Modular audio DSP solutions accelerate the development of high-end audio systems

Aguilera

Modular audio DSP solutions accelerate the development of high-end audio systems [Copy link]

In today's digital world, almost everyone can take out a pair of portable headphones with good sound effects. If you really love music, you will probably sing a few karaoke songs on your high-fidelity floor-standing speakers at home and listen to audiophile-level music. Imagine traveling 140 years through a time machine, listening to passive phonographs to the latest 16-channel audio and video receivers (AVRs). What an exciting and shocking scene that would be! In the 19th century, when phonographs were played in villages and towns, neighbors gathered to listen and enjoy music, and today 16-channel AVRs have entered the living rooms of ordinary people. The dynamic range and fidelity have undergone major changes, with an increase in the number of channels, a decrease in noise levels, higher resolution and higher precision, making the enjoyment of music even better. Among them, the digital signal processor (DSP) has made an indispensable contribution.

Floating-point digital signal processors usher in the era of high-end audio

Analog Devices introduced integrated digital signal processors in the mid-1980s, which were 16-bit fixed-point processors. These processors used Harvard architecture and were very efficient. The first audio products using these processors were players with two-channel decoding and post-processing capabilities. The two-channel decoders running on these processors did use double-precision math and output 24-bit audio.

Powered by a dual-core SHARC+ digital signal processor, Damson Global's compact wireless S-Series features the Dolby Atmos 3D audio codec.

However, engineers used to spend a lot of time tuning these fixed-point processors and getting the desired characteristics from the filters. For major issues like decimation and truncation errors, the only solution was to painstakingly tune the filter coefficients through trial and error. Later, some simulation software packages did generate coefficients for fixed-point processors, but they did not completely eliminate the manual tuning process.

Floating-point digital signal processors were a boon, bringing multiple advantages, including better dynamic range, higher resolution, and lower noise. Soon, the professional audio industry realized these advantages and used them in high-end studio equipment, with multiple processors on each board. Then, the audio decoders of cinema equipment ran on these DSPs. As expected, they were also used in AV receivers for decoding and post-processing, bringing the theater experience to people's living rooms. Among them, ADI's SHARC processor series has dominated the market in floating-point DSPs, with its excellent core and memory performance and excellent I/O throughput.

Modular audio solution simplifies high-end audio system development

ADI's 32-bit floating-point SHARC digital signal processor is based on the Super Harvard architecture, which is characterized by a balance between core and memory performance and I/O throughput. The architecture adds an I/O processor and its associated dedicated bus, further extending the original concept of independent program and data memory buses. In addition to meeting the needs of computationally intensive real-time signal processing applications, the SHARC processor also integrates large memory arrays and application-specific peripherals, effectively simplifying product development and shortening time to market.

Just recently, ADI announced that its SHARC audio module (ADZS-SC589-MINI) is now available. This hardware/software platform helps improve the prototyping, development and production efficiency of various digital audio products. The SHARC audio module realizes an innovative combination of high-performance audio signal processing devices and a comprehensive software development environment. It is very suitable for sound processors, multi-channel audio systems, MIDI synthesizers, and many other DSP-based audio project applications.

ADI launches an audio module platform designed to accelerate the development of audio DSP projects.

Traditional product development usually requires a lot of time and effort in device selection, circuit board prototype development, and basic software structure development before considering how to meet your product needs. SHARC audio modules can greatly accelerate this process by providing an audio platform that integrates a variety of analog and digital I/O options as well as basic software packages and development environments. SHARC audio modules can be used as stand-alone self-sufficient systems or expanded to support the creation of custom I/O and control interfaces.

This audio module makes SHARC an ideal platform for supporting the development of audio processing equipment. Redesigning the entire system originally required a lot of time, but now with this module, it only takes a short time to create a variety of powerful custom I/O peripherals. The core product of the SHARC audio module is the ADSP-SC589 processor from Analog Devices, an advanced engine for audio processing that uses two 500MHz SHARC+ DSP cores and an ARM Cortex-A5 core.

Also connected to the processor are the following carefully designed system peripherals: two 2 Gb DDR3 memories and one 512Mb SPI flash; UART (for MIDI, etc.); SigmaDSP ADAU1761 96 kHz, 24-bit audio codec; Gigabit Ethernet, S/PDIF, and 1/8" stereo jacks; USB OTG and HS; AD2425W A2B multichannel audio interface; and more.

In addition to the main SHARC Audio Module, Analog Devices also offers an expansion board called "Fin" that provides additional functionality. The Audio Project Fin (ADZS-AUDIOPROJECT) is a MIDI and instrument/FX daughter card that provides a control surface for the SHARC Audio Module with " stereo I/O, additional connections, and expanded signal acquisition capabilities.

The SHARC Audio Module Platform accelerates audio system development.

With the SHARC Audio Module, developers can focus on their own algorithm and user interface development work by leveraging a compact and cost-effective base platform and a highly optimized software and tool ecosystem.

The excellent toolchains for these processors provide assistance in writing C/C++ code and also use some highly optimized libraries to implement FIR, IIR, FFT/IFFT, etc. Programming in C reduces time to market and is portable across processors without requiring in-depth knowledge of the processor architecture and underlying features. For example, the IP holder can release multiple versions of a decoder to correct bugs or make improvements and provide new C/C++ code with some changes. An efficient processor compiler can create new libraries for the processor, requiring less effort and time to accomplish the same task than using assembly language.