream. This is by design. It is not a consequence of the
technology. It is how we have designed the technology.”
Bolin says he wants to see technology that “not only allows”
but “encourages the industry to mix and blend downstream and upstream
processes” to enable TV formats more tailored to the viewer or concepts that
allow viewers to contribute to the live programme.
BBC in the cloud
These are not the thoughts of one maverick vendor. The BBC is thinking along
identical lines.
Having started to use IP in production centres like BBC
Wales fitted with 2110 it now says, “the content-making capacity and
equipment in these facilities is still mostly fixed during the design and
fitout stages, meaning large changes can only occur during a re-fit. The
business operating model of a current generation IP facility is also fairly
inflexible, with large capital expenditure required upfront.”
Those are alarming statements given that they could equally
apply to SDI, the prison from which ST 2110 promised escape. Content still has
to be created using traditional broadcast equipment in a physical production
facility.
BBC R&D is therefore investigating how it can apply the
cloud computing technologies which run iPlayer to its production operations.
“R&D are working with colleagues from around the BBC to
join up these two areas, enabling broadcast centre-style production operations
to occur within a software-defined cloud environment,” it states.
“We think the benefits of this will be huge, making our
physical IP facilities even more flexible, and enabling us to deploy fully
virtualised production systems on demand. Ultimately, this will help the BBC
make more efficient use of resources and deliver more content to audiences.”
Edgeware is making similar explorations. “The idea is to
take web-based technologies and the tech stacks and concepts from games
development and esports and incorporate those into the TV stack,” says Bolin.
“Broadcast has always been about guaranteed bitrates and
guarantee framerates and guaranteed no drift in time. Video on the internet is
about accepting its imperfections, accepting that you will have drift and you
will have a problem guaranteeing perfect bit rates. The onus is on the industry
to build solutions that mitigate these imperfections.”
No matter how perfect the upstream there will always be
imperfection in the downstream. That’s true with SDI or 2110 since the source
is always degraded in some form during distribution. Bolin says the industry
should prioritise innovation in production and work with internet’s concept of
best effort distribution.
Indeed, there are a number of protocols for smoothing loss,
jitter and latency such as MPEG DASH, RIST and SRT which do mitigate the
internet’s deficiencies.
“We should facilitate innovation rather than seek
perfection,” he says. “Today’s best effort is pretty darn good.”
