1 Summary of the current course of the debate

1.1 The current direction of the debate on digital TV

The debate on the direction of digital TV and datacasting in Australia has taken a disturbing turn. FACTS, with so much at stake in the analogue TV industry to protect, has naturally been a very powerful voice in the debate to date, but it has recently tabled proposals on key issues that, if implemented, threaten the success of digital TV and datacasting in Australia. Through its submissions to the Datacasting and HDTV reviews, its proposed channel allocation plans and its position on transmission and STB standards, FACTS has recommended:

• A severely restricted definition of datacasting and little bandwidth for datacasters (through inefficient planning) which means few new services for Australian consumers.
• Set top boxes (STBs) that have only a single use (ie can't be used for datacasting or pay TV) and have features that make Australian STBs incompatible with the world standard. This will raise the price to consumers for access to digital TV, retarding the take-up rate and the development of the market for additional services.
• No requirement for HDTV production. Merely upconverting low quality material to meet HDTV legal requirements will not give Australians the improvements in sound and picture quality that the free-to-airs have promised.

1.2 The solution

Fairfax is committed to ensuring that digital TV and datacasting are a success in Australia - it is in our interests as well as the country's. Success requires consumers to accept the new technology quickly and that will depend on three critical factors:

• the number and attractiveness of new services, particularly datacasting, which are not available with analog;
• the availability of cheap set top boxes (STBs) which have multiple uses (ie datacasting and pay TV as well as free to air); and
• the degree of perceptible picture and sound quality improvement provided by digital TV.

Fairfax has consistently advocated positions that will maximise the success of digital TV:

• a broad datacasting definition to facilitate the development of attractive new services;
• efficient spectrum planning to maximise the number of new services; and
• interoperable transmission and STB standards that will allow the use of readily available, cheap, world standard STBs.

1.3 Worth correcting course

The tide of the debate can be turned if Government acts now, but without strong action digital TV will fail with these consequences:

• consumer backlash against expensive non-world standard STBs, the need for 2 or more STBs and restricted service options compared with the rest of the world;
• inability to reclaim analog spectrum by 2009 because of poor consumer take up of digital TV;
• Australia being a digital TV backwater and suffering consequent loss of international competitiveness in this crucial area of the information revolution.

2 Digital TV needs datacasting to suceed

2.1 New services are needed to drive consumer takeup

As Fairfax has emphasised in its various submissions, the success of datacasting is crucial to the success of digital TV in Australia because it is the main source of new services. Around the world the takeup of digital TV is being driven by new services (datacasting and new channels) not HDTV.

• In the US (the original champions of HDTV) HDTV is being abandoned. See the conclusion draw by The Forrester Report "HDTV Dreams, SDTV Realities" - "HDTV won't fly, but SDTV will" and numerous reports of US broadcasters adopting SDTV multi-channelling and datacasting instead of HDTV.
• In the UK, there is no HDTV. Consumer takeup is rapid and is being driven by vast number of new services (channels and datacasting) and cheap, multi-function STBs.
• In New Zealand SDTV has been adopted and new services (channels and datacasting) and cheap world standard STBs are the driver for consumer takeup of digital TV.

Datacasting will be critical to offering consumers a rich set of new services and ensuring digital TV succeeds.

2.2 But early recommendations need to be changed to support datacasting

FACTS' current set of proposals would have 3 negative effects on datacasting in Australia:

• by defining it so narrowly, through imposing a multitude of onerous and unworkable restrictions, that the business opportunity of datacasting is destroyed;
• by limiting the number and quality of channels available to datacasters through inefficient spectrum planning; and
• by making the free to air STBs unable to receive datacasting services, thus reducing the size of the market.

Action needs to be taken on 3 fronts to ensure that the datacasting industry develops and thrives in Australia.

(a) Adopt a broad definition datacasting

FACTS' submission to the Datacasting Review proposes 8 separate restrictions on datacasting prevent datacasters providing a commercial television service. Only one restriction, the limitation, proposed by Fairfax, of 10 minutes of unrepeated moving images per 30 minutes of transmission of non-interactive services is needed to achieve that policy aim. FACTS expansion of this restriction to include interactive services would prevent datacasting internet access (because the datacaster can not control how much video an internet user watches) and is unnecessary because commercial TV is passive not interactive. The other 7 restrictions add nothing other than reducing the commercial viability of datacasters.

FACTS' proposal to ban datacasting of entertainment programming is overly onerous. Given that much of the material currently available on the internet is entertainment programming (ie games sites, online games, music sites, film star fan sites, sports sites, chat rooms etc) it is not sustainable to assert that the provision of entertainment is the exclusive domain of broadcasting. Preventing the datacasting of such material would kill many of the most attractive and natural datacasting services, such as internet access, online games and high speed download of music, books and games software.

FACTS' proposals to restrict video frame rates and to limit video to one quarter of the screen would make datacasting unattractive for consumers and anchor datacasting against the already outdated benchmark of low speed internet technology. Neither of these restrictions advance the policy objective of preventing datacasting of commercial television services because Fairfax's proposed video time limit already achieves this goal.

FACTS' proposal to limit datacasting to subscription services would represent unreasonable interference with the commercial decisions of prospective datacasters and is based on the flawed assumption that being available free of charge is a distinguishing feature of broadcasting. Internet content is equally available free of charge to anyone with an internet connection. Such a restriction will severely limit the takeup of set top boxes and the viability of datacasters.

(b) Make spectrum available for datacasting

Early indications from the ABA channel planning process are that there will be very little spectrum allocated to datacasting. But a studied review has revealed that the process used to develop the early plans is flawed. The problems in the process are:

• the use of outdated spectrum planning tools instead of far superior modern spectrum planning software;
• the application of spectrum planning assumptions based on analog transmission which are inappropriate for digital transmission; and
• failing to make full use of the additional capabilities of digital transmission (such as single frequency networks and conditional access) which can provide increased coverage using fewer channels.

Fairfax has advice that if these problems are remedied the number of available digital channels can be dramatically increased, with no loss of quality, from the 7 or 8 proposed by the ABA to 11 or 12. It is worthwhile noting that New Zealand, with 15% less available spectrum (49 frequencies vs 57 frequencies) looks like having 3 more digital channels and 11 more total channels than Australia. New Zealand spectrum planners have adopted modern spectrum planning software and digital techniques.

In particular the widespread use of single frequency networks (SFNs) will allow the main transmission frequency to be used for translators (for gaps in the main transmitter's coverage) instead of requiring a new set of frequencies for each translator, as is required with analog transmission. This will free up a large number of channels for datacasting.

Equally, the use of conditional access rather than reduced transmission power to prevent reception outside of broadcaster's licence area will allow higher power transmission and thus fewer coverage gaps requiring translators.

In light of the above, FACTS' statement that the purpose of its Datacasting Review proposal to limit datacasters to 1MHz channels (instead of the 7MHz channels used by broadcasters) is to make "datacasting open to as many players as possible" is quite misleading. 1 MHz channels limit rather than expand opportunities for datacasters because they undermine datacasters' ability to provide high speed internet access and other high speed services (because the capacity must be shared with all users) and create STB incompatibility (see below).

(c) Have an interoperable STB for all TV signals

Properly implemented, the same STB used by consumers to receive digital TV can also receive pay TV signals and datacasting signals. This is the case in other countries around the world, but FACTS' proposals effectively prevent this from occurring.

FACTS proposition to limit datacasters to 1MHz of spectrum highlights this recurring consequence of FACTS proposals - datacasting and pay TV will be incompatible with free to air STBs. By restricting datacasters to 1MHz channels instead of the 7MHz channels used by broadcasters, consumers will not be able to receive datacasting on the STBs used for free to air digital TV. Instead, consumers will need to purchase a separate STB for datacasting.

The same problem occurs if datacasters are allocated substandard 6MHz channels, as proposed by the ABA's Digital Channel Plan Discussion Paper. Fairfax has been advised by a major STB manufacturer that the extra cost of an STB that will tune 6MHz as well as 7MHz channels is about A$80 wholesale and thus over $120 extra for the end consumer. Efficient spectrum planning will avoid the need to use substandard 6MHz channels.

FACTS proposal to transmit only a Dolby Digital sound signal and exclude the DVB standard MPEG II stereo sound signal will also mean that Australians can't buy world standard STBs and FACTS can control an Australia specific STB. Transmission of Dolby Digital only, as well as FACTS adoption of MP@HL rather than MP@ML transmission will also:

• raise the price of the STB by requiring proprietary decoding hardware not included in any world standard DVB STB (see below);
• prevent the tens of thousands of rural users who currently receive free to air television over direct digital satellite links from using their DVB compatible set top boxes to receive FACTS version of digital TV;
• prevent the tens of thousands of satellite pay TV customers who currently receive digital satellite pay TV from using their DVB compatible set top boxes to receive FACTS version of digital TV.

3 Digital TV needs cheap, powerful STBs

FACTS' proposals which lock out datacasting and pay TV from the free to air STB, also directly undermine the success of Digital TV itself by:

• significantly raising the price of STBs; and
• artificially limiting STB functionality.

3.1 Cheap STBs are essential to digital TV's success

Digital TV will fail if the STBs used to receive and decode the digital signal for display on either a PAL TV, a SDTV or an HDTV are too expensive for the average consumer.

There are two main factors which determine the price of the STB:

• the complexity of the components; and
• the number of that type of STB produced.

FACTS have proposed an STB that will be expensive on both these grounds.

(a) Costly features should be optional

FACTS' STB has Dolby Digital sound decoding as a mandatory feature instead of the standard MPEG II sound stream. Dolby Digital is a proprietary standard which requires complex decoding hardware and a licence from Dolby. This will add significant additional cost to the STB. As a guide, an entry level Dolby Digital decoding amplifier retails for A$300 more than the same amplifier without Dolby Digital decoding and a stand alone Dolby Digital decoder retails for A$695.

To take full advantage of Dolby Digital the consumer will require a 6 channel amplifier and 6 speakers. Very few consumers have such audio equipment and thus will be paying $100s extra for features they can't use.

FACTS proposed adoption of MP@HL transmission also raises the cost of the STB because it requires greater processing power and 6-8 times the memory that MP@ML requires. Only the very small percentage of consumers with HDTVs will get any advantage from MP@HL transmission and the majority of consumers will get a worse picture because of the downconversion required and the inferior propagation characteristics of MP@HL.

(b) FACTS wants non world standard STBs

Australia is the only country in the world to have proposed the hybrid transmission standard of DVB with Dolby Digital instead of MPEG sound. The UK, Europe and New Zealand use the DVB standard with MPEG II sound. The US uses the ATSC standard with Dolby Digital. By adopting a hybrid standard Australia gets the worst of both worlds because STBs will have to be made especially for the small Australian market.

Contrary to FACTS' assertion in its HDTV Review submission, we can find no evidence of any other country adopting such a hybrid standard and would welcome a full discussion of the available evidence. All countries actively pursuing digital TV at this time have adopted either the US or the European standard. Australia will be alone in the world and will be unable to use the world standard STBs manufactured in Europe, nor export STBs to the rest of the world. Australian STBs will cost far more than world standard STBs because the manufacturing run will be limited by the small size of our market.

3.2 Multi-function STBs are essential to digital TV's success

FACTS' proposals which mean that free to air STBs are incompatible with datacasting (see 2.2(c)) also directly undermine the success of Digital TV by artificially limiting the STB's functionality. STBs that can provide access to datacasting and pay TV as well as free to air TV will be much more attractive to consumers and will drive the overall takeup of digital TV. Failing to provide multiple functionality will significantly limit the success of digital TV.

4 Even FACTS' HDTV commitment looks doubtful

FACTS' submission to the HDTV review makes it clear that they have little interest in providing high quality high definition TV to drive digital TV takeup. Instead FACTS propose to take the cheaper, low quality option of upconverting PAL and SDTV source material and make no commitment to true HDTV content.

The quality of the ultimate picture received by the consumer is determined primarily by the technical quality of the source material. Upconverting PAL or SDTV source material to HDTV for transmission does not give consumers a better picture than transmitting SDTV. If the consumer has a HDTV set, the set itself will upconvert SDTV to HDTV prior to display. If the consumer has a SDTV set, the HDTV signal will need to be downconverted again prior to display leading to picture imperfection.

Indeed, technical studies have shown that because of the 2 conversions needed for consumers with SDTV or PAL sets, the picture quality of HDTV transmission is worse than for SDTV transmission. Also, because of the greater bandwidth required for HDTV, less error correction can be used and thus the signal is more susceptible to interference.