Developing Diveemo

Every year, the IBC exhibition in Amsterdam fills the halls of the RAI centre with so much innovation that some new exhibits, no matter how groundbreaking they may be, find themselves overshadowed. The 2010 edition of the broadcast convention was no different, for as visitors formed scrums on stands where new broadcast consoles were being launched, or talked earnestly in the aisles about networking protocols, one very important new technology may have slipped past their notice. Unlike many of the products launched at the show, it has the potential to improve the lives of millions of people around the world, and perhaps even save a few. It also boasts one of the strangest names to have graced the exhibition: Diveemo.

‘IBC saw the first real Diveemo transmission on the booth, with the BBC on-air live from the UK, broadcasting to the Netherlands,’ confirms Alexander Zink, project manager for Diveemo at Germany’s Fraunhofer IIS, the technology think-tank that stands behind the creation of the MP3 and, crucially, Digital Radio Mondiale (DRM). Speaking some weeks after the exhibition, Mr Zink describes the demonstration as ‘the first time that DRM had ever been used in this way’. Namely, to transmit video signals.

Anyone who is familiar with Digital Radio Mondiale will already be aware of its rapid acceptance around the world, and the potential that it holds. Created to make better use of the limited potential of shortwave radio, the system takes frequencies below 30MHz and applies state of the art codecs to transmissions, allowing FM-quality audio to be broadcast across vast distances, while squeezing additional features such as electronic programme guides, the Journaline text news service and more into a very narrow bandwidth.

The concept – designed to use cheap existing infrastructures such as AM transmitters – is quickly gaining global momentum, offering broadcasters the potential to reach remote parts of the world where even the internet cannot be accessed without an often unrealistic level of investment.

‘Let’s say for example that you have loosely populated areas in parts of Africa, South America, Asia and Australia, and you want to cover that,’ explains Mr Zink. ‘Sometimes people don’t have satellite capabilities or the ability to invest in expensive equipment.’ Thanks to DRM, and more recently the more powerful DRM+, which pushes the system into the VHF bands up to 108MHz, people living in those remote areas are now in touch with the rest of the world.

As groundbreaking as the development of DRM has already proven, however, the system’s true potential is only now becoming apparent. Using DRM to broadcast radio into remote areas is impressive, but squeezing video transmissions into that same, narrow pipe is an achievement that few believed possible. Yet this is what Diveemo represents – the delivery of visual content to areas of the planet that were hitherto unreachable.

‘The idea came from the market – that’s always a good place to start,’ chuckles Mr Zink, describing why Fraunhofer and its partners – transmitter specialist Thomson and Chinese receiver manufacturer Chengdu Newstar – initially hatched the Diveemo project.

‘It was actually Thomson driving this initially,’ he explains. ‘Thomson is a manufacturer of transmitter equipment and they have representatives and connections all over the world. They got the feedback that this would be a fantastic opportunity because broadcasters were challenged with providing information services and they couldn’t economically cover remote areas. We’re talking about areas where the internet and so on is not available. That’s how it started.

‘Initially everybody just thought it was plain stupid to try to do video over DRM due to the very limited traits of the system. But we gave it a shot, we tried, and the initial tests were quite successful. That’s when we started to optimise the system and come up with a transmission method that doesn’t waste any bit-rate, so the maximum is available for useful content.’

The result is, explains Fraunhofer’s Matthias Rose, suitable for a very specific kind of broadcast. ‘Due to the frame-rate and the file-size, you wouldn’t want to watch a Hollywood movie on such a service and it’s not like a new mobile TV service – that’s not what we were aiming at. It’s more of a far-distance education service or a platform for warnings and messages about the latest developments, such as news broadcasts.’

In technical terms, he continues: ‘It’s around 10 frames per second – 60kb in total for DRM30. If you take DRM+ then you can go up to 186kb per second and you can use those bits for better video quality and a larger file with more frames. It’s scalable, not fixed, so if you have more bits available then you can improve the quality.’

‘Diveemo really is there to make use of the great potential of shortwave,’ agrees Mr Zink. ‘It’s there to cover huge areas and make information available where otherwise you couldn’t easily get coverage, at least not with equipment that is affordable.’

As for how the system actually works, the process of developing Diveemo was less about creating new technologies than applying existing codecs to the concept. ‘The signal over the air is obviously heavily compressed,’ he explains. ‘The system uses MPEG4 high efficiency AAC v2 for the audio coding and ABCH 224 for the video coding – these are currently the most efficient codecs.

‘At this moment we’re using a shortwave or medium wave double channel – DRM 30. That means Diveemo uses 18KHz or 20KHz of bandwidth over the air instead of a regular channel of 10KHz, which is still enormous but we’re trying to come up with additional codecs to have Diveemo transmissions in a regular, licensed single-channel over shortwave.

‘On the broadcast side,’ he continues, ‘Diveemo uses a regular video encoding solution and can encode from file or live sources. It is important to understand that the broadcaster can really heavily influence the quality of the content in terms of what the bit-rate should be, the quality of resolution and so on. If you have an education program and you’re more interested in broadcasting white boards, then you might opt for a larger resolution and have less frame-rate repetitions, whereas if you want to provide a news programme, then the frame rate might be more important than the resolution.’

Finally, Mr Zink adds that ‘the multiplexing is all taken care of at the broadcast side automatically – this is then what makes it Diveemo. On the decoding side, the additional functionality besides decoding regular radio is that the specific transport structure must be decoded, after which the receiver has the individual audio and video elements. These are MPEG standard so any MPEG standard decoder can be used to render it on the screen.’

Described in such terms, the system sounds complete, but Mr Zink is swift to explain that it is still in development, with the IBC demonstration having marked a milestone in Diveemo’s development rather than its culmination. Nevertheless, the emergence of the technology is likely to be swift.

‘We’ve demonstrated that its working and we’ve tried to optimise it and fine-tune it,’ he reasons. Adding, however, that ‘we can probably bank on a new MPEG codec coming out. But that will be a future development and it will be backward compatible so it won’t be a problem. At the last steering board of the DRM Consortium, which was held just before IBC 2010, the decision was taken to go forward with Diveemo standardisation, so now it is effectively a community task to put all of this development and technology into a standard paper and file it with the European Telecommunications Standards Institute (ETSI). That’s the goal.’

In commercial terms, that means that Diveemo could be introduced to the market, Mr Zink continues, ‘at the earliest next year. First we have to have the standard published and then people can start implementing the technology.’ As for the production of transmitters and receivers, Mr Zink proudly declares ‘we are quite advanced already’, as the Chengdu Newstar prototype that was used for the IBC demonstration proved.

Indeed, Fraunhofer is not shy in pouring praise onto its partners in the Diveemo project. ‘Thomson is a reseller of the Fraunhofer content server, which is used for getting a Digital Radio Mondiale signal on the air,’ explains Mr Rose. ‘They are the equipment manufacturer and we are the technology development company. The third partner, Chengdu Newstar, provides the chips. So it all comes together – we develop the technology, we have the transmitter side with Thomson, and we have the receiver side with Chengdu.’

The creation of a technology as unlikely as Diveemo could only really have happened in one place, however, and it is Fraunhofer IIS. The institute –bankrolled by a combination of public funding and revenue drawn from its developments, such as MP3 – exists primarily to facilitate the kind of big thinking required to push technology forward.

‘That’s our mission and also what our customers expect us to do,’ agrees Mr Rose. ‘They don’t expect us to come with technologies that are already in the market, they expect us to come with something that could be of potential revenue interest not in the next year but in the next three to five to 10 years.’

Even the new technology’s name has been poured over by a Fraunhofer team. ‘It is an artificial name which is loosely related to Digital Radio Mondiale,’ explains Mr Zink. ‘And it’s available! We didn’t want to run into some other technology or company that already used the name somewhere in the world. When you come up with a technology like this and try to find a name for it, you have to check worldwide, which isn’t easy to do. There’s also a team of people who check to make sure that the expression doesn’t imply anything negative in any native language.’

Meanwhile, Mr Zink is adamant that while Diveemo represents an enormous step forward for DRM, the video broadcasts are not intended to replace the audio applications for which DRM was initially created. Nor is the potential of DRM tapped – there is yet more thinking to be done at Fraunhofer.

‘DRM as a platform is always open for additional data applications, for example,’ he enthuses. ‘It contains all the hooks and mechanisms to transport any kind of data. That makes it quite an open and flexible platform for the future and I’m sure that people will find additional ways to make use of it.’

For now, however, Diveemo represents a use of DRM that could change the way that millions of people learn and live. While the big product launches and industry talk at IBC 2010 might have overshadowed the system’s debut, Fraunhofer’s latest creation could well prove itself to be the most far-reaching technology to have appeared in Amsterdam in years.

www.iis.fraunhofer.de