Delivering Quality Ultra HD

Delivering Quality
Ultra HD
By Ian Trow, Senior Director, Emerging Technology
& Strategy, Harmonic Inc.
A White Paper by Harmonic Inc.

August 2013
www.harmonicinc.com
Delivering Quality Ultra HD

White Paper I August 2013
ABSTRACT
Broadcasters are keen to follow the lead of cinema and deliver video resolutions beyond 1080p. The introduction of the 2K and
4K digital cinema formats in the last decade enabled movie producers to replace traditional film production and post production
with complete high-resolution digital workflows. The 4K Ultra HD format offers similar promise for broadcasters, but calls for a
re-evaluation of the fundamental parameters key to real-time and file-based content delivery.
To date, Ultra HD has been marketed on improved spatial resolution, with little consideration given to frame rate, viable transmission
bandwidths, bit depth, dynamic range and interfacing. Key to the initial success of Ultra HD will be the high-quality spatial and
temporal up-scaling provided within television sets to display HD content. Given the availability of 4K movie content and the
desire for improved broadcast HD quality, video on demand (VOD) services will likely seed the market for Ultra HD adoption.
As acceptance increases, we can expect Ultra HD viewing to move away from the optimized experience of darkened home
theaters into the harsh ambient lighting found in the typical living room. This inevitability will reveal the inadequacy of
employing a 24-frames-per-second (fps) format in a television environment where high-motion sports are frequently watched.
The use of higher frame rates can overcome this challenge, but will require the development of new professional and consumer
interface technologies to handle the increased data throughput. Proposals exist for the continued use of video-specific interfaces
for Ultra HD delivery, but high-speed-based Ethernet technologies could also enter the picture. Bit depth and dynamic range are
the last two factors that need to be addressed to provide high-quality Ultra HD delivery to the home.
INTRODUCTION
Since televisions introduction in 1925, viewers have been treated to a steady rise in resolution; the only thing needed to enjoy
the improved picture quality is a new TV set. These improvements in video quality have sometimes been accompanied by an
increase in frame rate, which results in smoother motion video. More often than not, while television resolution and size have
steadily grown, the most visible quality gains have been realized by television-set manufacturers improving the delivered signal
by up-converting the frame rate.
This fact is particularly relevant when considering a transition from HD to Ultra HD television. The ideal scenario for displaying
Ultra HD is to shoot and process the content in its native standard. Herein lays the first major workflow hurdle for Ultra HD. While
the film community has implemented the Digital Cinema Initiative (DCI) to deliver a complete 4K workflow, the effort has been
largely focused on 24-fps content (a notable exception being the recent release of The Hobbit at 48 fps). The 2013 Consumer
Electronics Show in Las Vegas raised consumer interest in Ultra HD, with a vast range of monitors demonstrated in darkened
areas to show 4K film content at 24 fps without motion scenes appearing too jerky. These demonstrations have paved the way
for Ultra HD to penetrate the lucrative home video marketbut what workflow advances are required to deliver a genuine
improvement over existing HD services?
www.harmonicinc.com

CINEMATIC 4K WORKFLOW
Digital production workflows for cinema are similar to the legacy workflow associated with film; the critical difference being the
ability to immediately review shots during production. This improvement overcomes the need to process film negative in order
to review dailies (or rushes), a procedure that has traditionally introduced both delay and cost to film production. A further
benefit of digital workflows is that the source content is available for editing, processing and post production without the need
for film transfer (Figure 1).
Editing
Capture
Library
Storage
Compositing
Online
Storage
Rendering
Preview
Grading
Figure 1. A typical cinema workflow
www.harmonicinc.com
Mastering

FROM CINEMA TO THE HOME: THE ISSUES

The workflow for broadcasting Ultra HD is similar to that for digital cinema but is more complicated and has a number of workflow
and format differences (Figure 2). Cinematic capture is targeted towards acquisition of content, whereas a typical television
workflow also has to address factors such as rebroadcasting of turnaround content, file-based media management, content
archiving and the complexities of regionalization.
Editing
Camera
Turnaround
File
Playout
Automation
Ingest
Server
Central
Archive
Archive
Remote Feed
Browse
Serving
Multiscreen
Transcoding
Library
Archive
Figure 2. An Ultra HD broadcast workflow
www.harmonicinc.com
Encoding
Multiplexing
Transmission
CDN

Resolution
The resolutions for 4K adopted within DCI and proposed for Ultra HD are similar (Figure 3). The slight discrepancy lies in the
drive towards higher resolution for television that respects HDs 1920 x 1080 resolution. Consequently, Ultra HD doubles the
horizontal and vertical resolutions of HD, resulting in a format that stops short of fulfilling the full DCI 4K specification. This
approach makes sense: establishing Ultra HD resolution as a multiple of HD will allow legacy HD content to be more easily
up-converted to the new format.
Ultra HD
3840 H x 2160 V
8,294,400 pixels
HDTV 1080p
1920 H x 1080 V
2,073,600 pixels
SDTV
720 H x 480 V
345,600 pixels
2K
2048 H x 1080 V
2,211,840 pixels
Figure 3. Comparison of television resolutions
www.harmonicinc.com
Digital Cinema 4K
4096 H x 2160 V
8,847,360 pixels

Frame Rate
To date, most of the marketing behind Ultra HD has concentrated on resolution; consequently, the majority of the
demonstrations at CES 2013 featured highly detailed scenes with little or no motion. Where representative television content was
shown, it was in a darkened environment that avoided the awkward interaction between 24-fps content and the vagaries of
lighting. This same approach may be acceptable for a home cinema environment but not for a higher-quality replacement for
HD. Delivering higher resolution video without an increase in frame rate, on a monitor that will likely be far larger than current
screens, is a difficult proposition, especially for sports content. Given a choice between resolution and frame rate, frame rate
wins out. A similar justification is often made for sports content, with 720p selected over 1080i, namely for the emphasis on
temporal rather than spatial resolution for moderate- to high-motion content. At a minimum, Ultra HD will need to support 50 Hz
and 60 Hz frame rates, with the ideal being 100 Hz and 120 Hz.
Consumer Interface
Since its introduction 10 years ago, the High-Definition Multimedia Interface (HDMI) has shipped on 3 billion devices. The current
implementation supports both Ultra HD and 4K DCI formats. Ultra HD is supported at 24, 25 and 30 Hz, 4K DCI is supported at 24 Hz.
The current iteration of HDMI only supports resolutions up to 1080p at 50 Hz and 60 Hz; however, without further development
of the HDMI standard, any move towards Ultra HD will be temporally hindered unless the consumer is prepared to use dual
synchronized HDMI connections to their television. Though witnessed in some prototype screens at CES 2013, this approach is
unlikely to appear on consumer equipment. It is interesting to note, however, that multi-HDMI has been adopted on 4K cameras,
where four synchronized HDMI connections are used to accommodate the bandwidth required for baseband 4K video.
For Ultra HD to become a genuine successor to HD, consumer interface standards will need to evolve to support the data rates
required for Ultra HD at 50 and 60 Hz. The following table illustrates the approximate data rates for several possible Ultra HD
(3840 x 2160) format/frame rate scenarios1:
Application
Color Depth
Bit Depth
Frame Rate
Data Rate
Commercial Delivery
4:2:0
8-bit
60 Hz
5.96 Gbps
Commercial Delivery
4:2:0
10-bit
60 Hz
7.46 Gbps
Production Workflow
4:2:2
10-bit
60 Hz
9.95 Gbps
Sports Coverage
4:2:2
10-bit
120 Hz
19.91 Gbps
An important aspect for consumer interfaces is the additional bitrate required to sustain the needs of content-protection
mechanisms, essential to avoid pirated digital copies becoming easily attainable. Its worth noting, as well, that HDMI does have
a challenger standard in Thunderbolt, an Intel interface used by Apple and capable of 20 Gbps.
Bandwidth
Lets review the headline issue of how to compress the Ultra HD signal to make it viable for carriage over broadcast media. Both
MPEG-4 AVC (H.264) and the new High Efficiency Video Coding (HEVC) standard have been associated with Ultra HD
transmission. Due to the lack of HEVC-compliant equipment and workflows, most broadcast deployment trials to date have
utilized H.264. To enable processing of Ultra HD, four synchronized HD AVC encoders are typically used as an interim solution
while the industry waits for HEVC to deliver the latest compression performance improvements.
www.harmonicinc.com

An example of such a deployment is the recently launched demonstration Ultra HD channel by Eutelsat.2 While operating at a
very high bandwidth (40 Mbps), the emergence of Eutelsats Ultra HD channel so early in the market is indicative of a general
desire by broadcasters to embrace the format. It also shows the determination of broadcasters to utilize frame rates appropriate
for television (50 Hz), as opposed to taking the cinematic approach of sticking with 24 Hz. In time it is expected that, even using
H.264, bitrates can be reduced to a much more manageable 20 Mbps. Adopting such a configuration would allow for the carriage
of two channels within a 45-M transponder (Figure 4).
20 Mbps
Ultra HD
HEVC
2 Channels
20 Mbps
Figure 4. An Ultra HD/HEVC two-channel scenario
While this configuration represents an improvement, it still falls far short of the existing channel capacity for HD, where five
channels at 8 Mbps is the expectation (Figure 5).
8 Mbps
8 Mbps
8 Mbps
8 Mbps
8 Mbps
Figure 5. An HD/H.264 five-channel scenario
www.harmonicinc.com
HD H.264
5 Channels

Herein lays the challenge for broadcasters and Ultra HD: How to deliver the new format while matching the channel count required
to sustain revenues. HEVC is expected to deliver a 50% improvement over H.264, in terms of bitrate, once the standard is fully
optimized and implemented in second- and third-generation encoders. In the short-term HEVC will likely be launched with a more
modest 30-40% performance improvement over H.264. Even this gain would allow for an Ultra HD channel to be compressed
down to 14 Mbpsmaking room for a valuable extra channel on a 45-M transponder when compared with an Ultra HD H.264
implementation (Figure 6).
14 Mbps
14 Mbps
4K HEVC
3 Channels
14 Mbps
Figure 6. A 4K/HEVC three-channel scenario
When HEVC fully delivers on its promise of compression efficiency, a fourth channel could be offered with each service operating
at 10 Mbps. These aggressive compression rates are not possible with AVC (Figure 7).
10 Mbps
10 Mbps
10 Mbps
10 Mbps
Figure 7. A 4K/HEVC four-channel scenario
www.harmonicinc.com
4K HEVC
4 Channels

LG, in conjunction with KBS, demonstrated an HEVC-based system at CES 2013 for terrestrial broadcasting (DVB-T2) of Ultra HD at
60 Hz.3 The video bitrate deployed was 35 Mbpsoff the target 10 Mbps but, given that the HEVC standard had just been
officially released, a high-profile early endorsement.
While the bandwidth illustrations provided above were based around satellite payload, carriage over alternative broadcast media
should also be considered. Cisco predicts that Internet traffic will grow fourfold between 2011 and 20164 (Figure 8). While growth
in video-related traffic over the Internet is significant, the rate of growth is likely to slow. Therefore, it is reasonable to assume that
as overall capacity grows, more bandwidth will become available for Ultra HD over-the-top (OTT) carriage. In the short-term, HEVC
is likely to emerge as a strong proposition for HD OTT applications. Another area anticipated to grow significantly is VOD delivery
of Ultra HD, witnessed by recent announcements from Google5 and Netflix.6
Figure 8. Anticipated growth of video traffic over the Internet through 2016 (Cisco Video Networking Index)
www.harmonicinc.com

Professional Interface
As weve seen, the baseband signal throughput for Ultra HD is extreme, easily outstripping the capabilities of most commonly
deployed broadcast interfaces. While it is likely that a dedicated Ultra HD broadcast serial interface will emerge, an alternative
that addresses high-bandwidth IP network requirements could be adopted. As with the consumer interface, the emergence of
an Ethernet-based solution shows just how far standard IP approaches and solutions could penetrate future broadcast workflows.
In the case of professional interfaces, the rival IP format could be based on 40-Gbps and 100-Gbps technology. Recall that Ultra
HD (3840 x 2160), 4:4:4, 10-bit at 60 Hz would yield a data rate of 14.93 Gbps. Even catering for the most-demanding professional
applications, Ultra HD (3840 x 2160), 4:4:4, 12-bit at 120 Hz could be contained within 35.83 Gbps. This is within the 40 Gbps possible
with the next iteration of IEEE802.37 and is easily contained in the 100-Gbps variant.
Audio
4K in the cinema supports 7.1 audio within the DCI specification. Given the relatively moderate bitrate requirements for audio, it is
likely that Ultra HD audio support will be up to and including the DCI 7.1 surround sound specification. This capability would provide
backwards-compatible support for 5.1 surround sound, currently popular within both broadcast HD and home cinema systems.
Screen Technology
While improved resolution will be the main issue in the marketing of Ultra HD, frame rate and dynamic range will also have a
significant impact on screen quality. For Ultra HD to deliver on the promise of a truly immersive experience, screen manufacturers
must significantly improve the dynamic range of their TV sets, above and beyond standards for the latest HD screens. At CES 2013
various approaches were shown to accurately portray scenes more naturally by improving dynamic range. The improvement is
most noticeable in the ability to represent blacks and shadow outlines at the low end of the range while still being able to handle
areas of vivid color at the high end.
Related to dynamic range is the issue of compression bit depth. MPEG standards have largely adopted an 8-bit approach. The
exception is in AVC contribution, where the limitations of 8-bit are evident, especially on HD screens, manifested as contouring or
banding on large color washes within scenes. A typical example is often found in the rendering of clouds and sky. Ten-bit depth is
essential for resolving this issue, a fact accommodated in the first release of the HEVC specification. This is important for two reasons.
HEVC uses a similar approach to AVC, and so suffers from the limitation evident in high-end HD content. More importantly, the
increased screen size of Ultra HD televisions will make the limitations of 8-bit-based compression even more evident. The increase in
screen size also vindicates moves to at least sustain higher frame rates, and ideally endorse up-scaling to 100/120-Hz frame rates.
www.harmonicinc.com
10

ULTRA HD WORKFLOW
As weve seen, the Ultra HD workflow is considerably more complex than the cinematic workflow. This complexity is due to three
factors: 1) Broadcast involves high-bandwidth, real-time transfer throughout the chain. 2) Content is derived from a variety of
ingest sources, consisting of both national and regional feeds. 3) Even though the workflow is for Ultra HD broadcast, television
has to cater to an ever-expanding number of multiscreen options.
Over the next several years we can expect broadcasters to provision their production environment for Ultra HD in the same way
most production facilities currently do for HD. To cater for resolutions lower than Ultra HD, transcoding the original broadcast
content cost-effectively will be critical if broadcasters are to maintain a healthy margin. While this paper discusses the impact of
Ultra HD and advocates the use of HEVC, such a compression scheme is equally applicable to the need for compression efficiency
in multiscreen, all the way up to HD OTT delivery. Indeed, many of the challenges a broadcaster faces when trying to launch an
Ultra HD channel would be easier to solve in a multiscreen environment. This is because there is often little legacy STB or PVR
provision associated with multiscreen, which relies on software running on standard platforms that can be quickly repurposed
for new compression strategies, such as HEVC.
The key technology timeline shown in Figure 9 charts the most likely market and product availability dates for Ultra HD. As you
can see, OTT HD applications will utilize HEVC long before the launch of a full-blown Ultra HD live channel.
OTT
1st Gen live enc
PC/Tablet demo ware
2nd Gen enc
1st Gen file enc 4k24

Gaming Consoles/PC /tablet
1st Gen key decision
Today
H1 2013
Connected TV 4K Decode
2nd Gen key decision
H2 2013
H1 2014
1st Gen key decision

Sample
Dec silicon
Linear PayTV
H1 2015
2nd Gen key decision

STBs
4K p60 Live
4K nVOD
1st Gen enc
Product
H2 2014
2nd Gen enc
Market milestones
Figure 9. HEVC technology timeline

Another key point to address is the lack of native Ultra HD content when channels first launch. Consequently, Ultra HD screens
will most frequently have to deliver a quality picture from sub-optimum sources, such as 1080p-based HD. Therefore, the
perceived quality of Ultra HD screens will be reliant on how well up-conversion is implemented in the television set. Upconversion was a major issue when broadcasters transitioned from SD to HD. Given that the move to Ultra HD requires an even
greater stretch, sophisticated algorithms will need to be implemented in the most cost-effective manner.
www.harmonicinc.com
11

Ultra HD workflows are expected to be highly reliant on the compression efficiency made possible by HEVC. The consumer and
professional interfaces should be high-speed derivatives of Ethernet protocols using industry-deployed network connectors, which will
allow off-the-shelf networking approaches to replace dedicated broadcast provisioningwith obvious CAPEX and OPEX advantages.
In a best-case scenario, Ultra HD services should be broadcast at a higher frame rate than that currently deployed for HD, for
example 100/120 Hz. However, given that the majority of initial Ultra HD content is mastered at 24 fps, and the bandwidth
implications of employing the higher frame rate are considerable, 50/60 Hz will be the most likely transmission rate. Ultra HD
screen manufacturers will differentiate their products by offering optimized up-scaling techniques to achieve 100/120 Hz frame
rates from source content broadcast at much lower rates.
The adoption of 10-bit depth will yield additional, and significant, video quality benefits.
CONCLUSION
Ultra HD is an exciting prospect for broadcasters. While there are significant hurdles concerning interfacing, format
standardization and bandwidth requirements, solutions exist that will allow such services to launch in the next three years.
Understanding the workflow associated with Ultra HD will aid the cost-effective deployment of services at lower resolutions; for
example, OTT HD delivery. This is vital for broadcasters to improve their margins, especially as CAPEX and OPEX become an
increasing problem in a multiscreen world.
Native Ultra HD content is the ideal for a broadcast service, but this expectation will be seldom met with initial service launches.
Understanding how to repurpose HD and cinema-derived content will be crucial if Ultra HD television sets are to deliver the
wow factor desired by many consumers.
The prospect of Ultra HD is to fully immerse the viewer in the viewing experience, perhaps to the point of being indistinguishable
from watching the live event. If it can achieve this lofty goal, Ultra HD will assuredly be a game-changer for content providers,
broadcasters, service providers, television set manufacturersand, most criticallyfor the television-watching public.
REFERENCES
1. Video Bitrate Calculator.
http://web.forret.com/tools/video_fps.asp?width=4096&height=2160&fps=120&space=yuv422&depth=12
2. Eutelsat Launches Europes First 4K Channel, Broadband TV News, January 7, 2013.
http://www.broadbandtvnews.com/2013/01/07/eutelsat-launches-europes-first-4k-channel/
3. CES: 4K 60p on DVB-T2 by LG Electronics, AV Magazine, January 9, 2013.
http://www.avmagazine.it/news/televisori/ces-4k-60p-su-dvb-t2-da-lg-electronics-_7522.html
4. Cisco Video Networking Index.
http://www.cisco.com/en/US/netsol/ns827/networking_solutions_sub_solution.html
5. CES 2013 Highlights & Predictions, Red Bee, January 15, 2013.
http://www.redbeemedia.com/blog/ces-2013-highlights-predictions
6. CES 2013: A Complete Tour of Ultra HD, Ultra HDTV, January 15, 2013.
http://www.ultrahdtv.net/ces-2013-a-complete-tour-of-ultra-hd/#.UPzG5icdyJQ
7. IEEE 802.3-2012 Standard for Ethernet Expands to Address New Markets, Bandwidth Speeds and Media Types,
IEEE Standards Association, September 5, 2012.
http://standards.ieee.org/news/2012/802.3_12.html
www.harmonicinc.com
2013 Harmonic Inc. All rights reserved. Harmonic and the Harmonic logo are trademarks, registered trademarks or service marks of Harmonic Inc. in the United States and
other countries. Other company, product and service names mentioned herein may be trademarks or service marks of their respective owners. All product and application
features and specifications are subject to change at the sole discretion of Harmonic at any time and without notice.
08.26.13
12

Delivering Quality Ultra HD

Загружено:

Сведения о документе

Авторское право

Доступные форматы

Поделиться этим документом

Поделиться или встроить документ

Параметры публикации

Этот документ был вам полезен?

Это неприемлемый материал?

Авторское право:

Доступные форматы

Delivering Quality Ultra HD

Загружено:

Авторское право:

Доступные форматы

Delivering Quality

A White Paper by Harmonic Inc.