The architectural distinction between beds and objects is the defining feature of Dolby Atmos, and understanding it is fundamental to mixing immersive audio correctly. Beds and objects are not interchangeable — they serve different purposes, have different technical behaviours, and make different demands on the mixer. Getting this wrong produces mixes that don't translate across playback environments.

What Is a Bed?

A bed in Dolby Atmos is a static, channel-based audio layer that maps directly to specific speaker positions. Beds are defined in discrete channel configurations: most commonly 7.1.2 for film and OTT work. Audio assigned to a bed channel always plays from that fixed speaker position regardless of the playback system — the renderer doesn't alter the bed's channel assignments when adapting to different room sizes or speaker configurations (though the renderer may apply downmix coefficients when rendering to systems with fewer speakers than the bed format).

Beds are ideal for: diffuse ambience and backgrounds, reverb returns, sustained music content, and any sound source without a specific, trackable spatial position. Room tone, for example, belongs in the bed — it doesn't have a point of origin, it has an environmental character. Putting room tone on an object is technically possible but wastes object allocation and creates unnecessary metadata complexity.

What Is an Object?

An Atmos object is audio with associated spatial metadata — an XYZ position (and optionally size, gain, and zone restrictions) that the renderer interprets dynamically for the specific playback environment. Unlike beds, objects are not locked to speaker positions. In a cinema with 64 speakers, an object flying overhead will be rendered across multiple ceiling speakers to create a precise trajectory. On a consumer soundbar, that same object will be rendered using psychoacoustic virtualisation. The metadata is preserved; the rendering adapts.

Objects are ideal for: specific sound sources with clear spatial identity (helicopters, weapons, vehicles, on-screen characters), moving sounds with automatable trajectories, and height-active content that should feel positioned above the listener rather than diffused across the room.

How Many Objects Can You Use?

The Dolby Atmos specification supports up to 128 simultaneous audio objects. In practice, the internal Pro Tools renderer has a lower operational object limit that varies by system performance. For most film and OTT work, 16–32 active objects per reel is sufficient. Using more objects does not automatically improve the mix — each object should be justified by a specific spatial narrative purpose.

One common beginner mistake is objectifying everything — putting all dialogue, all effects, all music on objects. This creates excessive metadata complexity, strains the renderer, and paradoxically reduces mix quality because the renderer must make constant spatial decisions that a skilled mixer would handle manually. Use objects for spatial storytelling; use the bed for acoustic foundation.

Object Automation and ADM Export

Object automation in Pro Tools is written to the Atmos renderer's ADM (Audio Definition Model) metadata. When you automate an object's X/Y/Z position over time, that automation is embedded in the printed ADM BWF file and reproduced on every compliant playback system. This is the format that Netflix, Apple TV+ and Amazon require for Atmos deliveries — the ADM master carries both audio and spatial metadata.