Berlin-based ai|coustics, a startup that is democratising audio, has announced a €5 million seed funding round. The investment was led by new investors Partech, with Acurio, Intuition and Arc joining. Existing investors Connect and FOV Ventures, and angel investors including Mehdi Ghissassi (ex-Google Deepmind and CPO at AI 71), Gert Lanckriet (Head of Machine Learning, Amazon Music), Hazel Savage (Former Co-Founder and CEO of AI Company Musiio) and Thomas Wolf (CSO, HuggingFace) also renewed their participation. The startup will be using the funding to expand its platform and products.
With the AI sector heavily focused on generative uses, ai|coustics stands out for its commitment to retaining authenticity and using AI to create studio-quality voice recordings. TFN asked ai|coustics co-founder and CEO Fabian Seipel about the platform and the impact it will have.
The power of audio in creation
Audio is the unsung hero of creation. We all carry phone cameras capable of cinema-quality video in our pockets, but when we watch it later, it’s usually the sound that lets it down.
Many AI platforms address this with generated voices, but ai|coustics is taking a different approach. Seipel explained, “We believe that in a world of generated content, people will still be interested in creating and broadcasting original content using their own voices, and will want quality content rather than quantity.”
ai|coustics was founded in 2021 at Technical University Berlin by Corvin Jaedicke and Fabian Seipel, both experts in audio technology and machine learning. The company aims to revolutionise audio quality by leveraging AI to enhance speech clarity and intelligibility across various digital platforms, including voice AI, broadcasting, gaming, and digital communication
The ultimate goal is to set a new standard for high-definition audio quality, making studio-grade sound accessible to everyone, regardless of the environment or device used.
Their technology is designed to automate parts of the audio production process, saving time and resources for content creators and audio engineers ai|coustics uses generative AI to improve audio quality beyond just noise suppression, aiming to make every digital interaction sound like a professional studio broadcast.
Many AI platforms address this with generated voices, but ai|coustics is taking a different approach. Seipel explained, “We believe that in a world of generated content, people will still be interested in creating and broadcasting original content using their own voices, and will want quality content rather than quantity.”
Instead of recreating the human voice, ai|coustics’ platform focuses on authenticity. ai|coustics retains the original voice and instead recreates the audio, using its AI engine to remove background distractions and audio artefacts. It brings the focus back to the spoken words, with the timbre of the original speaker preserved.
The company’s AI leverages advanced machine learning algorithms to analyse frequency distribution, noise levels, and dynamic range before applying intelligent processing. This approach allows for nuanced enhancement that preserves the authenticity of the original audio while significantly improving its quality.
Breaking down the production barriers
Sound editing can be incredibly difficult. While human ears — and the brain between them — are incredibly good at filtering out extraneous noise and adjusting for conditions, current hardware is not. High-quality audio relies on professional equipment to capture the sound and experienced editors to enhance it afterwards.
ai|coustics focuses on transforming regular recordings into studio-quality audio across various platforms, while competitors like Cleanvoice AI are more specialized in podcast editing. Each competitor offers unique features, such as Xound’s pitch correction or ReMasterMedia’s echo cancellation, which cater to different user needs. Tools like Adobe Enhance Speech integrate well with existing software suites, while others like Xound focus on simplicity and ease of use.
But ai|coustics has found that their platform isn’t just being used for content. “Users have also used our tech to pre-process audio for voice cloning and automatic speech recognition, especially for acoustically challenging environments,” Seipel told us.
But whatever the end-use, improved audio is the result. “The market standard for sound quality is ever-growing,” said Partech General Partner Boris Golden, “but the average quality of voice recordings is still very poor. Bringing software-enabled ‘studio-quality sound’, without the need for expensive hardware or technical expertise, is the most efficient way to solve this growing problem at scale.”
Tech names like Elgato, DHD.audio, and Infineon Technologies have all partnered with ai|coustics. “For many customers, the idea of upgrading their product, especially the audio quality, with a few lines of code rather than redesigning the hardware is very appealing,” Seipel said. End users have also found the platform useful, and have already enhanced well over two million minutes of audio using it.
But even those who already have the highest quality audio equipment are turning to AI. “Our technology is a classic case of software eating hardware,” Seipel explained. “We are solving an audio hardware and acoustics problem with software, similar to smartphone cameras with AI processing.”
ai|coustics’ first clients, including national broadcasters Deutsche Welle and Radio France, are benefiting from the optimised post-production workflows made possible by AI, helping them to filter out the extraneous noise.
ai|coustics: an audio revolution for every user
Although initial users may be focused on creation, ai|coustics’ platform offers significant potential for wider use. In an age of online meetings, we might soon be free of hearing background conversations about household pets and meal preparations. And those with impaired hearing can benefit from audio that focuses on the human voice. Indeed, part of the founders’ inspiration came from their own mild hearing impairments, a result of careers in music production and playing in bands.
The company is using the funding to expand their platform and products. “We are looking forward to more commercial partnerships, doubling down on our real-time SDK as a quality layer for both human-to-human and human-to-machine voice quality, and introducing new features on top of our voice quality engine.”
They aim to be at the forefront of audio production. “The barriers to creating high-quality content are falling. Various AI tools are now the next step in content creation, but with a higher rate of technological development,” Seipel told us. “We aim to become an integral part of the Voice AI space, representing authenticity and quality.”
The post Bringing authenticity to voice: how ai|coustics is transforming sound appeared first on Tech Funding News.