c’t spoke to the music pioneer Jean-Michel Jarre on the occasion of his new album “Oxymore”, one of the largest audio productions in 3D to date.
The French pioneer of electronic music Jean-Michel Jarre (74) will release his new album “Oxymore” on October 21 – one of the most elaborate productions in 3D audio ever. c’t spoke to Jarre about the changes 3D audio is bringing to musicians and the technical hurdles the industry still has to overcome when it comes to playback.
c’t: Typically, musicians first compose, arrange and mix their tracks in stereo and then convert them to 3D audio. Did you do the same with “Oxymore”?
Jean Michel Jarre: No, I composed the music for Oxymore in a 360-degree setup with 7.1 speakers and later created a stereo version from it in addition to the multi-channel mixes. There is no stereo in nature. All the sounds we hear come from mono sources around us – for example when we are talking to each other here or when a car drives by. Therefore, 3D audio and binaural headphone playback come closer to natural hearing than stereo playback.
c’t: Which digital audio workstations and surround sound software did you use for the production of “Oxymore”?
Jarra: I’m used to Ableton Live, so I used it to create the mix and finalized it in Avid Pro Tools. For the surround sound we used L-ISA Studio from L-Acoustics, which we also use for my live shows. Finally, we created the multi-channel versions of Oxymore with Steinberg Nuendo. It converts various output formats such as Dolby Atmos and all conceivable speaker configurations from a 3D project. In this area, Nuendo is currently by far the best tool on the market.
c’t: What headphones have you used for mixing and mastering, and what would you recommend listeners use for the binaural version?
Jarra: In the studio, I test the mixes with neutral, open models such as the Ollo S4X and the Sennvoonzer HD 600. My team and I tested audience playback with almost 80 sets of headphones. Many models have too much bass, such as Apple’s AirPods Max. Sony’s MDR-7506, Bose 700 and Apple’s AirPods sounded good, for example.
c’t: What do you have to consider when composing and arranging in 3D and what changes compared to a piece in stereo?
Jarra: In stereo I have to arrange sounds like a painter in ever new layers in 2D in front of me. If I apply too many layers, everything turns gray slush. That’s why the principle applies in stereo: Less is more. In 3D, on the other hand, each sound occupies its own space. There it can certainly apply: More is more and maybe even better. As a musician, this gives me freedom for completely new approaches.
c’t: Their composition is quite dense, something is happening all around the listener. Did you omit or change elements when downmixing to stereo?
Jarra: The stereo mix contains all the elements of the 360-degree versions, so I didn’t leave anything out. In stereo, however, you have to mix differently than in 3D because the sounds merge differently. Stereo offers less space, frequency ranges of individual tracks overlap more strongly. You have to take that into account in the mix, especially in the bass and lower mids.
c’t: What do you think of turning old recordings into 3D?
Jarra: I’m not a fan of converting stereo recordings to 3D. It just doesn’t work. However, the industry thinks they can convert the old catalog of stereo albums into 3D. But that is a big mistake. From a creative point of view, you have to conceive the music for this new medium from the beginning. There’s no need to remix Frank Sinatra in 3D to make his voice circle around your head. Nobody cares.
c’t: But there are quite a few jazz recordings, for example by Blue Note, that sound very convincing in Dolby Atmos…
Jarra: That’s a matter of taste. Charlie Parker and John Coltrane were recorded in mono with large microphones. It was a different technique back then. It would also be nonsensical to digitize a Rembrandt painting in 3D. I see no point in that.
c’t: How satisfied are you with Dolby Atmos?
Jarra: We are still in the dark ages of three-dimensional music production. So far I haven’t been a big fan of binaural mixes. For three reasons: First, the instruments blend together differently. As a result, binaural mixes lose dynamics in the bass and lower mids. Second, Dolby Atmos mixing tools were not designed for music, but for movie theaters. There the dialogue comes loudly from the front in the middle, the rear channels are quieter. In music, however, we have an egocentric listening situation in which you sit in the center and all speakers around you are equal. Third, the filters used for binaural conversion for headphones were also designed for movies. If I shift the direction of a snare 30 degrees, its sound changes because of these filter characteristics, which you have to compensate for.
Dolby is aware of the issues and works closely with musicians. At the moment, however, we musicians do not have access to the necessary tools to adapt the binaural transfer filter to our needs, for example.
c’t: And what about Sony’s 360 Real Audio?
Jarra: Sony’s 360 Real Audio was developed by Fraunhofer. For speakers, it’s the best 3D format on the market. But the transfer filters for headphones are bad. We produced a test version of my album “Oxygene” in 360 Real Audio and it didn’t sound good at all. It’s a mystery. I spoke to Fraunhofer about it, but we don’t know why. You absolutely need to improve this.
c’t: The streaming services from Apple, Amazon and Tidal offer 3D music in different formats. As a musician, how do you get to grips with the differences?
Jarra: Apple wants to convince listeners that you can convert any stereo mix into 3D with their automatic and that it sounds better afterwards. However, that is not true. They boost the bass, add some reverb and increase the volume. The listeners then think it sounds better just because it’s louder.
In addition, the playback of 3D music on the streaming services differs due to the different processing by Apple, Dolby and Sony. In order to compensate for these format problems, we had to produce a separate 3D mix for “Oxymore” for each platform. That is a big problem. It is currently almost impossible for a studio to create a suitable multi-channel master for each of these platforms and formats.
c’t issue 22/2022
In c’t 22/2022 we give you tips and tricks on how to unmask power guzzlers with smart technology. We also put the new AMD CPU Ryzen 7000 to the test. Also in the test are the inexpensive mini barebones NUC11, markdown editors and nine programs for better headphone sound. We show how to manage image collections in macOS and how the post-quantum cryptographic selection process currently stands. You can read that and much more in the current issue of c’t.
Save electricity with smart technology
AMD CPU Ryzen 7000 in the test
Windows 11: 22H2 update tested
Child protection for Android & iOS
Inexpensive mini barebone NUC11
Markdown text editors
E-payment: debit cards in comparison
Practice: Manage image collection in macOS
Caution customer: Bose spoils earphones
FAQ: MS Office: Locked VBA macros
c’t 22/2022 in the Heise shop