GOING LIVE & REMOTE

LIVE REMOTE
PRODUCTION
April 2021

Enabled by
FOREWORD

Sometimes it takes unexpected events to make us look at life a little differently.

Just over a year ago we were all getting on with business as usual – preparing for
trade shows, planning product launches and discussing new projects with customers,
all blissfully unaware of the calamity that was about to knock everything sideways. In
the early days of the pandemic, the media production industry’s gaze turned almost
wholesale towards solutions and tools that would enable remote collaboration
and production. I think we all knew that remote production was going to become
the norm, but we imagined the adoption and transition as a gradual, orderly and
considered process rather than the mad scramble we actually experienced. Now,
as we slowly emerge from the grip of the pandemic, it’s clear that it has acted as a
catalyst for remote production principles and cloud based solutions. They do say
that necessity is the mother of invention, and the last twelve months have seen
the emergence of innovative new tools to enable collaboration and contribution to
continue, with low latency and high production values very much at their heart. If
there is one silver lining to take from this dreadful experience, it will be that
resilience and innovation. After all, the show must go on…

David Ross
CEO, Ross Video

APRIL 2021

LIVE REMOTE PRODUCTION 2

INTRODUCTION

For decades live content production was all about tonnes of kit, driven from place to
place by teams of people who were dedicated to operating that kit.

Tapes turned to files, hardware to software, on premise to cloud; but the OB truck
stood firm.

Tapes turned to files, hardware to software,

on premise to cloud; but the OB truck stood
firm. Until now.

The high data rates and real-time requirements of broadcast quality live video meant
that dedicated equipment still dominated. Bespoke processing hardware was tied
to specialist interfaces and control surfaces, acting almost as an extension of their
operators. Everything was designed to give production crew instantaneous control of
fast moving live content.

It’s no wonder that most live productions have operated at a single location, whether
studio or outside broadcast.

But in the past few years, that’s all begun to change. Computer processing power is
now sufficient to effectively process high quality video and audio in real time. High
end IP networks are now more than capable of transiting large numbers of high bitrate
video streams from location to location. Production teams can now be split across
sites, or even working at home.

The age of remote live production has well and truly begun.

Remote production is now becoming the norm for even the highest value, highest
profile live programmes.

But that is not to say that it is without challenges. Considerations of cost, connectivity,
and workflow are paramount. And indeed, there is no singular workflow or
architecture for ‘remote production’.

LIVE REMOTE PRODUCTION 3

 There is no singular workflow or
architecture for ‘remote production’.

There are tens or hundreds, if not thousands of different models for producing live
content with some level of remote or distributed operation.

So how mature is the state of remote live production? What improvements still
need to be made? And can we even reach a common understanding of what remote
production means?

The DPP has worked with experts from 44 different companies to shed light on one of
the most fundamental and important changes taking place in our industry today.

LIVE REMOTE PRODUCTION 4

EXECUTIVE SUMMARY

While it is true to say that remote live production has grown a great deal, there is still
a huge difference in technology and workflow between a major sporting event being
produced from the broadcaster’s headquarters, and an episode of a chat show in
which the guests are filmed at home via video conferencing.

Perhaps this is why there is no common understanding or language when it comes to

describing different models of remote live production.

So for the first time, the DPP has identified five clearly defined models of live production.

For the first time, the DPP has identified five

clearly defined models of live production.

Each has its own advantages and disadvantages, and its own specific technical and
operational considerations. By laying them out in plain English, we aim to arm producers
with the knowledge needed to choose the most appropriate model every time.

On location

The traditional model in which content

acquisition, signal processing, and production control are colocated; whether at a
studio complex or an outside broadcast.

It is the baseline by which other models are judged, using tried and tested workflows
that enable production teams to work closely together and maximise creativity.

Remote controlled

Remote controlled production moves the

control surfaces and operators to an offsite location such as the broadcaster’s
headquarters, while keeping processing at the acquisition site.

This approach has been popular during the pandemic to improve social distancing,
but also has broader applications in reducing crew travel, while making use of existing
outside broadcast equipment and infrastructure.

LIVE REMOTE PRODUCTION 5

Centralised

This model moves the signal processing and

production control to a central hub location, such as the broadcast centre.

A great deal of bandwidth is required to transport all the audio and video feeds from
the acquisition site, but the model delivers higher utilisation of equipment and crew,
lower capital investment, and lower carbon footprint. As a result, it is one of the two
models predicted to grow most dramatically in the next two years.

Distributed

This architecture also usually involves

centralised signal processing, but offers additional flexibility by enabling crew to
remotely control the production from multiple production hubs, or even from home.

It adds complexity to the connectivity provision, but in return it offers the potential to
reduce travel even further, while opening up access to the best talent no matter where
they are.

Cloud

The final option locates some or all of the

signal processing in the cloud, controlled remotely from production hubs or from home.

It offers similar benefits to distributed production, with additional flexibility and

scalability, and lower capital costs. As the capability of software production tools
improves rapidly, cloud production will join centralised production as a dominant model.

LIVE REMOTE PRODUCTION 6

CONTRIBUTORS

The content for this report has been gathered through a series of workshops and
interviews with over 60 subject matter experts from across the industry, along with
our expert sponsors, Atos, Ross Video, Telstra, and Kiswe.

Although the content of this report has been informed by those discussions, it should
not be assumed that every contributor shares all the views presented here.

Ed Abis Michael Clayton

General Manager, Business Development Executive,
[Link] BASE Media Cloud

Chris Ainger Charlie Cope

Head of Technical Sales, Technical Executive BBC Sport,
NEP Connect BBC Sport

Stuart Almond Tom Copeland

Industry Lead – Media & Telco, Head of Broadcast Technology and
Microsoft Technical Operations, Sunset + Vine

Roeland Awick Raoul Cospen

Project Manager, Director of Business Development,
RTL SaaS & Cloud, Dalet

Rich Baker Adam Dean

Senior Broadcast & Multimedia Operations Executive,
Manager, Formula E BBC Sport

Tim Baldwin Abi Dormieux

Head of Product, Managing Director,
Zixi Jackshoot

Narinder Ball Duke Duong

Outside Broadcast Operations Sr. Global Solutions Architect – Telco,
Executive, BBC Sport Media & Entertainment Specialist, Equinix

Hugo Bastos Troy English

Solutions Architect, CTO,
Ross Video Ross Video

Scott Bounds Dan Epstein

Principle Cloud Solution Architect, Senior Product Manager,
Microsoft Haivision

Richard Buchanan JJ Eynon

Global Media & Entertainment Engineering Manager – CNN,
Principle, Verizon WarnerMedia

LIVE REMOTE PRODUCTION 7

Larry Gale Joe LaMonica
Head of Production Network, Senior Director, Broadcast Engineering
Push Live & Operations, VICE Media Group

Bevan Gibson Benjamin Lardinoit

CTO, CEO & Co-Founder,
Euro Media Group OnHertz

Pete Griffiths Brian Leonard

CEO, Head of Engineering: Post & Workflows,
Idonix IMG Studios

Tim Guilder Arne Lievens

Technology Manager, European Ops,
ITV Kiswe

Simon Harrison Anna Lockwood

Head of Product, Head of Sales – Telstra Broadcast
M2A Media Services,Telstra

Gareth Havard Bruce MacGregor

Chief Engineer, Senior Project Manager & Consultant,
Tinopolis Atos

Buzz Hays James McAlister

Global Lead – M&E Content Live Workflow Manager,
Creation Solutions, Google DAZN

Simon Haywood Kevin McCue

Founding Director, Director of Technical Operations,
Jackshoot Sky Sports

Wendy Hendrickx Thomas Menguy

Head of TV Production, CTO & Co-Founder,
Formula 1 Wildmoka

Saskia Hook Rob Milchem

TVOB Product Manager, Media Project Manager,
BT Atos

Brian Jones Paul Moore

Senior Director, Media Production Head of Global Media Business
Operations, PBS Development, Atos

Anshul Kapoor Liam Murray

Head of Media Broadcasting Manager, Media Management,
Solutions, Google Cloud, Google Broadcast & Media Services, RTE

Rostam Kilgour Esben Nielsen

MCR Manager, Head of Broadcast,
RTE JTGlobal

Steve Kruger George Patchett

Head of Technology – Deputy Production Manager,
Entertainment, ITV Studios Promod esports

LIVE REMOTE PRODUCTION 8

Arran Paul Steve Teague
Location Engineer, Head of Technical Services,
BBC ITV News

Sean Pickering Rhian Thomas

Account Director, Managing Director,
Telia Tinopolis Cymru

Dan Pope Trevor Turner

Founder, Head of Media Systems Development,
Cingularity Formula 1

Andy Rayner Willem Vermost

Chief Technologist, Design & Engineering Manager,
Nevion VRT

Marc Segar Mun Wai Kong

Director of Technology, CTO,
NEP Grabyo

Corey Smith Matthew Webster

Director, Live Operations, Broadcast Technical Specialist,
Activision Blizzard Ross Video

Kelly Taylor
Technical Supervisor,
Sky

LIVE REMOTE PRODUCTION 9

REMOTE PRODUCTION
GOES MAINSTREAM

The trend towards remote production of live content is not new; some of the most
prominent examples are four or more years old. And the truly pioneering live remote
productions took place almost a decade ago. However, it is undeniable that the
pandemic in 2020 accelerated the trend, and moved it into the mainstream.

As Greg Burns of Arqiva put it, “Normally, live production is a logistical triumph of
in-person labour and a whole host of resources. This year we’ve discovered that if it’s
not feasible for teams to physically work alongside one another, it opens the door to a
whole array of new options.” [TVBEurope]

This trend was seen not just in the outputs on screen, but in the dynamics of the
media technology market. A number of suppliers formed strategic partnerships in
order to deliver compelling remote production solutions to eager customers. These
were significant moves from influential companies, with partnerships between the
likes of Avid and Haivision, LiveU and Grabyo, Zixi and Google Cloud.

Premium live sports events have been looking to undertake more remote working
for some time. Tournaments that involve a global circuit of events often move
large volumes of equipment and vast numbers of people from city to city, country
to country, or continent to continent. The potential savings in cost, time, and
environmental impact are therefore huge.

But COVID-19 meant that whole new genres, production teams, and content
companies got their first taste of remote production. According to Haivision’s
Broadcast IP Transformation report, 71% of those surveyed said COVID-19 had
pushed the move to IP in their organisation, while 39% are employing remote
production workflows already. [Haivision]

COVID-19 meant that whole new genres,

production teams, and content companies
got their first taste of remote production

LIVE REMOTE PRODUCTION 10

 What is remote production, anyway?

One of the greatest challenges of remote production is simply

understanding what others mean when they say remote production!
The terms are numerous – REMI, Home Run, At-Home, Centralised,
Hub & Spoke – and the definitions of these terms are rarely clear.

In this report, we use remote production as an umbrella term, meaning

any model in which some of the main production infrastructure,
functions, and/or crew roles are performed away from the primary
acquisition site. The simple term remote production is never used to refer
to a specific production architecture.

In later sections, we will explore this problem in detail, and define the
key models of remote production in use today.

Entertaining from a distance

Studio entertainment generally relies on the energy of the host, participants, and
audience. There’s an immediacy and atmosphere that thrives on the colocation
people, and so it is natural to have the production crew on site too. Before the
pandemic, it would be rare for such shows to be produced remotely.

Yet in 2020, Drew Barrymore’s live chat show The Drew Barrymore Show devised
ways to interview guests on the other side of the country from the main studio. They
used a green screen studio in Los Angeles to film west coast guests as they were
interviewed via video link by Barrymore, who herself was in New York. The green
screen setup enabled guests to be superimposed into the show’s studio, while the use
of remote controlled cameras meant that very few crew needed to be present in Los
Angeles. [ViacomCBS]

Meanwhile, Keeping up With the Kardashians, American Idol, The Tonight Show and
more have all been filmed remotely during quarantine, with stars sent equipment to
their homes and instructed over video conferencing. [Buzzfeed News]

In China, similar processes have been used to produce mainstream TV, with shows
like Singer having contestants and judges connect via video link from home. But while
these are temporary measures to cope with the pandemic, a new genre of online
video has emerged that is more likely to endure.

‘Cloud reality shows’ enable anyone with a mobile phone and a network connection
to be part of a show, and as discussed by Guangzhou Daily, they have struck a chord
with Chinese audiences. [BBC]

LIVE REMOTE PRODUCTION 11

iQiyi had a hit with Karaoke at Home, in which social media users send requests for
singers to perform songs of their choice, from their own home. And users of video
sharing site Youku have enjoyed cookery programme Eat Well, which takes viewers
inside the home of a different guest each day, to share their preparation of a home
cooked meal.

As the pandemic recedes, mainstream studio entertainment shows are of course

returning to having presenters in the studio. Few would suggest that having a guest on
video link from home is as good as having them in a professionally equipped studio.
But techniques such as those deployed by CBS for The Drew Barrymore Show may
well be retained for the long term as a way to reduce costs and environmental impact
of flying guests across the country. And online live shows built around audience
participation are demonstrating the new creative possibilities enabled by a cloud
native video production toolset that’s accessible to everyone.

Breaking news without breaking quarantine

Remote contributors are of course central to news broadcasting. Bulletins are often
built around such contributions from correspondents and guests around the country,
or the world. But most programmes are traditionally presented and produced from a
single studio location.

The exception would usually be major events coverage. A UK news producer like ITN
would historically have produced coverage of a major event such as a US election as
a full outside broadcast, with production taking place on site in Washington. But for
2020’s US election, camera feeds were backhauled from one capital city to another,
enabling programmes to be mixed in London.

For 2020’s US election, camera feeds were

backhauled from one capital city to another,
enabling programmes to be mixed in London

This allowed the organisation’s broadcaster clients – including ITV, Channel 4, and
Channel 5, to have an on site presence in Washington without having to fly entire
production crews across the Atlantic. [TVB Europe]

Of course the pandemic also created more extreme requirements for remote working,
with broadcasters like Sky News having to anchor whole shows from home during the
most severe lockdown restrictions. They created remote presenter kits with cameras,
tablet based teleprompters, high quality return video monitoring, and background
screens that could be driven remotely.

LIVE REMOTE PRODUCTION 12

While news anchors may not work from home in the long term, some production
crew might. In January 2021, ITN shared a three stage plan for increasing remote
working of their staff. It begins over the first quarter of the year with roles such
as production assistants who don’t require specialist control surfaces, and can
therefore be given relatively simple equipment kits to work remotely. This first stage
nonetheless involves a significant workflow shift, with key foundational technologies
such as intercom now required to work seamlessly across locations.

The plan then expands during 2021 to cover roles such as graphics operators who
have simpler control surfaces, or who can use software control surfaces more easily.
Phase three will target remaining roles such as vision mixers and sound operators.
[IBC]

Sports off site

If there’s one genre of programming that stands to benefit the most from remote
production, it’s sport. As tournaments travel the world, the costs of travel and freight
can be vast, while the need to move equipment and people from place to place can
create editorial constraints.

The one genre of programming that stands to

benefit the most from remote production is sport

Motorsports series Formula E has found that off-site multi-feed production enables
much greater flexibility in the editorial and commercial offerings they can provide to
broadcast partners. A single track feed mixed on site can be reversioned remotely,
mixing in different camera feeds, graphics, and commentary.

Formula 1 has been taking steps into remote production for some years, in an effort
to reduce the 200+ people and 160+ tonnes of freight sent to each race. Digital and
social content has been produced in a Remote Operations Centre (ROC) that has
been operational since 2014, but plans have been accelerated by the pandemic.

Their latest installation allows them to reduce freight shipping by 34%, and the
number of travelling staff by 37%. [Formula 1]

90 camera feeds and 170 audio sources are brought into their Master Control Room,
where they can be curated and produced by 53 operators all working in the Remote
Technical Centre, not at the race track. Other remote functions include camera
racking, data management, graphics, and media management.

LIVE REMOTE PRODUCTION 13

But it’s not just motorsports, nor even just the highest profile and highest budget
competitions that are affected. Sky Sports deployed remotely operated cameras
extensively to cover the 2020/21 PDC World Darts Championship, reducing the
number of camera operators from 16 to 7, with only one manually operated camera in
the venue. [IBC] And Animation Research, which provides their Virtual Eye graphics
to events ranging from PGA golf to America’s Cup sailing, moved their operation to
a cloud based production environment in 2020. This enabled them to cover events
around the world from their base in New Zealand. [IBC]

Beyond television

There’s also been an explosion of other types of live video. These include user generated
and social content, corporate, education, governmental, and houses of worship.

Online game streaming is one particular area that has grown significantly. Between
April 2019 and April 2020, there was a 101% growth in Twitch streaming, 65% growth
in YouTube Gaming, and 238% growth in Facebook Gaming. [StreamElements]
Although some might consider this content quite different from traditional television,
the elements of production are in fact very similar. The growth of software tools for
vision mixing, audio mixing, graphics and more, mean that the technology used for
different genres of live content is converging.

These topics will be explored further in an upcoming DPP report, The Business of Live.

The technology used for different genres

of live content is converging

Development continues

There is still considerable work happening to improve the technical and workflow
capabilities of remote production.

Major recent development tests include those run by the European Broadcasting
Union (EBU), that has collaborated with member organisations to air a biathlon
event using a cloud live production proof of concept, with just one engineer on site.
[TVBEurope] And new connectivity options will only accelerate developments, as
seen by 5G production trials from the likes of Sky Deutschland, that has tested the
technology for transmitting live camera feeds at Bundesliga. [Rapid TV News]

The growth of IP video production, as broadcasters move from SDI to protocols such
as SMPTE ST 2110, lays the foundations for more flexible production models. And the
increasing use of software tools allow more aspects of the production to be virtualised
and run anywhere. Although some steps taken during the pandemic will be rolled
back to more traditional operations in the coming months, there is no doubt that the
future is bright for live remote production.

LIVE REMOTE PRODUCTION 14

THE ELEMENTS OF
LIVE PRODUCTION

In order to explore the different architectures of live remote production, we must first
understand the key functions involved in live production generally. This invites the
question of which functions could be performed in which locations.

Through a series of workshops with our expert contributors, we were able to define
and agree upon a series of high-level functional elements. While the list is not
necessarily exhaustive, it provides a reference for the different functions that could be
deployed in different locations depending on the chosen production architecture.

On-screen & acquisition

The first thing to consider is the main content of the show. In a sports scenario this
might begin with the action at the stadium or race track; for an entertainment or news
programme it might be the action in the studio. But there is also a high likelihood that
there will be other sources coming from other locations.

Presenters, guests, commentators, or other contributors might be in the primary

production location, or they may be in a different studio or location. Increasingly, they
may be at home.

In each case, we must consider microphones and cameras. There is particular

diversity in the latter, which might consist of manually operated studio cameras,
fixed rigs, or simply webcams. They could also be robotic or PTZ (Pan/Tilt/Zoom)
cameras with remote control capabilities. In the case of professional cameras, many
of the functions (such as iris, colour balance, and so on) can be controlled separately
through a Camera Control Unit (CCU). These functions could be placed local to the
camera, or operated remotely.

For some genres, there are other types of content capture too. Data feeds are
increasingly important to many sports, while esports relies on gameplay feeds from
observer clients, which act as virtual cameras in the game environment.

Content doesn’t only come from the acquisition locations, of course. We must also
consider content being returned to these locations. These might include camera
tally systems, teleprompters for presenters, or video feeds for in-vision displays. In
almost all cases, a confidence feed or programme output will be provided back to
the presenter/contributor. They may also be provided with other video feeds; a sports
commentator may be supplied with multiple video feeds of the action, for example.

LIVE REMOTE PRODUCTION 15

Finally, the relationship with the audience is no longer always one-way. Depending
on the production, audience members could participate via video link – essentially
becoming a remote video contributor – or they could provide input via social media that
is ingested to create on-screen graphics. Fan engagement is an increasingly important
aspect of many live events, and it can be a significant part of the production. A number
of sports and music events have deployed in-vision video walls of remote audience
members, adding tens or hundreds of additional video feeds to the production.

Production & editorial

The individuals on screen are of course working hand in hand with individuals off
screen. Roles such as the director and producer are central, with others determined
by the type of production.

Depending on their location, each of the production crew may require video
monitoring such as multiviewers in order to perform their roles. They may also need
access to production planning or rundown tools, such as a newsroom computer
system (NRCS).

The production team and on-screen contributors need to be in communication

throughout any live broadcast, and especially when distributed across multiple
locations, they will rely heavily on intercom. Our experts were very clear that getting
intercom right is central to any successful remote production.

Getting intercom right is central to any

successful remote production

Content processing

Once captured by a camera, microphone, or other device, content is processed

and curated to create the programme output. Central to this are vision mixing and
audio mixing.

In addition, pre-recorded content may be inserted into the programme via VT/
playback, while content from the live programme may be reviewed and reused with
replay systems. There may also be editing required to create highlights packages or
other inserts, while some type of graphics (including overlays, full-frame 2D and 3D
graphics, or augmented reality) is crucial to most types of live content.

The programme may further be augmented with access services such as subtitles
or sign language. And graphics, subtitles, commentary and other elements may
be replaced for different territories in a localisation process. Finally there will be
distribution of the main programme output.

LIVE REMOTE PRODUCTION 16

 Processing vs control

Significantly, each of these functions can be divided into two component

parts: the signal processing, and the control system. For example, vision
mixing is performed by a person interacting with a control surface,
which controls the actual signal processing.

Generally speaking, most modern production tools actually separate

these functions even in traditional single-site production scenarios;
the control surface sends messages to the signal processing hardware
which is situated in a rack elsewhere in the OB truck, or in an apparatus
room. This logical distinction between control and processing will be the
foundation of the different remote production architectures discussed
later in this report.

Production locations

Having defined the primary functions involved in creating a live production, we can
consider the different locations in which each of these functions might take place. Our
experts considered a generalised case, and agreed on the following broad definitions
of the key locations.

3 Acquisition site

The location at which the on-screen action happens. This could be the venue, stadium
or racetrack; the studio or outside broadcast location. It is the place where cameras
and microphones acquire the media for the programme.

Of course there may be multiple acquisition sites, such as a stadium and separate
presentation studio, or multiple news reporters in different locations.

3 Production hub

The production hub is a location separate from the acquisition site, at which there is
signal processing and/or control surfaces. This is often the production company or
broadcaster’s headquarters.

The production hub can include data centre space, production areas, or both. There
are many names used for such sites, including the broadcast centre, control centre,
and remote operations centre (ROC).

LIVE REMOTE PRODUCTION 17

 Some broadcasters have built multiple
production hubs in different locations,
allowing crew spread across a country to
work on programmes without travelling

Again there may be multiple production hubs involved in a given production, allowing
processing or crew to be split across multiple locations. Some broadcasters have
built multiple production hubs in different locations, allowing crew spread across a
country to work on programmes without travelling to a central location.

3 Cloud

The cloud is an increasingly important part of the modern production landscape, with
a growing number of live remote productions making at least some use of the cloud.

For the purposes of this report, we take a very broad definition of cloud. While
public cloud is the main focus, third party data centres or colocation sites could be
substituted for the ‘cloud’ block in the production models below.

It is also important to understand that even within a single public cloud provider, the
cloud is not a singular entity or location. Infrastructure and processing may be spread
across multiple regions, or – increasingly – deployed at the network edge. This can
help with reducing latency involved in transmitting signals over long distances.

3 Homes

Finally, individuals involved in the production may be based at home.

During the pandemic, presenters and other talent have been working from home,
often using professional equipment supplied by the broadcaster or production
company. More generally, other contributors such as interviewees may join by video
conferencing using webcams. In both of these scenarios, the home essentially acts as
an additional acquisition site.

Production crew may also be able to perform their roles from home using remote or
software control surfaces. In these cases, their homes are acting more similarly to a
production hub.

Finally, the audience are of course likely to be at home. Depending on the playback
device, the player could be overlaying graphics or performing other on-device
rendering. And as well as viewing the content, the audience could be interacting
directly; joining the programme through video conferencing as a contributor, or
providing input that is included in the programme such as comments, social posts, or
virtual applause.

LIVE REMOTE PRODUCTION 18

FIVE ARCHITECTURAL
MODELS

One of the key difficulties in discussing remote production is that there are so many
different possible architectural models, making it hard to succinctly describe one
versus another. However, our expert contributors were able to agree on five high-level
architectures, which outline the most significant models.

In each case, the capture equipment, signal processing, and control surfaces are
assigned to one or more locations.

CAPTURE EQUIPMENT SIGNAL PROCESSING CONTROL SURFACES

Each individual production will have its own nuance, variations, and complexities.
But almost any live production can be described as broadly conforming to one of
these models.

LIVE REMOTE PRODUCTION 19

 1 On Location

3 Capture, processing, and control are in one location – whether a studio

complex or an outside broadcast

3 This model offers creative flexibility by ensuring easy communications

between all members of the production team

3 But it is costly, and in many cases environmentally unsustainable; its use

is therefore predicted to decrease

The first model is not a remote

production at all, but a traditional REMOTE ACQUISITION/
CONTRIBUTORS
single-site production. In this case,
the capture equipment, signal
processing, and control surfaces are
AV signals Return monitoring
all positioned at the acquisition site,
with the programme output being
originated from there. ACQUISITION SITE
This model describes a traditional
outside broadcast, where the OB truck
is the main production hub, housing
the processing and the production
crew. It also describes a traditional Programme
output
studio show, in which the studio,
control room, and apparatus are all
within the same building.

The On Location model does not preclude additional remote contributions via either
dedicated professional video link, or video conferencing. However, unlike remote
production models, the signal processing and control are retained at the primary
acquisition site.

Tried and tested

Our experts told us that this model continues to be valuable for a number of reasons.
These include familiarity; not to be underestimated when crews have to work together
in real time, making instinctive decisions as they go. Such familiarity may also lead
to a greater number of trained individuals available to build, support, and work on
these productions. And because the model is well proven, it may be easier to find,
understand, and fix any errors or problems that occur.

LIVE REMOTE PRODUCTION 20

 Familiarity is not to be underestimated when
crews have to work together in real time

On Location production can also be cheaper in the short term, due to existing
investment in equipment and locations such as studio complexes and outside
broadcast trucks.

Low latency

Even with a longer term view, however, our contributors highlighted that this model
still has benefits of low latency and reduced bandwidth requirements.

Latency is the delay experienced as signals move over a network (or through
processing equipment), and so placing all components of the production chain in a
single location ensures very low latency throughout.

The fact that all signals remain on one site until the final production output can also
make this the most cost effective model when broadcasting from sites which don’t
have good connectivity. The requirements for bandwidth could be as simple as a
single programme output feed, which could be distributed by satellite, for example.
Nonetheless, if a venue is expected to be used for live production regularly, it may
be more economical to invest in connectivity, to enable other remote production
architectures to be used.

LIVE REMOTE PRODUCTION 21

 2 Remote Controlled

3 Processing remains at the acquisition site, but is remote controlled from

elsewhere

3 This reduces both travel requirements and the density of crew on site, while
avoiding huge bandwidth requirements

3 While crew travel is reduced, the volume of equipment on location is not,

leading some to see this model as a sub-optimal compromise

REMOTE ACQUISITION/
CONTRIBUTORS

AV signals Return monitoring

ACQUISITION SITE Control PRODUCTION HUB

signals

Monitoring
feeds

Programme
output

In our second model, some or all of the control surfaces are removed from the
acquisition site, and moved elsewhere to a production hub. The primary signal
routing and processing still takes place on the acquisition site, hence the processing
equipment is being remote controlled from elsewhere.

Other names for this model include remote operation, home run production, remote
control surface, or remote surface.

Physical distancing

Remote Controlled production has been particularly popular during the pandemic,
as it reduces the number of staff in the OB truck or production gallery, enabling
social distancing. It is also a popular option where the connectivity available at the
acquisition site lacks the bandwidth required to backhaul all raw camera and audio
feeds to a remote production hub.

LIVE REMOTE PRODUCTION 22

Smaller footprint

As with On Location production, this model may be chosen because of previous

investment in infrastructure and equipment. Existing outside broadcast trucks or
studio equipment can be used, with the signal processing components staying in
place, and new remote control surfaces added.

Smaller vehicles may be able to reach locations,

or park in spots that would be inaccessible to
full sized OB trucks

However, when designing new trucks specifically for this model, our experts told
us that it is increasingly common to build smaller vehicles with reduced production
areas, recognising that there will be fewer crew on site. This has advantages not just
in cost, but also in flexibility, as these smaller vehicles may be able to reach locations,
or park in spots that would be inaccessible to full sized OB trucks.

In the case of fixed studios, a largely remote controlled facility can take up
considerably less space, which could be advantageous when located in expensive
real estate. Conversely, additional space is required in the production hub, with some
contributors telling us that they are having to convert office and meeting space in
their buildings to additional galleries, in order to enable more remote controlled
production.

Other advantages of remote controlled production include the potential ability to

fall back to On Location production in the event of a network failure, reducing risk
somewhat; though of course if the number of crew on site has been reduced then
capability may be limited.

It is also important to ensure that sufficient technical skills are available at the
acquisition site to manage the equipment there and deal with any failures. This can
call for multi-skilled engineers and operators, and in some cases the need to provide
skills at both locations can actually mean that the total number of crew required
increases, compared to an On Location production.

Connectivity choices

As noted on the diagram above, the primary traffic moving over the network is video
monitoring to allow crew to see the video feeds they’re controlling, and command
signals from their control surfaces back to the signal processing. In many cases, the
video monitoring need not be full resolution; indeed low latency is more important,
dictating the choice of video transport protocol. Latency is also critical when it comes
to intercom, ensuring that those at the acquisition site and those at the production
hub can communicate, including both on-screen contributors and production crew.

LIVE REMOTE PRODUCTION 23

 3 Centralised

3 Video, audio, and data feeds are streamed to a central location for
processing and control

3 This reduces travel and increases equipment utilisation, delivering

substantial cost savings

3 As availability of high bandwidth connectivity continues to grow,

use of centralised production is expected to increase significantly

ACQUISITION SITE AV PRODUCTION HUB

signals

Return
monitoring

Programme
output

Centralised production is perhaps the most challenging model to name, as evidenced

by the range of titles used across the industry, including REMI (Remote Integration),
split production, at-home production, and hub & spoke.

No matter the name, this model centralises the processing into a production hub
away from the acquisition site. Or to view it from the reverse angle: the content is
captured live at a remote location, while production takes place centrally at a hub.

High bandwidth

Because the raw camera and microphone feeds (and, where applicable, data feeds)
are being backhauled from the acquisition site to the hub, this model requires
significant bandwidth.

One of the most prominent examples of this model is the Distributed Production
Network deployed by Telstra in Australia, with anchor tenant Fox Sports. It connects
over 40 venues which host high profile events regularly, including sports such as
rugby, Australian rules football, cricket, and basketball. Featuring network speeds
of up to 100 Gbps at each site, this level of bandwidth is clearly not achievable at all
venues worldwide. But where it is possible, there are significant benefits to be gained.

LIVE REMOTE PRODUCTION 24

Higher utilisation

As well as the obvious travel expense savings, there are significant potential savings
due to better utilisation of equipment. Hardware processing can be used across
programmes or events, rather than being unused during transport from one location
to another.

There is also a related saving in wear and tear on equipment when it’s not being shipped
or transported in OB trucks. In general, this architecture – along with Cloud production –
was felt by our experts to have the greatest potential for cost savings overall.

This architecture was felt by our experts to

have the greatest potential for cost savings

Of course, for those who have previously used a traditional outside broadcast model,
longer term savings may be offset in the short term by a requirement for capital
investment at the hub, or in connectivity.

There are also advantages for the production team. While reducing the need for them
to travel, this model still places key roles of the production team together in a single
location, enabling easy collaboration and low latency communication.

Smaller OB trucks

When it comes to the on location aspects of the architecture, there is less equipment
required and fewer staff. Outside broadcast trucks can be even smaller than in the
Remote Controlled model, with little to no space given over to production areas. This
has enabled some OB companies to use half sized trucks, containing just a small
production area for disaster recovery scenarios. It should be noted however that with
processing equipment centralised in the hub, the surfaces in the trucks may actually
be remote controlling the hub.

Once again, small trucks bring abilities to access more difficult sites and park with
less available space. Meanwhile, less equipment to deploy on location means
quicker setup times and less potential for problems that need troubleshooting. Some
producers are exploring ways to further reduce the on location crew by using more
remote controlled cameras, or high resolution fixed camera arrays from which views
can be cropped out.

Flexible production

With the production team centralised in a production hub, there are many ways
to acquire the source video and audio. High end productions may need OB trucks
and huge bandwidth, but the Centralised production model also has more flexible
applications across a wider range of content.

LIVE REMOTE PRODUCTION 25

During COVID-19 lockdowns, there was explosive growth in the use of video
conferencing over domestic internet connections for this purpose. But the most
significant enabler in the longer term may be the use of bonded cellular connections,
harnessing 4G and/or 5G mobile internet to transport video from almost anywhere.

This technology means that camera feeds can be transmitted wirelessly back to the
production hub, proving especially useful for newsgathering, or for sports not taking
place in stadia, for example.

Our experts reported significant growth in these use-cases in recent years, and
expected that growing availability of 5G would drive more usage of the Centralised
production model.

Growing availability of 5G will drive more

usage of the Centralised production model

LIVE REMOTE PRODUCTION 26

 4 Distributed

3 Processing and control are distributed across multiple sites, most commonly
by enabling crew to work from home

3 This provides huge flexibility, though it comes with some technical challenges
around the latency of different networks including home broadband

3 The ability to use the best crew, no matter where they’re located, unlocks
creative potential – provided communication systems enable them to
work at their best

HOMES

Monitoring Control
feeds signals

AV
ACQUISITION SITE signals PRODUCTION HUB

Return
monitoring

Programme
output

A logical development of the Centralised model is an architecture in which the signal

processing is centralised, but operators are able to work remotely, either in secondary
production hubs or at home. We call this Distributed production.

Sustainability and creativity

While a Centralised model reduces the need for crews to travel to the acquisition site,
they still need to travel to the production hub. By enabling crew to work from home – or
from regional production hubs – some broadcasters are trying to reduce commutes and
travel requirements even further, to achieve significant sustainability benefits.

Others are using a Distributed production model to access the best talent wherever
they are in the country; or indeed around the world. The promise of the best creative
talent being able to work on any production is certainly tantalising.

LIVE REMOTE PRODUCTION 27

 A Distributed production model helps
access the best talent wherever they are
around the world

One sports organisation explained how they operate graphics in Spain and multi-feed
production in the UK. Another operated eleven control rooms in different locations to
cover a tournament, all controlling signal processing in a central hub.

Communications challenges

When production crew are dispersed, it is critical that they can communicate effectively.

Some experts felt that having crew working at home creates such a barrier to their
team working that it would not persist beyond the pandemic. They cited production
teams who work very tightly together using non-verbal cues, saying that such teams
need to be together in one location, even if that’s not at the acquisition site.

Others, however, felt that effective intercom systems mitigate these concerns, citing
that even in most On Location scenarios, key production crew are in separate rooms
relying on intercom to communicate.

Where crew are working from home (rather than from production hubs that are
connected to high quality networks), the potential latency experienced on domestic
broadband can be a concern. Communication through intercom can be disrupted
by high latency, impacting the production team’s efficiency. Nonetheless, major
manufacturers do provide virtualised intercom systems that can be used remotely,
and some of our experts reported good experiences working with such systems.

Domestic limitations

Similar latency concerns may apply to video monitoring, which is also affected by the
bandwidth limitations of domestic internet connections. Despite advances in video
transport and encoding, the number of video feeds that can be delivered to users was
cited by our experts as a limiting factor.

Signal timing is also an important consideration in unmanaged networks, where

latency may not be consistent. But the concern most often raised in our discussions
was redundancy.

For decades, broadcast engineers have been used to ensuring that every signal has
at least two separate paths available to its destination, in case either connection
should fail. Generally, domestic broadband suppliers do not provide such redundancy.
However, many productions have made use of cellular communications as a backup
for wired broadband.

LIVE REMOTE PRODUCTION 28

 Video monitoring is affected by the bandwidth
limitations of domestic internet connections

All of the factors identified here are dependent on the type of content, its budget, and
its editorial requirements. Genres such as esports have been leaders in Distributed
production, and many others are now following that lead. However, those producing
the most valuable content – or the content most susceptible to the challenges of
higher latency – may favour a Distributed production model using multiple highly-
connected production hubs, rather than crew working at home.

LIVE REMOTE PRODUCTION 29

 5 Cloud

3 Processing moves to the cloud, with control at a production hub or

distributed across multiple sites

3 Multi-region cloud deployments offer flexibility for global productions,

though latency concerns are still paramount

3 The growth of high quality software tools for live production will enable new
workflow opportunities at lower cost, making this a dominant future model

Programme
output
CLOUD

AV signals Control signals

Return Monitoring
monitoring Control Monitoring feeds
signals feeds

ACQUISITION SITE PRODUCTION HUB HOMES

Our final remote production model moves some or all of the signal processing to
the cloud.

As noted previously, ‘the cloud’ here is used broadly to refer to any scenario in
which the processing is performed remotely from both the acquisition site and the
production hub. This could be public cloud, private cloud, hardware cloud (managed,
multi-tenanted installations of dedicated hardware production devices), or simply a
remote data centre.

While we have grouped these different versions of ‘cloud’ together for the purposes
of understanding the most important remote production models, many of the distinct
characteristics, advantages, and disadvantages of each are discussed further in the
DPP’s The Cloud for Media.

LIVE REMOTE PRODUCTION 30

In general, the unifying characteristic of Cloud production is that video, audio, and
data feeds are ingested into the cloud processing environment, which is remotely
controlled from elsewhere. That could be from one or more production hub(s), or
from operators’ homes.

To a great extent, many of the same considerations apply with this architecture
as with Remote Control, Centralised, and Distributed production. It is still vital to
have the required connectivity; correct signal timing is still key; and latency must be
carefully considered, especially in relation to functions such as intercom.

A matter of region

To many, public cloud offers the promise of a truly dematerialised facility, with
almost unlimited flexibility and scalability. It is enabled by software driven workflows
running on commodity IT hardware.

However, the cloud is still ultimately powered by physical hardware, which is still
subject to the laws of physics. As such, the cloud is not location agnostic; which is
why every public cloud provider offers control over the regions in which infrastructure
is configured and hence processing is running.

The cloud is still subject to the laws of physics

In order to minimise the challenges of latency, it is important to choose your region

wisely. The objective is often to position the compute resources (and hence the
signal processing) as close as possible to the video sources. And depending on the
production, there could be advantages to deploying across multiple regions. Some
producers have, for example, deployed vision mixing functions in one region and
graphics in another. In doing so, they aim to balance the trade off between proximity
to the acquisition sources and proximity to operators; using the cloud provider’s fast
backbone to move content between regions.

Software catching up

Some contributors to this research reported that they could not yet move to wholly
public cloud solutions, due to the lack of a complete and fully featured production tool
set implemented in software. A common example is vision mixing, where software
tools originally designed for use cases like game streaming are finding increasing uses
in broadcast.

LIVE REMOTE PRODUCTION 31

One expert complained that none of these solutions offer multiple Mix/Effects buses
(essential functionality for their production), while another explained that even simple
productions required the use of three separate instances of the vision mixing software
connected together. And most of these setups require licenses of the software to be
deployed onto virtual machines, controlled by remote desktop connections.

This caused one expert to describe them as ‘a bit makeshift’, as he looked forward
to more capable tools that are operated through web browsers or dedicated
control surfaces.

Nonetheless, the software and cloud tools are catching up fast. Recent high profile
product announcements from cloud vendors and their partners have delivered
functions including high quality inter-instance video connectivity, while some SaaS
(software as a service) video production tools are now available. They may not yet
fully replace professional hardware solutions, but their capabilities are growing rapidly.

A flexible future

While there are undoubtedly technical challenges yet to be overcome, there are
perhaps as many perceptual hurdles. As one of our experts put it, “just as we see range
anxiety with electric cars, we collectively have a bit of cloud anxiety in broadcast.”

Just as we see range anxiety with electric cars,

we collectively have a bit of cloud anxiety
in broadcast

Yet despite the notes of caution from some, cloud processing is a viable option for
many live productions today, and its usage is certain to grow. There is huge benefit
to be gained from the ability to configure a production infrastructure without capital
investment or long build times.

LIVE REMOTE PRODUCTION 32

Variations

The architectures that have been outlined above are hugely helpful starting points, but
they are rarely deployed in a pure form. Each individual deployment will have nuances
and adjustments. Some of the most significant or common variations include:

3 Remote Controlled On Location production:

This is a form of Distributed production in which there is no production hub; signal

processing takes place at the acquisition site, with some functions remote controlled
from crew homes.

3 On Location audio:

Some producers find that audio and/or intercom processing must be performed
at the acquisition site in order to deliver low enough latency, even when other
processing is remote.

An example is the mix minus feed provided into a presenter’s earpiece. If that
presenter can hear a sound source – such as another contributor’s voice – both
directly (because they are physically colocated) and also through their earpiece, then
a delay of even a few milliseconds between the two can be intolerably disorienting.

3 Downstream production:

This is a model in which the main audio and vision mixing capabilities are deployed
at the acquisition site, but functions such as graphical overlays are implemented at a
production hub, from where the main distribution output is created.

3 Hybrid Remote production:

In some cases, capabilities such as a vision mixer are deployed at the acquisition
site and in a production hub. For example, a primary feed might be produced at the
acquisition site, but delivered along with some isolated camera feeds back to the
production hub, from where different variations and localisations are produced.

LIVE REMOTE PRODUCTION 33

THE RIGHT TYPE
OF REMOTE

Having defined five key models of live production, it is worth examining the reasons
why a production might choose one over another.

There are four primary constraints and considerations that were raised repeatedly
during discussions with our expert contributors: content, bandwidth, latency, and cost.

Content

3 Creative potential is maximised when the production team can

communicate seamlessly

3 This can be achieved remotely, but there are still advantages to having the
bulk of the crew in one location: On Location or Centralised production

Whichever production model is chosen must serve the goals of the content being
produced, so we asked our contributors which production model they’d choose if
basing their decision primarily on maximising creativity and production value. Despite
a general enthusiasm for remote production, On Location production was the most
commonly chosen answer.

However, a close second was Centralised production, reflecting a feeling that

production teams work best when they are physically together, whether they’re at the
acquisition site or a separate production hub. A number of experts commented that
the energy generated by a crew working closely together is hard to replicate when
they are geographically dispersed.

Production teams work best when they are

physically together, whether at the acquisition
site or a separate production hub

LIVE REMOTE PRODUCTION 34

We also asked about the creative impact of moving specific functions remote from
the acquisition site. Remote controlled camera operation was felt to be a creative
hindrance overall, while remote audio mixing was also seen by many contributors
as having negative editorial impact. This reflects the audio latency issues discussed
previously under the On Location audio model.

Conversely, the ability to operate graphics and post production remotely were seen
to have creative benefits. And for those producing multiple types of content, remote
production using virtualised tools can also offer new flexibility. A sports broadcaster,
for example, might centralise their processing in a hub, and create different control
room configurations for different sports.

Optimising video transport

Centralised production is generally performed using high bandwidth

managed networks, with protocols such as SMPTE ST 2110 for video
transport. It is not unusual for HD feeds to be uncompressed, but for UHD
– or where bandwidth is more constrained – it is more common to use
lossless or light lossy compression, such as JPEG XS, JPEG 2000, or VC-2.

The choice of codec essentially trades off three factors: bandwidth,

latency, and quality. Some prioritise low computational complexity,
reducing the processing latency involved in encoding and decoding
the video. Others reduce the required bandwidth, at the expense of
increased processing time, or reduced quality.

When working with unmanaged networks like the internet, a range of

video transport options is available. WebRTC is an increasingly popular
real time protocol, used in many video conferencing tools. It is designed
to run in web browsers, and prioritises low latency delivery above video
quality or reliability. As such, it can be used to great effect for monitoring
and confidence feeds delivered to users at home. Protocols such as RIST,
SRT, and Zixi, meanwhile, prioritise reliable high quality video delivery,
and although they introduce more latency, they are often better suited
for delivery of the main programme content such as camera feeds.

Further detail on the available options for streaming protocol and

compression is available in Recommendation DPP001 Live IP Profiles.

LIVE REMOTE PRODUCTION 35

Bandwidth

3 Bandwidth is a key limiting factor at many sites, and connectivity is now

part of the site assessment for outside broadcasts

3 When bandwidth is constrained, it is preferable to use models that avoid

transporting raw camera feeds: On Location or Remote Controlled production

The bandwidth available to the acquisition site will have a significant impact on
the production architectures that can be employed. Remote Controlled production
is possible with relatively low bandwidth, while Centralised production requires
considerably greater connectivity. Distributed production relies upon the available
bandwidth to each production hub and/or to the crew’s homes.

The video aspects of production are of course the most restricted by bandwidth,
due to the higher data rates of video compared with audio and control signals. Using
uncompressed video, or common live production codecs, a Centralised production
with tens of camera feeds could have a bandwidth requirement of between 10 Gbps
and 100 Gbps.

Using uncompressed video a Centralised

production could easily have a bandwidth
requirement approaching 100 Gbps

Such connectivity is often available at large and frequently used venues such as
major stadia or studio complexes. But at other venues, a choice may need to be made
between using a different production architecture (such as Remote Controlled), fewer
camera feeds, or heavier compression.

The need for connectivity has completely transformed the priorities of site surveys for
many outside broadcasts. As one of our experts put it, “the biggest problem we have
is getting fibre lines to a car park in rural China”.

For those using a Cloud production model, direct connectivity to their chosen cloud
provider can also be a key enabler. There has been significant growth in such provision,
and most major network providers can offer interconnection to major cloud platforms.

LIVE REMOTE PRODUCTION 36

Latency

3 Latency is a key consideration for all remote productions, but is especially

important for systems such as intercom

3 Latency can be managed in each remote production model, but is likely to

be minimised when using those with simpler network topologies:
On Location and Centralised production

Latency is the delay introduced into the signal path either by the time taken for a signal
to traverse the network, or by signal processing. It affects all of our remote production
models, as latency can be problematic for both control signals and media streams.

While latency can be carefully managed and reduced to a certain extent, it is

ultimately governed by the speed of light; there is a theoretical minimum time
taken for light to travel along a fibre optic cable between two points. As a result, the
latency involved in a remote production will differ depending on the physical distance
between the acquisition site and the production hub.

For example, it may be completely feasible for a vision mixer in Tokyo to be remote
controlled from Osaka or Seoul, but not from Paris or Chicago. When considering
Remote Controlled production, our contributors commented that many pieces of
production equipment cease to work reliably if the latency on the control signal
reaches 100 ms or more.

When moving media, such as in a Centralised architecture, increased latency can

cause more problems with audio than video. As noted previously, high latency in
audio mixing and intercom can cause problems for the on-screen contributors and
crew. By using a Distributed or Cloud architecture with appropriate choices of data
centre or region, it may be possible to reduce latency by locating critical processing
closer to the acquisition site.

The effects of latency also vary greatly depending on the content. A few
milliseconds of delay might be problematic when cutting between shots of racing
vehicles moving at 200 mph, but not when managing a camera feed of a prime
minister’s slow moving motorcade. In the former case, an On Location or Remote
Controlled vision mixer may be preferable, whereas the latter might better suit a
Centralised one.

Finally, it’s worth noting that it’s not always the absolute amount of latency that’s the
biggest concern; it’s also the consistency. Unstable latency can cause jerky responses
from remote controlled equipment, while variable latency in audio and video signals
may require buffers to be introduced – further increasing the delay.

LIVE REMOTE PRODUCTION 37

Cost & risk

3 Remote production reduces travel and transport costs, and the most
effective models enable greatly increased utilisation of equipment and crew

3 When considering new build infrastructure there are favoured models for
cost effectiveness: Centralised and Cloud production

Alongside creative and technical constraints, cost is a crucial factor in choosing a

remote production model. There are too many variables at play for it to be possible
to give a simple answer to the question, which architecture costs least? But when
asked, our experts gave strong indications that for a new build, Centralised and Cloud
production have the potential to be the most cost efficient.

Centralised and Cloud production have the

potential to be the most cost efficient

When asked to assess the potential cost impact of moving individual functions to
a remote model, the most positive impact was expected when a combination of
functions like vision mixing, graphics, replay, and vision engineering are remote.

With that said, the impact hugely depends on factors such as the travel cost of
sending people to the acquisition site. If the event to be filmed is a single match at
a local football ground with no connectivity, and the production crew all live within
ten miles, then the value assessment of Centralised production is unlikely to be
favourable. But for a multi-day event at a well connected arena 1,000 miles away from
the production team’s home base, the answer will be very different.

One source of cost in any production is the mitigation of risk, and this is highly
dependent on the content.

Productions on a budget might hugely benefit from the flexibility of Distributed

production, and be able to accept some level of risk from a remote contributor’s
internet connection failing. But a high budget, high profile music concert or sporting
event will require resilience and redundancy at every stage of the production chain.
This is of course achievable, but the more locations and network connections in use,
the more signal paths may need to be doubled up.

LIVE REMOTE PRODUCTION 38

Separating control surfaces from processing devices has potential benefits here.
Should a processing device or its connectivity fail, the control surface can be
switched rapidly to control a backup device. And whether for disaster recovery or
simply cost savings, some organisations have used software control surfaces to
enable a single operator to control multiple tools or functions that would traditionally
have been separate.

When things go wrong

There are two ways to plan for technical failure: have people on
hand to fix problems, or have the ability to switch to alternative
devices/connectivity.

When working remotely, on site technical support can be a challenge.

There is generally no avoiding the need to send some technically
skilled individuals to the acquisition location, but there are different
approaches available.

Some organisations have deployed multi-skilled engineers to work on

location, with the aim that they can handle many different types of
task. Others see advantages in developing deeper specialism within
a centralised support team, that can help any production remotely. A
camera specialist from headquarters might be able to work via video
call with the camera operator on location to troubleshoot an error. Or a
software engineer could remotely fix a problem with some processing
tools in the cloud.

The need for redundancy and spare equipment is hardly new to

broadcast engineers. But Centralised production promises to make such
provision more efficient by sharing spares across multiple productions.
Cloud production, meanwhile, offers the ability to use multiple
availability zones to increase resilience by duplicating virtualised
processing tools across different data centres while sharing resources
such as storage.

LIVE REMOTE PRODUCTION 39

THE FUTURE
OF REMOTE

The DPP’s expert contributors were bullish about the future of remote production.
They estimated that by the end of 2022, use of On Location production will diminish
significantly, while usage of all of the remote production models would grow.

On Location

Remote Control Distributed Centralised Cloud

Growth in Remote Controlled production and Distributed production will be

significant, according to our experts. But Centralised production and Cloud production
will become the dominant models in the coming years.

Centralised production and Cloud production

will become the dominant models in the
coming years

Post-pandemic normalisation

In estimating future production trends, our experts were asked to compare against
a baseline of 2019. This is significant, because in 2020 there was huge growth in
Distributed production as media organisations found ways to enable home working.
But in general, there is an expectation that most high value television production will
step back from home working as requirements for social distancing recede. This leads
to an expectation of more Centralised production in place of Distributed.

LIVE REMOTE PRODUCTION 40

Nonetheless, home working will continue to be useful for some productions.
Those on lower budgets may find it cost effective, while certain genres such as
esports and social shows are generally more comfortable with this working model.
It should also be noted that the Distributed architecture doesn’t necessarily imply
home working; it also incorporates configurations with multiple production hubs.
This model is becoming increasingly popular with some broadcasters.

Technology improvements

From open source video transport protocols to global fibre networks, virtualised
production tools to cloud infrastructure, recent technology developments have
been critical to enabling remote production. But there is plenty of room for
further improvement.

Our experts foresee further virtualisation of the production environment, enhancing the
flexibility to deploy any equipment and any role in any location. Continuing rollout of
higher bandwidth fibre networks is critical of course, while wireless technologies such
as 5G have the potential to revolutionise acquisition and contribution in particular.

Cloud providers and 5G networks alike are working to bring more edge compute
capability, distributing computing infrastructure around the network. This offers the
potential to locate processing power closer to where it’s needed, reducing network
latency. The laws of physics cannot be overcome, but continual optimisation will
enable the most appropriate architecture to be deployed for each production.

Workflow familiarity

As technology develops, so too will producers’ familiarity with the tools and
workflows. While many challenges will be solved by technical improvements, others
will be managed by adjusting workflows.

One contributor explained that, “I see a lot of benefits for remote operations if we
rethink our workflows and operations”.

But in the fast-paced live production environment, it can be hard to find time to
rethink and reconsider. Crews are familiar with their environment, their workflows,
and their tools. They have learned highly effective communication mechanisms, and
can react instinctively to the situations they find themselves in.

At times it can be challenging to learn new workflows, new ways to communicate,

and new tools. But for every challenge comes a new opportunity. An opportunity to
deliver a better production at lower cost. An opportunity to reduce the environmental
footprint of our industry. Or an opportunity to produce coverage of an event a
thousand miles away, and be home in time to put the children to bed.

LIVE REMOTE PRODUCTION 41

 Remote production offers the opportunity to
deliver a better production, at lower cost, and
to be home in time to put your children to bed

By reaching consensus on five key production architectures, the DPP members

who’ve contributed to this research have enabled us all to better understand the
future of remote production. And that’s important because remote production – in all
its forms – is here to stay.

LIVE REMOTE PRODUCTION 42

This report was written by Rowan de Pomerai and designed by Vlad Cohen.

Workshops for Live Remote Production were led by Rowan de Pomerai, and organised
by Abdul Hakim and Anh Mao. Background research was by Alex Fenton.

Management of the Going Live & Remote project is by Abdul Hakim, with support from
Jayne de Ville and Anh Mao. Content for the project is led by Mark Harrison and
Rowan de Pomerai.

About the DPP

The DPP is the media industry’s business network. It is a not-for-profit company with
an international membership that spans the whole media supply chain, covering global
technology companies, production companies, digital agencies, suppliers, service
providers, post production facilities, online platforms, broadcasters, distributors
and not-for-profit organisations. The DPP harnesses the collective intelligence of its
membership to generate insight, enable change and create market opportunities. For
more information, or to enquire about membership visit
[Link]

About Ross
High Impact, High Efficiency Production Solutions

Ross powers video productions for billions of global viewers daily with the industry’s
widest range of high impact, high efficiency production solutions. Ross makes it easy to
create compelling news, weather and sports broadcasts, engaging material for sports
stadium screens and entertainment shows, remote productions, educational institutions,
legislative assemblies, corporate applications and inspiring content for houses of worship.

Ross solutions have impressed the audiences and marketing partners of NBC Sunday
Night Football, Eurosport, BBC World, Google YouTube Space London and China’s
eSports powerhouse VSPN. Ross delivers an unrivalled range of products and services
comprising virtual studios, real-time motion graphics, cameras, robotic camera
systems, production switchers, video servers, infrastructure and routers, social media
management, newsroom systems and mobile live events.

This publication is copyright © Digital Production Partnership Ltd 2021.

All rights reserved.