BEDS, OBJECTS, AND NEW TOOLS FOR IMMERSIVE AUDIO PRODUCTION

How do you mix music in Dolby Atmos? What are the technical differences compared to a stereo or even 5.1 mix? And, above all, what tools and approaches are best suited for this format?

In this post, I want to share some essential technical concepts of the Atmos workflow, address them from practical experience, and also offer a critical look at their implications.

BEDS AND OBJECTS: A NEW SPATIAL LANGUAGE

One of the first things to understand in Dolby Atmos is the difference between beds and objects.

• Beds: These are traditional channels grouped together—generally 7.1.2, which comes from cinema but also scales to formats like 7.1.4, 9.1.6, and 11.1.8—that function as a fixed base within the sound field. Here, we could say that it usually contains the material that we want to maintain its spatial stability, such as the drum, bass, or fixed effects stems.
• Objects: These are mono or stereo tracks to which dynamic XYZ coordinates are assigned in a three-dimensional space. These objects can move, change position over time, and sound differently depending on the system on which they are played.
• oBeds (Object-based beds): This is a hybrid technique where a custom “bed” is built using static objects. Instead of sending multiple signals to a traditional bed, individual objects are created that occupy the exact same positions as the bed’s channels. This allows the content to have the same spatial behavior as a traditional bed, but with the additional control that objects provide. It is especially useful when you want to preserve a feeling of a “fixed” mix but with advanced rendering or specific automation options.

BINAURAL RENDERING

One of the most discussed challenges is how an Atmos mix translates to listening on stereo headphones. For that, we use binaural rendering, which simulates the three-dimensional space using HRTF (Head-Related Transfer Function) filters.

The Dolby Atmos Renderer allows you to apply different binaural modes for each object: “Off”, “Near”, “Mid”, and “Far”. These modes affect the behavior of the signal during the binaural rendering process and work as follows:

• Off: The object will not be included in the binaural spatial field. The signal will be heard centered in the middle, without spatiality.
• Near: The sound is perceived very close to the listener. This mode uses an HRTF response with less arrival delay and a higher direct signal/reverberation ratio. The simulated reverberation time is short (generally < 100 ms), and the signal feels intimate, as if it were a few centimeters from the head.
• Mid: The object is positioned at a medium distance. There is a moderate increase in reflection and delay, which generates a more open sense of location. The simulated reverberation time can range between 150–250 ms.
• Far: The sound seems to come from a great distance. The algorithm adds a greater arrival delay, a greater ratio of reverberated vs. direct sound, and a more dispersed HRTF response. The simulated reverberation time can exceed 300 ms, generating an immersive but more diffuse perception.

SPECIALIZED PLUGINS: NEW TOOLS AND NEW PERSPECTIVES

Mixing in Atmos requires tools designed for a three-dimensional environment. Some of the most prominent in my workflow are:

• Genelec GLM: For those who have physical multi-channel systems, speaker calibration is essential. Genelec’s GLM system allows you to align levels, times, and frequency response in 7.1.4 or higher configurations. I have implemented this tool in “safe” places like our studio but also in live situations such as the Organysmo project in Monopol, Berlin, where we practically used the speakers in a resonant bunker of 15×15 meters of brutalist concrete. Without GLM we would still be looking for a way to calibrate the system…
• Sound Particles: This software allows you to model moving sound sources using trajectories, dispersion, virtual mass, and randomness. It is ideal for designing dynamic and organic scenes. They also have a series of plugins for panning creatively and even SkyDust 3D, which is the first spatial synthesizer to hit the market. It is super interesting how we can create new narratives with tools like these.
• LiquidSonics – Cinematic Rooms: A reverb designed from scratch for multi-channel spaces. It allows you to position reflections and adjust spatial behavior with precision. Initially, we had to emulate all this with multi-mono concepts, imagine the processing load, although it worked, this solves many things.
• Audiomovers – Listento: It is not exclusive to Atmos, but it has become crucial for sharing immersive sessions in real-time, especially when working with clients or collaborators remotely. As well as for creating custom routings with tools like the Omnibus.

ARTISTIC DECISIONS VS FORMAT LIMITATIONS

One of the great temptations when starting to mix in Atmos is to move everything, fill the space, demonstrate that we are working in 3D. But in my experience, less is more and each project has its requirements.

The space must be functional to the content. If all the elements are in constant motion, the auditory reference is diluted and the focus is lost. For me, it is important to maintain the idea that technology should be at the service of the work and not the other way around.

Furthermore, not all genres or musical styles benefit in the same way. In electronic productions, environments and textures can take on a new life. But in acoustic styles, a good stereo mix with a good room is often more effective.

Mixing in Dolby Atmos implies a change not only technical, but conceptual. It requires rethinking the spatial narrative, the sound design, the final listening experience, and the use of the correct tools, but it also requires understanding its limits: from the quality of the binaural rendering to the difficulties of translating a speaker experience to headphones.

As with any tool, the important thing is to know when and why to use it. The potential of Atmos is enormous, but its effectiveness depends on the artistic criteria with which it is applied. We are in a very entertaining process where developers, manufacturers, creatives, and professionals from multiple areas are working on these developments. Nothing is yet taken for granted, so it is an invitation to continue exploring and contributing to this developing medium.

Christopher Manhey.-