Founder and CEO of Brandenburg Labs

The Birth of MP3—The MP3 format revolutionized the way people consume music. Could you take us back to the early days? What was the biggest challenge you and your team faced in developing this technology?

The biggest challenge by far was to ensure that the predecessor to MP3 worked for all signals, including music in its many diverse forms. While early versions worked quite well for certain types of music, they fell short for others. To solve this, we had to deepen our understanding of human hearing. The so-called masking effect had been known in science for long but not yet modelled with the necessary accuracy. Some sounds, even if physically present, don’t trigger nerve responses in our ears and are thus inaudible. So, we had to model this perceptual behavior within our algorithms. And we had to do all this while keeping the bit rate extremely low. That’s where advanced techniques from digital signal processing came into play. The whole process was iterative, new ideas were implemented, and we tested them on our (then)very slow minicomputers, which, compared to today’s processors built into mobile phones, etc. Then, a listening test was conducted to evaluate performance. Each test revealed new challenges, prompting further refinement. It took years before we reached a point where quality met our expectations. Interestingly, the human singing voice turned out to be the most challenging element to encode, perhaps because our brains are especially attuned to voices, given their central role in human communication.

Innovation in Audio – With nearly 100 patents to your name, you've consistently pushed the boundaries of audio technology. What excites you most about the future of digital sound, and where do you see the next breakthrough happening?

For a long time, the focus shifted from being able to duplicate the audio quality available on vinyl records or CDs to services with only a fraction of the original bit rate available. To reproduce sound in a room faithfully was the topic of research for around 30 years. For loudspeakers, this has already been scientifically solved a long time ago, and we enjoy great audio quality and audio effects in movie cinemas etc.

However, achieving the same immersive quality over headphones has been tried for over 50 years. Despite a lot of small progress, this had not been solved until recently. Headphone reproduction of sound in a room moved from in-the-head to near the head, but it didn’t feel truly realistic. Based on in-depth research into how our brain perceives spatial cues, we have developed a breakthrough. We finally managed to reproduce sound over headphones in a way that people can’t distinguish between the actual sound source or the virtual sound source over headphones. The spatial audio precision we achieved with our Deep Dive Audio technology is a real breakthrough that offers countless applications in terms of auditory virtual reality, professional audioproduction, education, and beyond.

Intelligent Headphones & Super Hearing—At Brandenburg Labs, you are developing intelligent headphones designed for "super hearing." Can you tell us more about this project and how it could redefine the way we experience sound in our daily lives?

By combining our spatial sound reproduction technology via headphones with artificial intelligence, we can do AI-supported selective hearing. Headphones for super hearing improve your hearing and make it even better by adding information if needed. Imagine being in a noisy airport or on a crowded train. These headphones can automatically amplify the sounds that matter, like someone speaking to you, while minimizing background noise. They can also provide useful contextual information, such as travel updates or destination details, without requiring you to look at your phone. We call them PARty headphones, short for Personalized Auditory Reality. These headphones will do to our hearing what glasses are doing to our visual sense. Over time, these intelligent headphones could become an essential everyday tool, being a nice gadget first, but at the same time improving the quality of life for millions.

The Ethics of AudioTech – As AI and audio technology evolve, there are increasing concerns about deepfakes and privacy. How do you think the industry should balance innovation with ethical responsibility?

I believe that innovation should never be pursued without responsibility. It is important that developers, researchers, and companies build safeguards into their systems and advocate for clear ethical guidelines. Transparency, consent, and traceability should be fundamental. Companies need to collaborate with inventors, policymakers, and educators to ensure that people understand how these technologies work and where the boundaries should lie. This does imply a basic understanding of technology on the side of the policymakers. Unfortunately, I often find that this understanding is lacking.

AI & the Future of Audio – Emerging technologies like artificial intelligence and machine learning rapidly transform many industries. In your view, how will these technologies reshape the future of the audio tech industry, from production to personalization and beyond?

Machine learning and AI are basic tools which can be applied at the heart of new systems, like understanding of the acoustic environment and the noises around us in a PARty headphone. At the same time, they are practical tools helping in writing etc. But I don't believe everything automatically becomes better just because AI is involved. Just as in my field, new programming tools over the last decades have massively improved the ways to get from ideas to working code, AI will help us become more productive if we use it in the right way.