SVG Europe Audio: Weaving threads together at the Future Audio Forum

Next generation audio (NGA) is one of those woolly umbrella terms that covers a lot of ground. Encompassing everything from immersive audio to personalisation, a healthy variety of broadcast technology companies, standards agencies and broadcasters have been pulling at the threads of NGA for years.

However, there are so many threads that need to be weaved together to make it work. If it’s not delivery formats, it’s the need for common codecs, appropriate metadata, the effective monitoring and QC of audio objects, and the higher channel counts required to capture everything. And all that is before companies work out how to make everything more straightforward for the viewing public to use in the home.

Watch more Future Audio Forum – Personalisation, intelligibility and accessibility: Watch on-demand

In the second SVG Europe Audio Forum of 2024 the theme was Personalisation, intelligibility and accessibility, and many of these threads came together with a range of participants from across the broadcast ecosystem.

While all demonstrated a common understanding of what was required to move everything forward, this session marked a crucial shift. Not only were broadcasters on board, presenting real-life business cases to encourage faster deployment, but the presentation of a cast-iron argument examined how NGA will directly benefit over 1.5 billion people worldwide.

All the players

With representatives from every part of the production chain, it took SVG Europe Audio lead and editor Heather McLean a full two minutes to introduce all the players. The session boasted technology developers in the form of Dolby Laboratories’ senior director for AV ecosystems, Tim Carroll; Jünger Audio product manager Roman Rehausen, who works closely with France Télévisions on Serial-ADM (S-ADM) solutions; Dolby Laboratories’ senior staff architect James Cowdery; BBC R&D’s Dave Marston, senior R&D engineer; and Salsa Sound CEO Rob Oldfield.

It starred France Télévisions technology engineer, ISP exchanges department, Yannick Olivier talking about the broadcaster’s progress in providing live NGA content of the Roland-Garros tennis tournament in Paris, as well as its plans for the Olympic Games next month.

The discussion featured Felix Krückles, Professor for Broadcast Production Systems Design at the University of Darmstadt in Germany. Meanwhile audio consultant and former BBC Radio head of technology, Rupert Brun, shone a light on a use case which few people attending the meeting had even considered – neurodivergence.

With the widespread adoption of IP, the meeting kicked off with some good news from Dolby’s Carroll, swiftly followed by a challenge: “We don’t need to talk about codecs,” he said. “It’s done. It’s rolling out. It’s happening. The ease of something like Dante to connect things together was really hard not that long ago, but now things like immersive audio are just table stakes. Now it’s about how we do all that and provide accessibility?

“Since we met last time, there are more things happening throughout Europe and it’s even leaking into North America. Thank goodness we have IP networks; thank goodness we have S-ADM; now we can make all of that creative stuff actually happen. It’s exciting and I’m prepared for more of the hard stuff.”

Time for tennis

Five years after it was published, S-ADM has become the metadata format which facilitates a range of NGA projects, but the publishing of the SMPTE 2110-41 Fast Metadata standard in April is a huge development in that it enables NGA codecs like Dolby AC-4 and MPEG-H in 2110 networks. It is already encouraging more broadcasters to push the boundaries of personalisation, and France Télévisions’ Olivier shared a generous insight into the work that the French broadcaster has been doing with NGA at the Roland-Garros.

Olivier revealed that this work will develop futher at the Olympics, providing access to more signals to create more options for viewers of the tournament, using Junger’s AIXpressor technology to transport object-based audio and its associated S-ADM metadata.

Both Dolby’s Cowdery and the University of Darmstadt’s Krückles acknowledged that use cases like this are the best motivation for vendors to prioritise the development of equipment which can add more value. But as all the attendees learned, there are many more people who can benefit.

One fifth of the entire world

Brun’s presentation entitled, Enhancing Accessibility: Utilising Audio Objects for Neurodiverse Audiences, revealed that around 20% of people worldwide are thought to be neurodivergent and he highlighted the positive influence of audio personalisation on neurodivergent viewers.

Neurodivergence encompasses a range of conditions including autism, ADHD, dyslexia, brain injury and dyspraxia, presenting many sets of unique sensory needs. As neurodiverse individuals perceive and process sensory information differently, Brun explained how traditional audiovisual materials may not cater to the needs of neurodiverse individuals. Sensory overload, difficulties with focus and challenges with traditional learning methods are common.

Many like absolute quiet in the background, clarity, simplicity and good sound design, Brun went on. They may have difficulty with high pitched noises, high background noise, sudden changes in loudness and characters talking over each other. And although audio objects have well-established benefits for those with hearing and visual impairment, they can also deliver alternative versions of the audio and simple ways for neurodivergent viewers to switch between them.

There is little published research into the audio needs of neurodivergent audiences, Brun noted, and while it is acknowledged that personalisation through NGA systems can help, more research and trials are needed to understand the diverse needs of this audience [links to research are at the bottom of this article].

Tools which can help

All this puts more pressure on the audio mixers, and the session touched on ways that technology providers might deliver automated quality control (QC) processes in audio production. Salsa Sound’s Oldfield talked about how the company has been developing system to automate audio QC processes with artificial intelligence (AI), including detecting anomalies and early warnings, with a focus on improving speech intelligibility.

As the industry goes futher down this path, Oldfield also asked is there a scenario where the role of a mixer might morph into where the engineer is creating clean, monitored audio stems that then get sent down an S-ADM stream and are automixed as objects further down the chain. Krückels took the idea a stage further.

“In live broadcast 80% of the stuff we do in the control room is not mixing,” he replied. “80% is engineering the signal flow and 20% is just beautifying the sounds. Right now we’re talking about cleaning up the audio and mixing it automatically to create different presentations. In the future even the engineering path of getting microphones from a football match could be automated as well.”

On the right path

We are on the right path, but there is more work to be done and BBC R&D’s Marston stressed that the industry must continue to put pressure on manufacturers to support these workflows.

“I think organisations and companies who make audio equipment need to be a bit braver,” he said. “They need to implement all the latest S-ADM and 2110-41 related standards and not restrict equipment to 16 channels. Producers need to get into the habit of thinking in an object-based way in the future. Whether it gets used or not, it’s better to have metadata there. It can be ignored if it’s not relevant, but if it’s there it will always be helpful.”

In the meantime, it was noted by several panellists that the breadth of people in the conversation has changed over the last couple of years, and it’s no longer the technologists who have the reigns.

Oldfield agreed: “I’ve been involved in object-based audio for my whole career, and it’s great to see it moving forward,” he concluded. “It is important to get the buy-in from everybody in the audio ecosystem and I think that’s starting to happen. It’s going to be driven by the people with boots on the ground doing the audio mixing, as well as the networks understanding the needs of their end customers, but everybody in the chain has got to buy into it.

“I see that happening. We’ve got the whole ecosystem represented on this panel from vendors to broadcasters to operators and distributors,” he concluded.

Links to research into the benefits of audio personalisation for neurodivergent audiences:

Dr Lauren Ward & Samera Haynes-Khan A3i “Accessible Audio for Autistic Individuals” published a survey: https://screen-network.org.uk/project/a3i-accessible-audio-for-autistic-individuals/

Paul Obey’s Thinking Ahead podcast about recovery from brain injury: https://www.thinkingahead.uk/