Page 17 - EngineerIT September 2021

AUDIO PROCESSING

Scalability processing needs of the larger system, instead offloading
While latency concerns are problematic for SoCs in applications that processing to the low latency audio DSP. Moreover,
such as noise control, another key shortcoming for SoCs aspiring audio DSPs that offer several different price/ performance/
to perform acoustic processing is in scalability. In other words, memory levels across a comprehensive code-compatible and
SoCs that control large systems (such as automotive head-end pin-compatible roadmap, offer maximum flexibility for system
units and clusters) with many disparate sub-systems cannot designers to right-size the audio performance offering for a
easily scale from low end to high end audio needs because given product tier.
there is constant conflict between the scalability needs of
each sub-system component, requiring trade-offs in the overall Upgradeability
SoC utilisation. For instance, if a head-end SoC connects to a As over-the-air firmware updates become more common in
remote tuner and, across automotive models, that tuner needs today’s vehicles, upgradeability to issue critical patches or
to scale from a few channels to many channels, each channel provide new functionality becomes increasingly important.
configuration will amplify the real-time concerns mentioned This can cause major issues for an SoC because of the
earlier. This is due to each additional feature under the SoC’s increased dependencies among its various subsystems. First,
control changing the real-time behaviour of the SoC and the on SoCs, multiple processing and data movement threads are
resource availability of key architectural components used by vying for resources. This increases competition for processor
multiple functions. These resources include aspects such as MIPS and memory when new features are added, especially
memory bandwidth, processor core cycles and system bus fabric during bursts of peak activity. From the audio perspective,
arbitration slots. feature additions in other SoC control domains can have an
Apart from the concern about other sub-systems connecting unpredictable effect on real-time acoustic performance. One
into the multi-tasking SoC, the acoustic subsystem itself has its side effect of this situation is that new functionality must be
own scalability issues. There’s a low end to high end scaling (for cross-tested across all operating planes, resulting in myriad
example, increasing the numbers of microphone and speaker permutations between various operating modes of the
channels in an ANC application) and there’s also the audio competing sub-systems. Thus, software verification increases
experience scaling, from basic audio decode and stereo playback exponentially for each upgrade package.
up through 3D virtualisation and other premium features. Though Viewed from a different angle, it could be said that
these requirements do not share the real-time constraints of ANC improvements to SoC audio performance are dependent on
systems, they nonetheless relate directly to the choice of audio available SoC MIPS, in addition to the feature roadmaps for the
processor for a system. other sub-systems controlled by the SoC.
Utilising a separate audio DSP as a co-processor to an
SoC is a perfect solution to the audio scalability problem, Algorithm development and performance
enabling modular system design and a cost-optimised solution. It should be apparent that, when it comes to developing real-
The SoC can focus much less on the real-time acoustic time acoustic algorithms, audio DSPs are purpose-built for the
task. As a significant differentiator to SoCs, stand-alone audio
DSPs can offer graphical development environments that allow
engineers with minimal DSP coding experience to add quality
acoustic processing into their designs. This type of tool can
lower development costs by reducing development time without
sacrificing quality or performance.
As an example, ADI’s SigmaStudio graphical audio
®
development environment offers a wide variety of signal
processing algorithms integrated into an intuitive graphical user
interface (GUI), allowing the creation of complicated audio signal
2
flows. It also supports graphical A B configuration for audio
transport, greatly helping to catalyse real-time acoustic system
development.

Audio-friendly hardware features
In addition to a processor core architecture that’s specifically
designed for efficient parallel floating-point computations
and data accesses, audio DSPs often have dedicated multi-
channel accelerators for common audio primitives such as
fast Fourier transforms (FFTs), finite and infinite impulse
response (FIR and IIR) filtering, and asynchronous sample
rate conversion (ASRC). These allow real-time audio filtering,
sampling and frequency domain conversion outside of the core
CPU, increasing the effective core performance. Additionally,
they can facilitate a flexible and user-friendly programming
Figure 2: ADSP-2156x DSP, illustrative of a highly scalable audio model due to their optimised architecture and data flow
processor management capabilities.

EngineerIT | September 2021 | 15