Is Dolby Atmos a Flop For Home Theater Like 3DTV Was?

An anonymous reader writes: Object-based audio is supposed to be the future of surround sound. The ability to pan sound around the room in 3D space as opposed to fixed channel assignments of yesterday's decoders. While this makes a lot of sense at the cinema, it's less likely consumers rush to mount speakers on their ceilings or put little speaker modules on top of their existing ones to bounce sound around the room. Leading experts think this will be just a fad like 3DTV was. What do you think?

Comments

Ambisonics

[wiredlogic]

A more rational pannable surround could be implemented with just four speakers using Ambisonics. It isn't patentable and doesn't sell lots of speakers so it will continue to be ignored.

Re:Ambisonics

[iluvcapra]

The problem with Ambisonics is it tends to favor a strong Sweet Spot, which is OK in a home theater but will fail in a large room, where people are seated to the four corners of the space. Speakers near the walls will always tend to be perceived as louder, and the further you are from the tuned center of the room, the more the sound field will appear to be warped toward the closest wall. This happens with 5.1 but the effect is mitigated by the fact that there's a center speaker behind the screen, and the mixers have individual control over speaker levels and panner divergence.

Ambisonic mixes are almost by definition not mono-compatible and don't allow the mixers to address sounds to individual speakers with unlimited panner divergence. There's always some situation where you want a sound to come from every speaker in the room, or to come from speakers on the opposite sides of the room, with equal intensity: the latter is impossible with B-format (and only possible in the limit with n channels), and the former is impossible with any theoretical pure ambisonic sound system.

Here is TFA

[Matt_H]

The missing link is http://www.audioholics.com/audio-technologies/5-reasons-dolby-atmos-is-doa

It's all a lot of fun

[sjames]

until the Sontarans invade.

3DTV a fad?

[jd2112]

That would imply that it was popular at some point.

Pssssh

[gman003]

It's not even a fad - it's dead on arrival. Most people don't even use 5.1 speakers. Hell, most don't even use 2.1. Anything that requires that much dedication of the room to audio is not going to sell to the mass market. Period.

3D TV at least had a vague hope of succeeding in the mass market. If they can ditch the glasses, they might actually succeed. But people are lazy and don't want to put any effort into their mindless entertainment. Putting glasses on to watch a movie was too much for them. Do you really think setting up a shitload of speakers all around the room is going to pass?

Re:im a music mixer in hollywood...

[iluvcapra]

and yes, you dont have enough speakers and amps for atmos at home. sound bars wont make it. hell, most people i know have their 5.1 systems setup wrong.

I'm a sound designer in Hollywood, my credits include Men in Black 3 and Zero Dark Thirty.

The main promise of ATMOS was that it wouldn't matter how many speakers you had -- a mixer could prepare a final mix in Atmos in his 60-horn room, but then when the bitstream on the DCP or Blu-Ray was decoded in the theater or home, it wouldn't matter if the end-user had a 60-speaker Atmos rig, a 9.1, a Barco Auro speaker system, a 5.1, a stereo or even a mono. The Dolby renderering algos would simply take the panned objects and automatically render the correct audio stream for each speaker, as a function of the speaker's position relative to the listener. The Dolby RMU is just a glorified OpenAL audio engine, it gets fed audio streams that have an alt/azimuth data envelope, and this envelope is transformed down to whatever speaker array the end user has.

What's even more interesting is you could have a significantly more complex speaker array than the person who mixed it -- maybe he mixed it with 32 speakers, and you have some future-ready system with 100 -- and the renderer will still do the Right Thing and expand the spatial resolution accordingly. Atmos mixes are future-proof for any simple, non-phase-related speaker array.

Re:im a music mixer in hollywood...

[SuricouRaven]

What could be done by reversing this process from the stream? There's a nifty trick on surround sound - some sounds you'll find the center channel is used for lyrics and the sides for instrumentals in songs, making it trivial to isolate them and get clean audio for redubbing with. If it works as you describe, would that make it easy to pull out individual instruments or effects? That could be useful for hobbyist remixers.

Re:im a music mixer in hollywood...

[iluvcapra]

The way Atmos works is it can carry up to 128 individual audio channels. 20 of these are set aside for two discrete 9.1 mixes (mixers choice what goes in those), the remaining 108 are set aside for individual pannable objects. In the file themselves, these audio objects are full-rez and lossless; however, these objects don't "live" all the time, the mixer can use them for a few seconds here and there. Nothing as general as "all the dialogue" or "all the car sound effects" lives in the pannable objects throughout the entire project.

There are discrete sounds in the Atmos bitstream itself though, and in principle it would make remixing easier, so I suspect you'll never see an Atmos bitstream in a consumer format without DRM.