Back in March, my colleague Carl Wakeland wrote an excellent blog describing all the reasons why audio is so important in VR. It’s a long read but well worth the time. As he puts it so articulately, what we hear is vital to our perception of reality. And achieving lifelike presence in virtual reality demands audio that sounds as real as the graphics need to look. Suspension of disbelief.
Audio that Feels Real
Creating audio that feels real when you’re in VR – when you can turn your head in any direction with continuous visuals - requires both physical acoustics-based sound modeling, and concurrent real-time audio physics calculations and processing. Conventional pre-baked approximations can offer acceptable experiences, but will fail to create the true presence necessary for full immersion.
The way to achieve immersive VR audio is by enabling a significant number of audio sources and processing them with real-time dynamic physics at low latency. This process allows humanly-unnoticeable delays between an input being processed and the corresponding output. However, it requires achieving a level of real-time performance in gaming scenarios that is not possible on the CPU alone. Because of performance limitations, physics-based audio engines have been forced to rely on statistically-derived physics – until now.
Physics-Based Audio Acoustics Rendering
Enter AMD TrueAudio Next. As part of AMD’s LiquidVR™ technology initiative aimed at enabling a fully immersive and comfortable virtual reality experience, TrueAudio Next is a scalable AMD technology that enables real-time dynamic physics-based audio acoustics rendering. It uses Radeon Rays (formerly AMD FireRays) to enable the entire soundscape to be modeled physically, with more than 32 stereo 2-second convolution sources.
We are thrilled today to announce that the TrueAudio Next open-source library is now available on GPUOpen.
TrueAudio Next stands out from the crowd thanks to the Asynchronous Compute Engines enabled by Radeon™ GCN and Polaris graphics architecture. It is a conventionally-held belief that using a GPU to render audio can cause too much latency, while also interfering with graphics performance. However, TrueAudio Next has the ability to leverage the powerful resources of GPU Compute, safely allowing the GPU to accelerate audio rendering. This is mainly thanks to a core element of this technology: Compute Unit (CU) Reservation.
AMD’s CU Reservation feature allows a limited set of CUs* to be partitioned off and reserved for as long as is required, offering a flexible and predictable reliable audio acceleration solution – isolating audio usage from graphics usage.[Note: AMD delivers the CU Reservation feature to AMD approved partners via the driver. The TrueAudio Next library can be used with or without CU Reservation.]
We are excited about the potential of TrueAudio Next, as it truly has the capacity to deliver spatially- and positionally-accurate audio to millions of consumers. It enables developers to integrate realistic audio into their VR content in order to achieve their artistic vision, without compromise. Combining this with AMD’s commitment to work with the development community to create rich, immersive content, the next wave of VR content can deliver truly immersive audio – that will sound and feel real.
Sasa Marinkovic is Head of VR and Software Marketing for AMD. Links to third party sites are provided for convenience and unless explicitly stated, AMD is not responsible for the contents of such linked sites and no endorsement is implied.
*Discrete AMD Radeon™ and FirePro™ GPUs based on the Graphics Core Next architecture consist of multiple discrete execution engines known as a Compute Unit (“CU”). Each CU contains 64 shaders (“Stream Processors”) working together.