Post Magazine

January / February 2019

Issue link:

Contents of this Issue


Page 15 of 51 14 POST JAN/FEB 2019 TECHNOLOGY ounded in 2003 in Paris, Audionamix offers audio sep- aration solutions to the post produc- tion market. In its first few years of development, the company's audio separation technology was used for vocal extraction, dialogue isola- tion and stem creation. Soon after, Audionamix opened an LA branch and started providing separating services to Hollywood-based film and music studios, and many other prominent companies. Here, Audionamix CEO Maciej Zielinski speaks exclusively with Post about how the company established itself as a global leader in the audio source separation market and how it's looking ahead towards the growing role artificial intelligence is playing within the audio post community. Tell us a little bit about your background in audio and when you were first introduced to artificial intelligence (AI). "I am a composer, songwriter and music producer, and an owner of Soundflower Studios, a Warsaw-based post production facility. I've composed music for 20 feature films, six TV series and a few hundred commercials. I am a huge fan of music technology. I've been using most of the software and hardware solutions since the mid- '80s, and I am very passionate about all the latest audio tech- nology. I've had some basic knowledge about AI ever since Deep Blue won the chess match with Garry Kasparov in 1997. I got more into the practical use of AI and the theory behind it when I started to work with Audionamix in 2014." How did Audionamix professional separation services originate in the film and music industries? "Many content owners couldn't repur- pose their older titles because of certain problems; for example, in many cases they didn't have the stems or multi- tracks, only final mixes of their TV series or feature films. So, we offered them our separation services. For example, we could remove existing music that was too expensive to re-license from a full TV series, so that they could license and replace cues with more affordable alter- natives. We also removed music so that iconic films such as North by Northwest could be screened with a live orchestra performing the score. Over the years, our separation services have been uti- lized by Warner Bros., Disney, Universal, Paramount, Canal+ and others. We also collaborated with record labels, sepa- rating vocals, helping to produce duets across the ages, including Rod Stewart's 2012 virtual duet with Ella Fitzgerald from his "Merry Christmas, Baby" album. A wide range of creative projects re- quire audio separation services and we still offer these unique solutions." When did Audionamix move from a service provider to a business that offered tools/products for audio source separation? "In addition to offering our profession- al services, we decided to convert our technology into software, and in 2014 we released Trax, the first commercially available vocal separation software. For the first time, users could separate vocal parts from any song to troubleshoot or create new productions. Since this groundbreaking initial release, we've expanded our standalone software offerings to include speech separation as well as stem separation. XTrax Stems, our most popular software, has recently become a mainstay for DJs and remix- ers. It can automatically separate any song into three stems: vocals, drums and remaining music, stimulating creation of new exciting productions. "Our latest release, the IDC: Instant Dialogue Cleaner plug-in, is a very important tool for the post communi- ty because it separates speech from background noise in realtime. By using a simple one-knob interface, IDC can instantly reduce and remove unwanted noise. We approach this de-noising from the other side, by using machine learning artificial intelligence to sepa- rate and preserve speech, allowing for many different types of interference to be reduced or removed from dialogue recordings. IDC was recently nomi- nated by the Cinema Audio Society for Outstanding Product for Post Production." What kind of role do you see artificial intelligence (AI) playing in the pro audio post market and where you see it going in the future? "AI is empowering a variety of applica- tions in the audio world. Audio source separation is a great example. AI has increased the quality of our separation techniques as we continue to train our ARTIFICIAL INTELLIGENCE AND AUDIO BY LINDA ROMANELLO AUDIONAMIX DISCUSSES AI'S ROLE IN AUDIO TOOLS F "AI IS EMPOWERING APPLICA- TIONS IN THE AUDIO WORLD. AUDIO SOURCE SEPARATION IS A GREAT EXAMPLE."

Articles in this issue

Links on this page

Archives of this issue

view archives of Post Magazine - January / February 2019