Introducing dsprak™ High Performance Audio

dsprak inside the box

Danville Signal’s dsprak™ high performance audio platform is a semi-custom solution for OEMs. Optimized modules are combined with a custom motherboard to create a no-compromise platform that meets specific customer requirements.

• Optimized High Performance Audio Module
• Expandable to 32 Channels
• DSP Concept’s Audio Weaver Support
• dspblok 21369zx (ADSP-21369) DSP Module
• Two Channel ADC Module (AKM AK5394A)
• 8 Channel DAC Module with Analog Attenuators (AKM AK4396 & Cirrus Logic CS3318)
• 8 Channel DIT (AES 3id/SPDIF)
• 8 Channel DIR with ASRC (AES 3id/SPDIF)
• Low Phase Noise Clock with Word Clock Sync

dsprak from the right

The basic 1U system supports 2 to 8 analog and digital inputs and 8 analog and digital outputs. Additional motherboards can be stacked to create larger systems with up to 32 channels.

Danville engineers are available to discuss your instrument requirements. Or you can start the process by completing our online Request for Quotation form.

Music delivery systems are rapidly changing. The days of a CD changer and a huge library of CDs is quickly being replaced by a computer with the content residing on a large harddrive.

Unfortunately, the internal soundcards in most computers are not nearly good enough for the demands of high quality audio.

USB Audio Class 2 has become the interface of choice for extracting the digital content from the music server computer to a target DAC or headphone amplifier. This is a relatively recent development. Although USB audio has been around since the early days of USB, it was basically a 48k, 16 bit stereo interface (USB Audio Class 1.1). USB Audio Class 2 can support multiple channels at high sample rate and full bit depth. It uses an external clock from the USB client that helps insure minimal jitter at the target device.

The leader in USB audio class 2 chip solutions is XMOS. Danville Signal is a third party partner with XMOS. We have utilized and extended their technology in several ways, primarily by combining their chips with Analog Devices' SHARC DSPs. This allows the XMOS device to concentrate on communications while the DSP to manages signal processing and control.

An audio class driver is mostly concerned with streaming audio data at various sample rates and bit depth. It has limited control, such as volume up and down. Danville has extended the control capability of our XMOS targets by appending a HID interface into the XMOS code. This allows us to download code or instructions to the DSP or interface to a graphical design and tuning interface such as DSP Concepts' Audio Weaver. We supply a Window DLL so that custom programs can be easily created to take advantage of these features without the need to be an XMOS programmer.

There is native USB Audio Class 2 support for Apple and Linux. Microsoft has not opted to provide similar support with their operating systems. This may be because USB Audio Class 2 places demands on the computer that not all PCs can support. We have driver support for Microsoft Windows (XP, 7 & 8). We provide these drivers with our general purpose dspblok modules such as our dspblok 21479u8a and with our more dedicated targets such as our wireless audio transmitters.

If you are considering your own design, we suggest you consider our dspblok 21479u8a. This module combines an XMOS XS1-U8A processor with an Analog Devices ADSP-21479 DSP. We use this module in many of our turnkey OEM audio products. With the dspblok, you don't need to program the XMOS device at all (XMOS programming is tricky). You also get Windows driver support. If you are interested in a more custom solution, we can provide a complete hardware and software package with optional graphical design and tuning capability. Check out the dspblok 21479u8a page for more information.

 

dspblok 21479u8a Xmos DSP Audio Class 2 dspblok 21479u8a
XMOS-SHARC Audio Class 2 Solution

 



  • Up to 384 kHz sample rate
  • S/N in excess of 120 dBA
  • Balanced/Unbalanced I/O
  • USB Audio Class 2
  • Analog Devices SHARC DSP
  • Low phase jitter clocks
  • AKM DACs
  • Optional analog input module
  • Optional custom configurations

The dspMusik™ 2/8 is a stereo in, eight out multichannel DAC platform suitable for high performance DSP loudspeaker crossovers, music decoders and many other high quality audio algorithm applications. An LCD/Encoder provides source selection, volume control and access to special features. You can also use an Apple Remote for control.

Semi-customizable, Professional Audio Converter Platform

The dspMusik 2/8 platform delivers audiophile performance, featuring state-of-the-art components from Analog Devices, XMOS and AKM. Each part has been carefully selected to build a precisely integrated system.

The dspMusik starts by using low phase jitter clocks driving all the data converters. The optional ADC incorporates very high common mode rejection / low noise balanced inputs and DACs with minimum phase anti-imaging filters (desirable to avoid pre-echo effects). Signals are processed by an audio-optimized 32-bit/40-bit floating point DSP capable of up to 384 kHz performance.
 

Audio Weaver Programming Software

The dspMusik 2/8 is supported by a custom version of DSP Concepts’ Audio Weaver™ programming software. This is a powerful graphical design package that lets you create your signal processing chain by simply dragging and dropping highly optimized audio processing modules. You can use any of the dozens of pre-built audio functions—EQs, crossovers, mixers etc—or enhance and combine them with your own custom processing to create precisely the functions you need. Once the design is done, the file can be directly saved and run stand-alone on the dspMusik platform.

  • PC-Based Development Environment
  • Drag and Drop Signal Processing
  • Over 400 processing Modules
  • Real Time Tuning and Profiling
  • Visit www.dspconcepts.com for more info

Audio Weaver Software Screen Shot

 

OEM/ODM Customers

The dspMusik 2/8 platform is available in a variety of configurations including PCB assemblies where you provide your own housing, or it can be provided in an extruded aluminum enclosure. Panels can be rebranded with your company’s logos and trademarks to fit your company’s specific requirements and branding. Danville also provides customized display software that includes your company and product name as well as access to special features that may be part of your system design. Contact us to discuss your project needs.

 

dspMusik 2/8 features:

DSP:
  Analog Devices SHARC ADSP-21469 or  
  Analog Devices SHARC ADSP-21479

Inputs:

   USB Audio Class 2: XMOS 
   SPDIF Digital Input
   Optional Analog Input Module:
       ADC: AKM AK5397 (S/N > 120 dBA)

Outputs:
   DACs: AKM AK4495 (S/N > 120 dBA)

Configurable Input / Output:
   Balanced or unbalanced circuits
   Consumer (-10dBV) or Pro (+4dBu) levels
   Standard 6.3mm TRS connectors
   Optional Tascam style DB-25 connections

Operating System Support:
   Microsoft Windows XP, 7, 8 &10 (included)
   Apple OS X (native support)

Input Power (Complete Version)
   IEC Mains Connector
   Autoswitching 115/230VAC, 50/60Hz

dspMusik 2/8 Options

   dspMusik 2/8 Complete

Includes ADC, UAC2, S/PDIF & Display/Encoder powered by dspblok 21479u8a

P/N A.03708E-479-AD-DE

   OEM/ODM Options:

There are many OEM/ODM options including individual board assemblies, power supply adapters for Pascal & Hypex power amplifiers, 16 channel expansion cappability and alternative dspblok modules with OEM/ODM quantity pricing.

Contact Danville for available options.

 

 

dspMusik 2/8 Multichannel DAC

 

dspMusik 2/8 Interior View  

Download the dspMusik 2/8 Multichannel DAC Platform Datasheet

 

Download the dspMusik 2/8 User Manual  

 

 

The benefits of active loudspeaker crossovers has been understood for many years. In an active loudspeaker system, each individual speaker can be driven by its own power amplifier. A band limited signal is fed into each amplifier that can have its phase and frequency response tailored individually. There is a big advantage to driving each speaker directly from a power amplifier since the current draw of a speaker can be quite non linear. If the amplifier has essentially a zero ohm output impedance, these affects a minimized. It is easy to compensate for sensitivity differences of the drivers and clipping can be confined to the output of single driver.

Passive loudspeaker components tend to be large and not very ideal. Capacitors are not really just capacitors and inductors are not really just inductors.

Historically, active crossovers were usually constructed by analog filter circuits placed in front of a set of power amplifiers in a separate enclosure. The filtering would usually be simple butterworth, bessel or linkwitz-riley implementations with limited adjustability. The loudspeaker would have separate inputs available for the tweeters, midrange and woofers. Of course, this meant lots of speaker cables and a stack of amplifiers.

DSP based crossovers can extend the active crossover idea in several important ways. First, the filters can be substantially more optimized. The same types of filtering that analog implementations relied upon are still available via IIR based DSP filtering. These filter options can be complemented by other topologies such as finite-impulse response (FIR) filters that are not available as analog filters. In a modern DSP crossover, the complexity of the filter is somewhat arbitrary, the computation power is generally available to do as much as you want. With a DSP crossover you can adjust delay, add individual look ahead limiters (to avoid warranty repairs) and even individually match the response of each production loudspeaker.

Another technology change that has made DSP crossovers much more practical is that high quality switching amplifiers are now available. Most studio monitors today, use internal amplifiers and certainly any wireless audio system would probably be self contained as a practical matter.

Perhaps, the easiest way to create a DSP crossover is to use a single chip solution such as one of Analog Devices' SigmaDSP processors. Typically a system is created using a graphical design package that simplifies the development and tuning process of creating a workable solution. This is a great way to create low cost, high volume DSP based solutions for mass market applications. With this approach you can get 15-16 bit CD quality.

Danville's dspCrossover™ technology goes much further. We generally start with a SHARC DSP which is much more powerful processor than any of the integrated parts. More importantly, we use very high performance DACs, low phase jitter clocks and low noise analog support circuits. Inputs can be digital or analog, wireless or wired. We have a power graphical design and tuning interface (DSP Concepts' Audio Weaver) that allows real time tuning with optimized code. When you are satisfied with the result, you just write to flash. The result is a fast time-to-market solution with very high performance.

Our target markets are high end home stereo, home theatre and professional audio studio monitors. Although not strictly a crossover application, we can also use this techology for high quality headphone DACs.

Almost all of our dspCrossovers are custom designed to meet the individual needs of the target speaker and customer. Working with Danville, complements the skills of your acousticians and speaker designers with a skilled audio oriented DSP technology company.

 

 

The wireless revolution has been in full force for several years. Many of us use wireless connections for computer networking, hands-free cell phone communications and many other day to day activities. We all get the benefits of wireless connectivity.

So why has high quality wireless audio been missing?

The short answer is bandwidth. High quality audio takes lots of bits delivered in a timely consistant manner. Unlike WiFi, you may not have the opportunity to retransmit the data again if the data fails the first time.  The second, less obvious answer is economics. Let's explore each of these in turn.

Most wireless audio solutions aim very low. The thought is that convenience is everything and that 16 bit 48k stereo is more than enough. Anyone who has listened to MP3s understands this point of view. Suffice to say, this is not an area that we are interested in pursuing. 

If you set the bar much higher, then it is reasonable to require higher sample rates and certainly much wider bit depth. This limits the playing field to only a few candidates. If you also extend the requirements so that multichannel solutions are supported, then the field narrows even more. Imagine you would like to support either wireless surround sound for a home theatre application or a system with multiple subwoofers for a stereo application or both?

Summit Semiconductor has a premium wireless audio platform that meets these basic requirements. It supports up to a 7.4 system, sampled at 96k and 24 bit uncompressed bit depth. The Summit solution uses 24 RF channels in the 5-5.8 GHz band. It is adaptive to minimize interference from competing wireless solutions. 

This is a great start and the basic solution is supported by an industry consortium called WiSA (Wireless Speaker & Audio) Association. 

So it would seem like Summit has done a great job of addressing the "bits" problem and they have. But there is still more to creating a great wireless audio system and this is where "economics" comes into play. If we can insure that bits are delivered error free, then any platform, wired or wireless should have equivalent results given the same sample rate and bit depth. This is true as long as the bits arrive at precisely the right time. The right data at the wrong time is the wrong data!. This is why jitter performance is so important. If the data stream is improperly sample rate converted, you can permanently encode jitter into the stream and this cannot be fixed. If the data is correct but not jitter free, it can be fixed by jitter attenution techniques.

So lets get back to economics. If you are a chip manufacturer, it follows that you need to sell your products to a sufficently large audience. This means that you will need to create a sufficient least common denominator of features, performance and cost that insures that your devices are both affordable and meet the primary needs of your target audience. Summit has done this quite well. But due to economic realities, there is more that can be done.

Danville wireless audio technology uses SHARC DSPs and very low phase jitter clocks to complete the solution. There is a SHARC DSP in our transmitter implementation and SHARC DSPs in each receiver circuit. The SHARC also implements our dspCrossover functions. We complete the process with high quality DACs and related single conditioning circuits. 

At the 2013 CES show, we had a well known audio reviewer state that the wireless speaker we were showing was a wireless system that didn't sound wireless.  Isn't that the way its supposed to work?

Our wireless audio implementations are available to loudspeaker manufacturers. Let's start a conversation.

WiSA Member - Danville Signal

Danville Signal is a WiSA Member

Learn more about the WiSA Association  here. 

Audio Weaver from DSP Concepts is an innovative design environment for developing optimized embedded audio software. It embodies years of audio product development experience and enables algorithm and product developers to more quickly and efficiently develop products and technology. Audio Weaver is suitable for the entire lifecycle of an algorithm, from basic research, to optimization and productization, integration into end products, and ongoing legacy support. Audio Weaver is supported on many of Danville Signal's dspCrossover and dspMusik platforms.

Audio Weaver differs from other algorithm design tools in that is it focused specifically on audio and geared toward the needs of product developers rather than researchers. Audio Weaver automates many of the mundane tasks associated with developing embedded products and allows the designer to focus on adding sophisticated functionality or reducing time-to-market.

Whether you are a seasoned audio developer or in need of audio expertise, Audio Weaver will accelerate your development effort. 

 

dsprak Audio Platform

The world of high performance audio is changing. Sure, you can still find vinyl enthusiasts and advocates for 10 watt triode amplifiers, but you will also find many more listeners that are taking advantage of newer digital technologies. What they all have in common, is that they all want great sound.

Danville Signal is uniquely positioned in that we are a partnering company that has extensive experience in both DSP (> 20 years) and high performance audio solutions (> 35 years). We are active members of Analog Devices and XMOS third party partner programs.

Here are a few of the application areas that we address:

  • DSP-based Active Crossovers
  • High Performance Wireless Audio
  • USB Audio Class 2 DACs and Headphone Amplifiers
  • Bass Management
  • Room Correction and Beam Forming
  • Vinyl to Digital Converters (Translation Phono Preamplifiers)
  • Algorithm Decoders and Demo Systems

The bulk of our audio products are semi-custom solutions tailored to individual customer requirements (ODM & OEM). You probably won't see them on this web site, but they are in brands you know. We invite you to start a conversation.

We also have off-the-shelf products that perform well, but are not fancy. These may be useful as take off points or repackaged to fit your needs.

 

© Copyright 2003-2017 Danville Signal Processing, Inc. All rights reserved.

Danville Signal Processing, Inc.
29687 82nd  Avenue Way,  Cannon Falls, MN 55009
Phone: (507) 263-5854 Fax: (877) 230-5629