Mark Cerny talks about Dolby Atmos support for the PlayStation 5

On the heels of our report on the PlayStation 5’s new Dolby Atmos support in DF Weekly Live Past WeekSony’s marketing team was contacted by a response from the PS5’s lead system architect, Mark Cerny, about how the new surround sound setup actually works. In Direct, we speculated that the PS5’s already impressive Tempest 3D audio data was likely recalculated and injected into the Dolby Atmos container.

If you think about it, this is an ingenious solution but it’s also the only way the existing library of PS5 titles can actually work with Atmos settings so far, developers won’t master audio to support the 7.1.4 speaker system.

In his statement, Mark Cerny takes us through the entire PlayStation 5 audio pipeline and how support has been extended to accommodate the additional amplifiers offered by the Atmos setup. The nature of Sony’s solution means that Atmos support is completely lag-free, unlike some of the issues we’ve had with Xbox and Windows devices. Is this solution actually a combination of Atmos? It’s ‘as good as it gets’ from a mathematical perspective, but with full support now added, audio engineers can master their audio mixes on full gear from the Atmos spec, which means there’s room for further improvement.

Here is the full breakdown:

It’s probably easier to talk about Tempest-based 3D audio and Dolby hardware support in terms of Ambisonic audio, which is getting more and more popular these days (note that there are other strategies for 3D audio, including those using discrete 3D audio objects, but the situation is somewhat similar).

Ambisonic sound can be considered a very radical extension of stereo sound. With stereo audio, the game’s audio engine (or the middleware used) will add an audio source to one or both channels based on its location – if the source is to the right of the listener, it will essentially be added to the correct channel, and so on. With Ambisonic sound, there are a lot of channels—the fifth arrangement is very common and uses 36 channels, so it makes for a good place for the sound. Then the audio source is added to those 36 channels based on the location; The math is a little more complicated than when using stereo but not overwhelmingly. Since the audio processing is channel based (albeit on 36 channels instead of 2) the sound designer keeps very good control over the mixes, filters, etc., and strategies such as dynamic range compression (where a particular important sound such as the player character can be heard) is guaranteed. ) can be used as normal.

The Ambisonic audio channels are then delivered to the Tempest 3D AudioTech engine for rendering, i.e. the Tempest engine uses the player’s HRTF and speaker locations to create an appropriate audio stream for each speaker. Ambisonic audio channels encode all directions, including above the player; Even if it’s the width for headphones, this is very important, because it allows the player’s “louder” sound to be processed in a way that sounds like it’s actually coming from above – and that, of course, is where HRTF coding for the head-and-ear shape comes in.

Until the latest update, the Tempest engine would display information in Ambisonic channels to headphones, stereo TV speakers, and 5.1 and 7.1 audio settings. Now 7.1.4 has been introduced, with its four speakers, but really nothing changes in Tempest’s overall rendering strategy—Ambisonic’s 36 channels include sound coming from all directions, including the top of the player. To put that differently, the four speaker support is “top notch” support, handled just like any other speakerphone. Also note that the latency of these new speaker settings is identical to what it was in the past for stereo, 5.1 and 7.1.

As a result, the 7.1.4 experience for current games should be pretty good. It’s true that the game teams couldn’t test with these speaker setups but the support should be somewhat automatic, the necessary audio data for the game is already in the form of Ambisonic. Going forward, there is an opportunity for improvement as sound designers can check for the highest sound quality in the 7.1.4 speaker settings as well.

– Mark Cerny, PS5 Lead Systems Engineer

This statement made its way into the upcoming DF Direct Weekly #124, going public tomorrow, where John Linneman and I discuss Mark’s comments and also our “from the audience” test on surround sound latency. That’s in addition to a host of other news topics that have combined to present what may be the longest live-action we’ve ever done, so look out for that tomorrow at 16:00 BST.