Dolby Atmos

[Dave2002]

Does anyone know what Dolby Atmos is? It appears to be a surround sound system with many speakers, including overhead, yet the same source feed can be used for many different speaker cofigurations.

[richardfinegold]

Does anyone know what Dolby Atmos is? It appears to be a surround sound system with many speakers, including overhead, yet the same source feed can be used for many different speaker cofigurations.

Yes, I do. The goal is ceiling speakers to provide a more immersive experience . If you don't presently care for the concept of Surround Sound, you won't be interested

[Dave2002]

Yes, I do. The goal is ceiling speakers to provide a more immersive experience . If you don't presently care for the concept of Surround Sound, you won't be interested

I wonder why the system doesn't have floor speakers as well!

The bit which interested me was the notion that the same source material could/would drive as many, or as few, speakers as actually deployed in the system. Michael Gerzon believed that this could be done with as few as four channels, and to a certain extent he was right. Where Dolby systems I think differ is that they don't simply treat the system and multi-channel recording as behaving like a linear system, Perhaps the simplest example is of a three channel front positioned speaker arrangement. In some linear systems (based on stereo, perhaps) sound from the centre would also have feeds to the left and right speakers, though at a lower level. This might be acceptable for some material, but for a true point source centrally positioned the simplest is to have only the centre channel driven, with no cross talk to the other channels. However this approach requires each original source sound to be located exactly at the horizontal and vertical angles of the reproducing speakers, which is often not the case, and a "hole in the middle" type of problem will emerge for any sound source which is not located in a position correspondig to the reproducing speaker locations.

The kind of surround sound in films is often highly artificial, and I have found the sound tracks on some DVDs absolutely awful. Maybe the same effects do work in a cinema, but in a domestic environment special effects seem often to be completely overwhelming (in a bad way - too many explosions etc) and speech is often muddled. I recall watching, or trying to watch the DVD of War of the Worlds a few years ago. It wasn't a great film, but the sound track was dreadful, and I think I may even have reverted to stereo and the TV speakers to make it watchable.

I think it is possible to get good effects for music with some forms of surround sound, but the recording approaches have to be different, and more subtle. Perhap Atmos uses some form of steering channel, for essentially 3-D pan potted spot sources, but there'd also need to be some way of capturing the overall ambience, which spot mics and pan potting don't really do well IMO. For films, which are often highly artificial anyway, this kind of approach might work well enough.

[MrGongGong]

Ambisonics works as long as you are in the right place and don't move your head

[Richard Barrett]

Gerzon was of course a genius, but that's right about Ambisonics. My favoured way of doing surround sound is DBAP (distance-based amplitude panning) which can be done with 3 or more speakers plus - crucially - information about exactly where they are relative to one another (perhaps Dolby Atmos is based on similar principles, I haven't been able on a quick search to find this out), though 16 speakers in a 3D array has for me been an optimal compromise between complexity/expense and sharpness of the sonic image.

[Dave2002]

Gerzon was of course a genius, but that's right about Ambisonics. My favoured way of doing surround sound is DBAP (distance-based amplitude panning) which can be done with 3 or more speakers plus - crucially - information about exactly where they are relative to one another (perhaps Dolby Atmos is based on similar principles, I haven't been able on a quick search to find this out), though 16 speakers in a 3D array has for me been an optimal compromise between complexity/expense and sharpness of the sonic image.

That's very interesting. Do you not use phase differences at all then? Do you have details of your "minimal" 16 speaker recording/replay systems?

Gerzon was certainly unusual.

[Richard Barrett]

That's very interesting. Do you not use phase differences at all then? Do you have details of your "minimal" 16 speaker recording/replay systems?

There will be phase differences naturally (if I understand what you're asking) since DBAP allocates volume levels to different speakers according to the distance of the virtual source from each speaker (hence the 3-speaker minimum). If you want more detail about the system I'm talking about you could take a look at the PhD thesis by my friend Michael Hewes, who was the principal sound projectionist at the performance of my big work CONSTRUCTION at Huddersfield Town Hall in 2011.

https://researchbank.rmit.edu.au/eserv/rmit:160490/Hewes.pdf

The section dealing with this project begins on page 81. Previous sections deal with the gradual evolution of the system through the course of a series of smaller-scale "tryouts".