Dolby Atmos for Home Cinema How is it Different?

To better understand Dolby Atmos and its potential integration into the consumer market, it's important to  consider how its different than the  “classic” 5.1 and 7.1 surround decoding schemes. In 5.1 and 7.1 mixes the individual channels of similar content are mixed to “stems” of 5.1 or 7.1 content. For example “music”, “dialog”, “adr” (additional dialog recording, that is dubbing the original dialog), “ambient” (sound of birds in a park, the roaring sea etc.), “foley” (sound effects recorded in a studio as footsteps, door slamming etc.), “fx” (artificially designed sound effects as laser beams, screaming monsters etc.) and others. Inside a stem, the individual channel can be “surround-panned” from the center to any of the extremes (from -63 to 63 on two axes: X and Y, so Front/Rear, Left/Right). In the end of post-production these “stems” are mixed together to a final 5.1 or 7.1 mix, the end product your hear when watching a movie.

Dolby Atmos is not just a system with more channels than 7.1.  It works with three primary elements:

1.      Bed Audio

2.      Object Audio

3.      Metadata

Bed Audio: these are channel-based stems, again in 5.1 or 7.1, so basically the same as in former surround-formats. The individual channels are static inside the bed.

Object Audio: these are mono- or stereo channels with dedicated surround panning. These channels are not in a bed, but remain individual.

Metadata: Surround panner metadata for Object audio and additional metadata

Illustration of Dolby Atmos Object Audio

These “Objects” can be surround panned into the space (from -100 to 100 on three axes: X, Y and Z, Front/Rear, Left/Right, Up/Down) and they can be given a “Size” (from 0 to 100, singularity to fill-the-room). The Metadata defines where the objects are positioned and how big they are in time. In normal surround these surround pannings are transformed in output levels of the classic 7.1 mix buses: L, C, R, Ls, Rs, Lsr, Rsr and LFE. The Objects on the contrary are “rendered” by the Atmos engine, to the actual number of speakers it has to drive, up to a maximum of 64 individual speaker feeds. So if there are “only” 32 speakers present, Atmos will render the Object audio to 32 speakers. These up to 64 speakers are divided in speaker arrays just like in classic surround L, C, R, Ls, Rs, Lsr, Rsr and LFE, with the addition of the newly introduced speaker arrays Lts and Rts (Left Top Surround and Right Top Surround).

So what is the minimum number of speakers to “get” the Atmos-certification in a cinema theater? The minimum requested number of speakers in the smallest theaters require 3 front speakers (L, C, R), 4 surround speakers on every side wall (Ls, Rs), 4 surround speakers on the back wall (Lsr, Rsr) and 4 surround speakers for both top speaker arrays (Lts, Rts). And of course a powerful subwoofer for the LFE channel and, if the mounted surround speaker model of the theater is so small it doesn’t reach 40 Hz, a minimum of 2 surround-subwoofers is required.

Typical Dolby Atmos setup for big theaters. Note the Side Wall Speaker Arrays start much more in the front as in “classic surround” theaters. Top Speaker Arrays have specific angles and aimings. Courtesy of Dolby.

What would that mean for a possible Dolby Atmos-capable system at home? As in Home Cinema, in classic surround all speaker arrays are represented by one single speaker (L, C, R, Ls, Rs, Lsr, Rsr and LFE). In addition one would need the top speaker arrays Lts and Rts, so at least two more speakers are required. But if an Object has to move into the room with a given size, one could think at least two speakers for each top array are needed, for a total of 11.1. Ideally in-ceiling speakers, but perhaps more practically (or less unpractically) the 11.2-setup introduced back in 2008 in Yamaha’s RX-Z11 does the job, as it would be able to pan the objects in the room on all three axes, and, with the front/rear subwoofer setup, would be able to represent the main LFE and surround-subwoofer. A Dolby Atmos bitstream will be rendered by the AV-receiver into the number of speakers detected, but if the AV-receiver is not Atmos-capable, Atmos could be decoded by the source (an Atmos-capable player) into LPCM of a given number of channels, presumable up to 32 channels as the latest HDMI 2.0 version specifies.

Typical 11.2 system Courtesy of Yamaha

As Dolby Atmos can be “re-rendered” in the post-production studio to classic 7.1 (or even 5.1) for monitoring purposes, it may not be necessary to create a separate mix for classic surround (e.g. Dolby TrueHD). This re-rendering in studio should be similar to the rendering to 7.1 of an Atmos stream by an Atmos-capable AV-receiver as an end result. Would an Atmos-stream rendered to 7.1 sound better than a native 7.1 TrueHD-stream?

To get the real benefits of Dolby Atmos, more speakers are required in the house. The real paradox in the audio industry is, the progress on the technical side is “more speakers”, but on the commercial side it’s “less speakers” (2.1 systems, sound projectors etc.). But who knows, if Dolby Atmos really will be introduced in consumer audio, the trend (at least on the high end side) could be go up again. The future will be exciting as always!

Many thanks to Tony Verkuijl of Yamaha Music Europe GmbH - Branch Italy for contributing this article.