A high-tech industry like broadcast television has always been a hotbed of technical innovation. But never more so that today, when the growth of digital media has resulted in a huge number of technical innovations - from media management systems to visual effects software.
To keep up with the fast-moving pace of change, well-funded research and development (R&D) is vital in keeping UK broadcasters and post-houses at the cutting edge, with most making substantial annual investments in developing new, more powerful or more efficient solutions. Take The Farm Group, for example, which has developed a reputation for innovation. Says director of operations Dave Klafkowski:
'We have a dedicated software developer, so we're very committed to R&D.'
One successful product is its purpose-built media management and logging system for live reality TV shows. The fully networked and edit-platform independent system has been used on all the streamed reality shows that the group has worked on - including BBC3's House of Tiny Tearawaysand Channel 4's Space Cadets.
'A producer or director can use it to cut a rough story out of rushes which they can then post anywhere,' explains Klafkowski. 'It's been in ongoing development for the past two years, with the basic bones of the database management staying the same but with parts of the system modified as each job requires it.'
This kind of continuous development can be seen at effects house The Mill, where the internal visual effects pipeline tools, render farm as well as software plug-ins and Mental Ray rendering Shaders, have all been subject to some heavy R&D by the company's five full-time developers. 'All projects we do will include some facet of our development work,' says The Mill's CG technology director, Dave Levy, who adds that R&D on rendering in particular allows the facility to take on projects which would otherwise be impossible - such as effects on the second series of Doctor Who. 'In terms of productivity we are able to render multiple passes simultaneously, which can speed up the process by an order of magnitude,' he declares.
Software developers also aim to make life a little easier for facilities through R&D. One such is Root 6, whose most successful product to date is ContentAgent, a one-box solution for streamlining the capture, multi-format encoding, management and distribution of digital media. According to Root 6 managing director Marcus Hume Humphreys it was developed because customers such as the BBC, MTV and
C4 were struggling to meet changing demands for deliverables. 'We already had an R&D team in place for earlier projects like the [tapeless video distribution system] BeamBox which we created with BEAM.TV,' says Hume Humphreys. 'About 15% of our employees are software developers, with the cost of running the R&D department accounting for about 30% of our total.'
Like similar manufactures in the broadcast sector, the Root 6 R&D team usually starts by defining the scope of a project and drawing up detailed specifications, and then assigns various coding tasks to individual developers.
The typical development cycle for a new integrated solution product at Root 6 is between 18 and 24 months, but even for a software-only outfit, such as the Foundry, development times can be lengthy. This London-based software developer took two years of R&D to deliver a shippable version of its Furnace set of 11 image-processing plug-ins for post-production systems. It was initially developed six years ago when clients such as the BBC, The Mill, Pepper Post and Skaramoosh voiced their concerns about the rapidly growing number of VFX shots being requested on productions. 'It was becoming impossible to deliver everything being asked of them whilst keeping projects within a reasonable budget and time frame,' says Dr Bill Collis, Foundry commercial director and head of research and development. In response the Foundry started to analyse sequences submitted by these companies, identifying the most recurrent workflow blockages with a view to engineering automated solutions.
'Before we release any product we test it to smithereens,' says Collis.
'For example, The Moving Picture Company was having problems with wire removal on a Jackie Chan film at around the same time that we were developing our WireRemoval plug-in, so it provided the perfect forum for putting it through some pretty heavy-duty testing.'
VFX facility Time-Slice Films also works with the Foundry to hone its range of products. 'We always test alpha and beta products on real jobs that come in,' says managing director Tim Macmillan. 'You have to have that deadline and the pressure. It works well reciprocally with the Foundry as it gets nice clips of the product in action.'
Macmillan is certainly no stranger to R&D, being inventor of the Time Slice frozen-time effect seen in everything from VFX sequences in Hallmark movie Merlin(1998) to ITV's recent drama-doc Inside a Bank.
He continues to push the boundaries of imaging technology, ploughing most of the profits back into research. But for Macmillan it's not all about coming up with innovations and patenting the results. 'I've never thought it useful to try and patent something that a couple of years down the line will be superseded by something else,' he says. 'It doesn't seem practical in visual effects. I'm not a manufacturer or software developer. We develop our own technology, which we keep in-house, to set clear blue water between the competition and ourselves.'
It's a view shared by the Pixel Farm, another software company heavily focused on research. 'We don't tend to look at what everyone else does and then try and do the same, we look for new ways of doing things,'
says managing director Richard Spöhrer. Like the Foundry, the Pixel Farm's products have been road-tested on high-budget film projects - the company tuned the capabilities of its PFTrack plug-in (a camera tracking and motion analysis tool) when working with Stan Winston Studios on some tricky head replacement shots for its 2005 film Fantastic Four.
According to Spöhrer, development time on solutions can be anything from a few days to a few months, with R&D costs met by the customer if they want to keep what is done solely for themselves. The company, on the other hand, will absorb costs if the work is something that the Pixel Farm is going to roll into a release of a product with wider applications.
Says Spöhrer: 'You have to be very careful when working closely with facilities as you can end up with an application that suits them and no one else. As well as building great tools we try and design workflow and process-based applications that can work in any environment.'
According to Spöhrer finding funding for R&D can be difficult. 'The grants system is a mess,' he says. 'To be honest starting up a technology business in the UK is about as hard as it gets. What is needed is a clear system of tax breaks for innovation and research.
Often grants are designed to focus your research in one area and can end up restricting what you do. In most cases grants, especially from the EU, are built around co-operation with a university and other commercial companies.'
This was the route taken by Time-Slice Films, which partnered with Snell & Wilcox and Bristol University on a Department of Trade and Industry-funded two-year project. 'It was some blue-sky work looking at getting 3D information out of a sequence,' says Tim Macmillan. 'It was aimed at developing products, but it's not really long enough to enable you to bring a product to market.'
Televirtual, which designs real-time animated solutions for interactive TV, has on the other hand successfully leveraged around £3m of public funding from Europe and UK sources. No less than five EU technology projects, followed by a DTI/EPSRC (Engineering and Physical Sciences Research Council) project in set-top box content, a Lottery grant and a SMART award has come the way of the company in the past 10 years.
'The truth is we never made a fortune out of funded R&D,' says Televirtual managing director Tim Child. 'Take EU-funded projects - if we agree to deliver our part of the project then we'll get about 70% of our effort in R&D paid for in return. However, that only applies when the goodies your project has developed have a real commercial value. If they don't then in reality you're going broke to the tune of 30% each time you complete a project.'
Top five industry R&D software products
Massive
This 3D animation system for generating crowd-related visual effects was developed by Stephen Regelous for Weta Digital to create the battle sequences for the Lord of the Rings. Now used commercially for commercials, film and TV.
Automatic Duck
Invented by software engineer father and film editor son team Harry and Wes Plate, the Automatic Duck plug-in is the leader in timeline translation between most popular NLE and compositing systems - allowing a Final Cut Pro to talk to a Media Composer.
Holoset/Chromatte
Developed by Reflec in conjunction with the BBC, Chromatte, also known as Holoset in the US, is a chromakeying solution for virtual sets that causes only the light shined onto it to be reflected back to its source.
Kronos
Cross-fertilised from research by Time-Slice Films and the Foundry's advanced motion estimation technology, Kronos builds in-between frames in a sequence and can be used to speed up or slow down an image sequence. Recently used to great effect in films such as Batman Begins, King Kongand the Harry Potterseries.
Nuke
Short for New Compositor, this began as the in-house compositing and effects application at Digital Domain to pull keys, colour correct, combine, transform, and filter images.
Kingswood Warren - Calling time at the BBC's high-tech home
It's the end of an era. Kingswood Warren, the BBC's famous R&D site which has played a huge role in some of broadcasting's biggest innovations, is to be sold off when the R&D division's 160 staff move to Manchester by 2010.
A 19th-century mansion in leafy Surrey, the site has played host to the first demonstrations of colour television and high-quality sound from magnetic tape as early as 1951. In the same year boffins developed the first television standards converter, using a special camera tube and a long-persistence CRT picture monitor. Major innovations were almost an annual occurrence. In 1959 the first transatlantic television transmission took place.
Early Ceefax tests were carried out in 1971, while 1978 saw digital stereophonic sound experiments that lead to invention of Nicam.
High-definition TV studies commenced in earnest at the department in
1982 and an HDTV picture store and high line-rate picture monitor was demonstrated in 1986. The 1990s saw pioneering work on digital radio and television broadcasting and the launch of the BBC's website.
'Work on virtual studios started at BBC R&D back in the early 1990s,'
says Graham Thomas, project manager, BBC R&D at Kingswood Warren. 'We developed some of the key bits of technology that made such systems practical. For example, our free-d camera tracking system, which is manufactured under licence by Vinten, is now used by many broadcasters around the world, and has also found its way into film production.'
This led to the formation of the Production Magic team, which contributed to the RTS award-winning BBC's General Election Results Programme 2005 and has contributed advanced on-air graphics and 3D synthetic elements to many programmes since.
Although the R&D workforce is to leave Surrey for London and Manchester over the next few years, work still continues on spectrum planning for TV and radio, digital TV services, electronic programme guides, and wireless camera technologies.
BBC Technology Group
No comments yet