Some of us are old enough to remember going to record stores and looking through the racks of LPs. Occasionally we would even be able to afford to buy one of these precious vinyl discs.

The age of vinyl and Hi-Fi hardware

In the days of vinyl, audiophiles would spend their money on Hi-Fi equipment that could reproduce the music exactly – looking to match the sound that was first heard in the recording studio when the track was originally created. This first wave of Hi-Fi enthusiasts focused on the speakers – purchasing the same type of professional reference monitors that the studio engineers used.

Soon however people realized that there was a lot more information that could be extracted from the tiny indentations on their vinyl discs. Turntables, tone-arms and stylus became the focus of peoples attention and the age-old computer mantra of "Garbage in – Garbage out" was adopted by all who were interested in hearing the music with all its original detail.

The arrival of digital audio CDs

Then audio CDs came along. For some they marked a backward step compared to the leading vinyl systems. There appeared to be less detail. Musical ambience was lost and the words "less musical" were often used to describe the early CD players. Some experts put this down to the loss of dynamic range. The16bit data used for CDs provided a theoretical 96db of dynamic range, which should be enough, but perhaps other issues were at play.

Others said it was the sampling resolution. At 44kHz this should allow reproduction of the whole 20Hz to 20kHz, human auditory range. But experts pointed to the sharp high-frequency cut-off filters required to avoid anti-aliasing effects and highlighted the potential loss of harmonics and other subtle effects. CD players did improve over time. Noise isolation, better timing control and improved digital to analog conversion techniques all helped to recover some of the lost information and improved the musical experience. However Hi-Fi enthusiasts hung on to their vinyl records and continued to claim that their analog systems still had an edge.

MP3 - the dark ages

For audiophiles the dark ages came with the arrival of portable music players. Storing music on a flash device in a portable music player, and later on a mobile phone, made access to music so much more convenient. You could carry your whole music collection with you and download new music whenever you wanted.

But there was a problem with the music quality. With slow Internet connections and expensive flash memory storage, digital music files had to be compressed. The result was the MP3 file. All the effort and expertise of the sound engineer to capture the music, the ambience and the musicality, was squished out as the music was compressed down into small digital files. One compression technique was to look for zeros or data close to zero and remove this information, to be replaced later by synthesized zeros during playback. We still got the loud heavy beats but lots of the subtle information that hovers around zero in music, was lost. The initial musical attack that you hear when a guitar string is plucked, or the breath you hear when a brass instrument is played, were compressed away and replaced by a synthesized, averaged-set of digital numbers.

The arrival of music services

The concept of buying and carrying all your music on a portable device is starting to fade. Now people stream their music from online services, giving them access to, much more choice. They can take advantage of online big-data analysis tools that help them find new artists and new music that they would never have discovered otherwise. Some services will let you find all the music created by your favorite music producer. You can go even further and find music that has been mastered by a particular sound engineer or music that has been created at a particular concert hall or studio.

No one wants to lose the convenience of digital music. The record store was great but times have moved on and getting your music over the Internet is so much easier and more convenient. But do we have to put up with squished, compressed music files?

Hi-Res headphone sales boom

Not any more. Internet connection speeds are now fast, flash memory is cheap, and in the Cloud I can have as much storage as I need. We're starting to see a backlash against the compressed music files and a growing demand for Hi-Res digital music that delivers the original quality of the music and recording.

Sales of expensive headphones have been booming and people can now hear the inadequacies of the compressed music in their MP3 files and the poor quality of most smartphone audio outputs. Just as before when people first bought studio monitor quality speakers, people with studio quality in-ear monitors can now hear that there is information missing as they listen to their digital music. There is much more musical information available and they want to make sure that they can get every last detail. The old mantra of "Garbage-in – Garbage-out" has returned. We now want high quality musical sources to go with our expensive headphones.

We don't need to limit ourselves to lossless–CD quality either. We can have higher dynamic range and more accurate sampling that extracts much more detail from the music, getting us back to the musicality of a great vinyl system or the experience of a live performance.

Most digital music today is recorded at 24bits and the DSP algorithms in most leading pro-studio equipment will store intermediate DSP-math calculations at 64bits to ensure that no detail is lost (something our xCORE devices do too). Digital music is now sampled at 96kHz, 192kHz or even 384kHz. Up sampling can help to restore even more resolution. Working at these sampling rates and these bit resolutions reduces distortions but most importantly it allows more detail to be extracted at the low sound levels which are so important to the auditory experience. Hi-Res music sources are starting up fast. Download services and streaming services are now available in most countries – with Japan and China leading the charge.

Real Hi-Res digital music systems on-the-go

Wired or wireless - that is the next question? Well let's have both. I can use my smartphone to access the Internet to find music and discover many more new artists and tracks. I can use the best online services that make discovering music easy and fun. As new services and new online user interfaces are developed, I can switch to use these. I can do all this over a wireless connection. My smartphone is my remote control.

I can then download my music files and store them on my device or I can stream them over the Internet as part of a Hi-Res music streaming service. I can even stream the Hi-Res music file to my music player over a wireless connection.

However for playback let's make sure we store the file ,or buffer it, so that we don't lose any detail. Then connect over a wire to a specialized Hi-Res audio playback device, so that we don't lose any musical data. Let's make sure that the digital components in our device shield the analog music playback from the digital noise that we find in our multimedia server, PC, or on our smartphone. And make sure that it can support the latest digital Hi-Res music formats like DSD and new Hi-Res lossless formats like MQA.

In the past our audiophile LP turntables were great at extracting every last detail from a vinyl LP and our professional studio monitors let us experience the full musical performance. Now we must make sure that our digital devices extract all the information that is contained in our new Hi-Res music files. We need to ensure that the digital components don't add any noise or timing anomalies. We need to select the components carefully. We can't just focus on the DAC or the Op-Amps, we need to make sure that we have great digital audio components too, including the connectivity over USB and specialized Time-Sensitive connections for audio over Ethernet. Let's make sure we don't put ‘Garbage' back into our new Hi-Res digital music system.

Enjoy your music – the way it was supposed to be heard.

What do you think about the future of digital audio? Please use the Comment button below to add your thoughts.

Is the audio market finally getting to grips with Hi-Res Audio (HRA)?

Here are 5 key trends emerging in this competitve and fast paced market.

1. Entering the mainstream
2. Going mobile
3. Increasing Hi-Res content
4. Cross-industry collaboration & support
5. Consumer demand for sound fidelity

Click to Enlarge     

1. Enters the mainstream

With consumer electronics giant Sony entering the market at the end of 2013 the HRA movement gathers momentum. Sony launches a full product range from hi-res capable headphones, amplifiers and portable players to high-end audio systems and speakers, catering for both the most demanding audiophiles and the average consumer who simply wants their music to sound better on the go!

Other manufacturers soon paid attention as where Sony leads others quickly follow.
By the end of 2014 there were more than 50 consumer electronics companies offering HRA products with notable announcements from Yamaha and TEAC. This trend continues in 2015 with other major consumer electronics manufacturers such as Panasonic set to launch products.

Is HRA finally on the cusp of becoming a mainstream technology enjoyed by the mass market and not just audiophiles?

2. HRA goes mobile

Smartphones offer convenience. A single device allowing each of us to stay in contact, access information and listen to our favourite music whenever and wherever we want. However not everyone is happy with the sound fidelity or listening experience on the average smartphone, a fact recognised by the leading smartphone manufacturers who have introduced support for playback of Hi-Res Audio files (typically 24-bit / 192kHz) on their high end smartphones.

LG led the way in 2013 with the release of the G2 and now all the leading brands including Samsung, Sony and HTC support HRA playback on their flagship smartphones & tablets. Today Apple doesn’t support hi-res audio playback on the iPhone 6 over the headphone jack, though it is possible to access Hi-Res Audio via the lightening connector.

3. Increasing availability of Hi-Res content

You’ve got the equipment but where do you find the Hi-Res files to play? Well the good news is that there are an ever growing number of online stores offering Hi-Res audio downloads and as well as the emergence higher quality streaming services laying the first tentative steps towards high resolution streaming.

No longer is Hi-Res content just the preserve of specialised classical download stores or high-end AV brands, such as Linn and Naim.
2014 saw a rise in independent stores across all regions offering high resolution downloads catering for all genres including SuperHiRez by Acoustic Sounds, PonoMusic and Quobuz, adding to the original pioneers such as HDTracks and HiResAudio. Recently Technics and Onkyo have also lent their weight to Hi-Res audio with the launch of the Technics Tracks and e-Onkyo download stores.

If download isn’t your thing, don’t worry. There are several CD-quality streaming services already available from Tidal, Qobuz and DeezerElite with the first ‘true’ Hi-Res streaming service is expected before the end of the year.

But the story doesn’t stop there. One of the most exciting announcements of last year was the announcement by the world’s 3 largest record labels (Sony Music Entertainment, Universal Music Group and Warner Music Group) that they are making their back catalogues available for re-mastering to increase the availability of content and are encouraging artists and studios to record new music in high resolution.

So 2015 could be the year when critical mass is reached and high resolution content becomes as freely accessible as MP3s?

4. Cross industry collaboration & support

For any market to reach its full potential requires cross-industry cooperation and collaboration. In the case of Hi-Res Audio this effort was led by the Consumer Electronics Association (CEA) in conjunction with retailers, artists, record labels and major electronics manufacturers including LG, Samsung and Sony to promote the technology and benefits.

This collaboration has also led to a formal definition for Hi-Res Audio being agreed by industry bodies, the CEA, Digital Entertainment Group (DEG) and The Recording Academy, and record labels, Sony Music Entertainment, Universal Music Group and Warner Music Group.

The CEA has also partnered with the Japan Audio Society (JAS) to promote Hi-Res Audio and nurture the nascent market, including the rollout of a registered trademark and logo program to enable consumers to easily and quickly identify compliant equipment and software.

Last, but not least, Hi-Res Audio has high profile backing from artists such as Neil Young, and Moby as they strive to bring the listener into the studio and experience the music as the artist intended. Watch YouTube videos here…

5. Consumer demand for improved sound fidelity

Finally and perhaps most importantly is the continuing demand from consumers for improved sound fidelity. This is reflected in the increasing sales of headphones and the dominance of high-end brands such as Beats and Sennheiser, with global sales for 2014 forecasted to exceed 286 million units.

The strong growth has seen the return of legendary audio brands such as TDK, and a raft of new entrants including audiophile equipment vendors Onkyo, Oppo and Music Fidelity, consumer electronics giant Samsung, as well as legendary brands McIntosh and KEF.

Now some would say that headphones are more about fashion than a reflection on sound quality but the Consumer Electronics Association (CEA) disagrees. In studies they found that sound quality is the most important component of a quality audio experience for nine out of 10 consumers. And more importantly amongst consumers with only a moderate interest in audio, 60% were willing to pay more for high-quality audio devices.

&nbsp

Against such a backdrop of market forces, converging technology, industry support and latent demand, is Hi-Res audio set to become the digital music format of the future? Let us know what you think.

So you love your music. You value those immersive moments in sound. You’ve got the high end headphones. But still something is missing. The music lacks the depth and clarity of a live performance. Could a DAC (Digital to Analog Converter) be the answer? In this blog we explore the world of DACs, what they are and why you should consider using one.

What is a DAC?

Simply put, a DAC takes an input of digital audio data and translates it into an analog signal for use by speakers, headphones, amplifiers etc. Computers, smartphones and digital music players (such as iPod) store audio data as digital signals, which has to be converted into an audio signal that humans, speakers and headphones understand. The quality and volume of the music you hear is determined by the accuracy of the DAC and the precision of the conversion process.

Many devices include in-built DACs, including your phone, MP3 player, receiver, AV processor, computer, laptop, CD/DVD/Blu-ray player with analog outputs, wireless speakers, clock radios, and more. Essentially anything that can accept a digital signal and produce sound includes a DAC.

Why use an external DAC?

If your devices already have in-built DACs, why do you need an external one? Well, an external DAC bypasses the internal DAC and fixes several common problems such as:>/p>

• Noise floor issues (i.e. background hiss)
• Sound quality
• Jitter introduced during digital to analog conversion
• Old equipment
• Low headphone volume and loudness

Noise floor problems are easily identified. If you hear a hiss during the quiet sections of your music, or the playback is disrupted by noise, you can use an external DAC to clean up the signal as it is almost impossible to identify and eliminate the source of the noise. The noise source could be almost anything from fan noise to hard drives spinning to changes on our computer screen or layout issues with the in-built sound card, making it difficult to eliminate.

Sound quality and fidelity are subjective and dependent on a number of factors. However the accuracy of the system DAC directly determines the sound quality of the music you hear. A high end DAC can let you experience your music with better clarity, finer detail, and more sonic realism. With their advanced processors and electronics, external DACs can deliver greater precision and accuracy than most in-built DACs.

Reducing jitter. Sound quality issues can be harder to diagnose than noise floor problems. If you hear warbling (where it sounds like the music is being played on a slightly warped vinyl record) this is a sign there is “jitter” in the system. Jitter can be a sign of clock issues due to the misalignment between the source clock and DAC. An external DAC which supports asynchronous clocking can resolve this by using an independent clock source.

Upgrading older equipment. Digital technology is constantly changing and evolving. Even a five-year system has DACs that have probably seen improvements since that time. Adding an external DAC upgrades the player without replacing it, extending its useful life. If your player (or computer) doesn’t play the latest high resolution audio formats (such as PCM384kHz or DSD256), or you want to enhance the sound quality of your music, or improve the sound quality of video games, then it could be time to upgrade to an external DAC as the in-built DAC may be the limiting factor. In short, external DACs are an effective way to upgrade sound quality of many audio sources without replacing them.

Increasing headphone loudness. When using high-end headphones, you might find that the output of your device just isn't loud enough, even at full volume. To overcome this issue you need a headphone amplifier. Many external DACs integrate headphone amplifiers as well, so you can solve your loudness problem and experience the benefits of a DAC as well.

Finally a couple of words of caution. Remember the old adage “garbage in, garbage out”. If you are listening to low quality source (i.e. low bit rate audio or a poor recording) then no amount of electronics is going to solve the problem. Digital audio quality has also improved significantly over the last five years, but you won’t hear the full quality of the latest high resolution audio files with old hardware.

Conclusion

External DACs can be used to solve noise issues, combat jitter and playback newer audio formats. If you want to hear music as it musicians intended you to, or simply to get the best from those expensive headphones, it’s probably time to try an external DAC.

In future blogs we will take a look at the key features of precision DACs, why they matter and how they are implemented.

&nbsp

&nbsp

Got a suggestion for a future blog topic, then let us know.

Emphatically yes! Here we will take a quick tour of the streaming landscape and which service providers are on track to deliver true high resolution streaming not just souped up MP3 at 320kbps, this year.

First some background. The way music is consumed has undergone a seismic shift, and 2015 is likely to see another momentum shift with the rise of Hi-Res Audio (see our 5 Key Trends blog for more on this) and the launch of the first true Hi-Res Audio streaming service.

In the last decade the music industry has seen the decline of the CD, the rise (and fall?) of digital downloads and the emergence of on-demand streaming.

On-demand streaming has grown exponentially since the first service was launched, and the latest reports estimate that streaming accounts for more than a quarter of digital music revenues compared to just 3% in 2007. And this trend is expected to continue.

While streaming offers the ultimate in convenience and accessibility, historically it has come at the cost of audio quality. Thankfully for music lovers everywhere the industry is refocusing on audio quality. 2014 saw a number of ‘CD quality’ streaming services launched and the first true high resolution streaming services is not far away.

Lets take a quick tour of the CD quality streaming services now available and who is in the race to launch Hi-Res streaming.

OraStream

www.orastream.com

Launched in 2013, Singapore based OraStream was announced as the world’s first high resolution cloud music service . OraStream is unique in the marketplace. It offers lossless audio in conjunction with a proprietary adaptive bitrate streaming technology that dynamically adjusts the bit rate to suit the device and connection bandwidth being used for playback. So if you are using WiFi then it is possible to receive the full glory of streaming of high-resolution music up to 9,216 kbps (or 192kHz/24-bit), but if the connection weakens then the service will stream a lower quality track while keeping the music playing without any disruption! Unfortunately, OraStream has only a small catalog which is restricted to just specialized classical and electronic dance music, although you can upload your own tracks for playback in the cloud.

Qobuz

www.qobuz.com

Qobuz is an on-demand streaming service and download store. Launched in in France in 2008, the service went global at the end of 2013 and is now available in 10 countries across Europe.

Qobuz offers CD quality lossless streaming at 16 bit/44.1kHz as well as Studio Master downloads at 24 bit/192kHz. With a catalog covering 28,000 labels and producers, and offering more than 20million tracks it caters to all music tastes.

Deezer

www.deezer.com

Deezer is best known for its MP3 based streaming service and is one of the top 5 vendors globally. 2014 saw them enter the lossless streaming market with the launch of Deezer Elite an on-demand streaming service for the Sonos platform.

Deezer Elite debuted in the US in September 2014 and as a result of stronger than expected demand is now being rolled out in more than 150 countries . With more that 25million tracks out of a 35million catalog available in lossless format, Deezer Elite offers the widest selection today.

Tidal

www.tidalhifi.com

In the world of high quality music streaming, Tidal are the new kid on the block. Launched at the end of 2014 by Scandavian company Aspiro , Tidal is currently available in 26 countries and has a catalog in excess of over 25 million tracks.

Like Qobuz and Deezer, Tidal offer lossless on-demand streaming. What differentiates Tidal from the others is the accessibility of the platform via their partnerships with leading audio hardware brands including Sonos, HEOS by Denon, Bluesound, NAD and Meridian.

With the backing of Jay-Z and the recent demonstration of MQA (Master Quality Authenticated) streaming in conjunction with Meridian, Tidal look set to win the race to high resolution audio streaming.

Apple, iTunes & Beats

No round up of the music scene would be complete without a word about Apple. There are plenty of rumours circulating about what Apple may or may not do with the Beats on-demand streaming music service they bought last year. Will it be standalone, will it be integrated into iTunes, will it replace iTunes?

What is certain is that as the undeniable leader in online music sales, Apple are seeing declining revenues from digital downloads. So surely it is only a matter of time before they offer a streaming service. The question is whether they will back quality in the form of lossless high resolution audio, or whether they will continue with the existing Beats model of convenience over quality. Ultimately only time will tell.

Conclusion

Hi-Res Audio is here to stay as consumers demand an improved audio experience and evidenced by the increasing sales of audio accessories (such as DACs , headphone amps and high-end headphones).

Until now listeners have had to choose between the convenience of streaming MP3s or the quality of Hi-Res downloads. Not anymore. The race is already on to win the accolade of the world’s first Hi-Res streaming service, with Tidal and Deezer edging in front. The question is who will be first over the line?

&nbsp

Let us know what you think.

Welcome to the first in our series of featured guest blogs, where XMOS customers discuss their experiences of using xCORE multicore microcontrollers.

First up we talk with Jason Liao, CTO and VP Product Development of OPPO Digital, Inc. OPPO Digital designs and market high quality digital electronics that deliver style, performance, innovation, and value to A/V enthusiasts and savvy consumers alike. OPPO’s attention to core product performance and strong customer focus distinguishes it from traditional consumer-electronics brands.

XMOS: How long have OPPO being using xCORE processors?

BD-105 Blu-Ray Player

Jason: We started using XMOS in our award winning BD-105 Blu-Ray Player which launched back in 2012. Since then we’ve used xCORE processors in our HA-1 Headphone Amplifier and most recently in the HA-2 Portable Headphone Amplifier & DAC to deliver bit-perfect audio over USB.

XMOS: What are the key features OPPO look for when choosing a supplier for USB audio?

Jason: Every product selection we make is driven by a single question, “Does it deliver the best possible customer experience?” In terms of USB audio, the customer experience is influenced by three key factors: audio fidelity; file format support; and device /platform support. At OPPO, audio fidelity means that we faithfully reproduce the original recording as the artist heard it in the studio. To deliver the audio fidelity our customers demand, we use the XMOS Asynchronous USB solution to realise a low-jitter, bit perfect audio system.

HA-1 Headphone Amplifier

High resolution audio files come in a bewildering area of different formats and sample rates, and our customers demand audio equipment which can handle any of the formats available today. XMOS enables us to natively support both PCM and DSD files without any conversion or pre-processing. On our latest product, the HA-2, we offer PCM audio up to 384kHz/32-bit and DSD audio up to 11.2MHz.

When investing in high-end audio equipment, consumers are looking to use the product with multiple devices, including the latest smartphones and tablets! The XMOS platform offers the broadest platform/OS support of any USB audio vendor, with PC, Mac OS X, iOS and Android all available and verified to XMOS’ high quality standards. This means that we can focus on what we do best, building great analog products.

XMOS: How would you sum up your experience of designing and developing products with XMOS?

HA-2 Portable Headphone Amplifier

Jason: OPPO and XMOS share a common ethos, to exceed the customer’s expectation and deliver excellent audio products which excite the listener and bring them closer to the live performance. This shared understanding makes working together incredibly easy as we are striving for the same goals. The combination of XMOS’ unique xCORE microcontrollers and the feature rich USB Audio reference software mean that we are able to rapidly develop products with the distinctive OPPO feel.

Find out more about OPPO Digital and the products at http://www.oppodigital.com/

For more information on how XMOS makes audio sound better go to the High Resolution Audio page.

audio-format_graphic_small.png

This is the first instalment of a two part blog, which provides a brief introduction of audio formats that have emerged since the 1980s.

Why the 1980s? Nothing to do with frilly shirts and leg-warmers, it was when the tide firmly changed from analog to digital audio. Sure, analog audio still existed in the 1980s (and continues even now) but the focus of technology development switched firmly to digital.

To kick off, a couple of notes. Arguably, more than any other market, the audio space is beset with passionate campaigners, so may I first apologize for offending any of them with any ignorant and/or over-simplified explanations. Secondly, like modern life, this article would not have been possible without Wikipedia.

CD – the arrival of digital audio

The story of Digital Audio begins in the early 1980s with the introduction of the first ubiquitous digital audio source, the humble Compact Disc (CD). CD’s changed the audio world beyond recognition. Thanks to the uniformity of the physical media and its robustness, for many CDs heralded an era of audio clarity the like of which we had never seen (or heard).

However sound is an analog waveform and CD’s store data in binary which is digital. So a way was needed to ‘digitise’ or encode the analog sound wave. The solution used on CDs was to encode the analog sound waves using Linear PCM (Pulse Code Modulation).

PCM

Linear PCM (often referred to as just PCM) is an encoding scheme in which the audio is sampled at regular intervals, and the amplitude is then digitized (or quantized if you are in the know, which you are now). The linear bit just means that the range of possible values used to represent the sound at any instant is spread equally between the minimum and the maximum volume (amplitude is the techy word to use here).

Sample rate & bit depth

Audio quality is determined by two parameters – sample rate and bit depth.

Sample rate is the number of samples taken per second. The more samples taken per second, the more accurate the digital representation of the sound can be. PCM sample rates range from 44.1kHz for CD beyond 192kHz for high resolution audio.

Each sample is really only a number that contains the amplitude value of a waveform measured over time. This brings us to bit-depth.

Bit depth is the number of bits used to describe each sample and defines the number of possible values that each sample can take. Higher bit depth audio allows for more dynamic range (enabling louder peaks and softer quiet sounds) and a lower noise floor (less noise). Smaller bit depth audio takes up less hard drive space but are inherently noisier. So higher bit depth audio should sound better than smaller bit-depth audio. PCM bit depths range from 16-bit for CD up to 32-bit for studio masters.

A final word on bit depth. Bit depth is exponential, so a 24-bit sample holds 256 times as much information as a 16-bit sample. A 16-bit sample has a range of 65,536 values, while a 24-bit sample has a range of 16,777,216.

PCM remained as the dominant encoding technique for digital audio until the introduction of the Super Audio CD (SACD) in 1999 which used DSD to encapsulate a higher resolution layer alongside an optional CD layer with the familiar 16-bit, 44.1kHz PCM encoding.

In the next instalment we will discuss DSD. If you can't wait click here to download the full article.

&nbsp

Welcome to the second instalment of our Hi-Res Audio Formats overview. In the fist instalment we looked at the Liner PCM (Pulse Code Modulation) encoding scheme used on CDs and which continues to dominate the digital audio world today. If you missed it, you can read the blog here.

Today we are going to look at the other primary format for high resolution audio files, DSD or Direct Stream Digital.

PCM remained as the dominant encoding technique for digital audio until the introduction of the Super Audio CD (SACD) in 1999 which used DSD to encapsulate a higher resolution layer alongside an optional CD layer with the familiar 16-bit, 44.1kHz PCM encoding.

DSD

So what is DSD? Time to don your technology trousers. To kick off, it stands for Direct Stream Digital (which is not particularly enlightening but you’ll imagine by now that finding an unused acronym is tough in this space).

DSD recordings use Pulse Density Modulation (PDM), which have a much higher sampling frequency than PCM signals. The standard DSD sample rate is 2.8224MHz, which is 64 times the 44.1kHz used by CD. The factor 64 is not randomly chosen. It's a power of two, meaning that it's relatively straightforward to upsample from typical PCM frequencies including 44.1 kHz and multiples like 88.2 kHz.

Click to Enlarge

Unlike PCM signals, PDM takes the same signal and encodes whether it is higher (the amplitude of the signal is increasing) or lower (the amplitude of the signal is decreasing) than the long-term average of that signal. Translation? Being a 1-bit signal means that every sample can only have two values, 1 or 0. So every bit in a PDM stream only tells whether the amplitude of the signal was higher or lower than the previous sample. The result is that the signal is encoded by the frequency or density of the 1s and 0s as shown the diagram below.

Any noise that occurs in the DSD signal as a result of the fast but clunky 1-bit quantization can be moved out of the audible range by more magic, called noise shaping. With some cleverness, the original signal can be recovered from a DSD stream using a simple analog low pass filter. Whereas if you did the same with PCM you would get only noise rather than your favourite track.

While the SACD has faded into the annals of history, DSD (PDM) continues to be a popular encoding technique used across the recording industry thanks to the might of Sony Music and support for DSD (PDM) on Playstation 3. This has driven the continued development of the technique with the introduction of double-rate and quad-rate DSD.

As the name suggests double-rate DSD (aka DSD128) uses a sampling frequency of twice standard PDM, 5.6448 MHz, which is 128 times the sample rate of a standard CD. Whereas quad-rate DSD (aka DSD256) uses a sample frequency 4-times standard PDM, 11.2MHz or 256 times a standard CD.

MP3 – A change in the perception of audio

In similar time to the emergence of DSD, MPEG-2 Audio Layer 3 (more commonly known as MP3) came on the scene. I don’t intend to say much about MP3 here, as its lossy format is not suitable for Hi-Res Audio, but it’s important to recognise that the arrival of MP3 revolutionised the audio market and it remains the most commonly listened to format today.

There was nothing special about audio any longer, it was just data that you could download, copy, keep and share - just like any other data. All you needed was a device to store the data, and a device that could take the digital data and convert it into an analog signal for playback over speakers or headphones (generally referred to as a DAC: digital-to-analog convertor).

DoP

There was a problem, however, for people who wanted to play DSD files over the Universal Serial Bus (USB), which had become the most common connector for audio devices. The USB Audio Class 2.0 (UAC2.0) specification included PCM support for high resolution audio, but there was no native support for DSD encoded audio; it could only be transferred using a general “raw data format”. To make matters worse, when Apple introduced MAC OS X 10.7 the USB driver only supported PCM. If you wanted to listen to a DSD file, it first had to be converted to PCM before playback.

To overcome these issues , a consortium of companies defined a standard for transporting DSD data inside a PCM frame, hence “DSD over PCM”, or DoP was born.

Standard DSD (aka DSD64) has a data rate of 2.8224Mbits/s (that is 1-bit resolution at a rate of 2.8224MHz.) This data rate is equivalent to a 16-bit PCM at a rate of 176.4kHz, but a method is needed to identify when a PCM stream contains DSD data. So the DSD bitstream is encapsulated in a 24-bit PCM frame and the additional 8 bits are used as a flag to mark the frame as containing DSD data.

Even though DoP enables DSD data to be transferred over USB, you still need a suitably equipped audio device that recognizes the DoP signal and converts it to an analog output the rest of your audio system can handle.

Conclusion

So what of the future? We can be sure that streaming audio over USB is here to stay, and very high channel counts are also becoming common-place in many areas of our lives.

What of DSD and PCM? DoP has given a new lease of life to the firey debate over the superiority (or otherwise) of DSD over PCM. For now, they are both locked in an ever higher resolution audio spiral - LPCM by increasing sample rates and bit depths (96kHz, 192kHz, 384kHz and beyond at 24, 32 bits) and DSD by increasing 1b sample rates (double @ 5.6448 MHz, quad-rate at 11.2Mhz, octuple rate at 24.576MHz).

Which one is better? That is something worth listening out for…

&nbsp

Welcome to the second in our series of featured guest blogs, where XMOS customers discuss their experiences of using xCORE multicore microcontrollers.

This time out we talk with Thorsten Loesch, Director of Technology for iFi Audio.

XMOS: Can you tell us a little about iFi Audio and their products?

Thorsten: iFi Audio is part of the Abbingdon Group Limited (AGL) which also owns the renowned Hi-Fi brand Abbingdon Music Research (AMR).

iFi Audio drills down to the silicon-die level to design and manufacturer portable and desktop ‘ultra-fidelity’ audio products. Our in-house hardware and software development teams enable us to bring advanced audio products which continue to push the audio frontier.

XMOS: When did iFi Audio start using xCORE processors?

Retro Stereo 50

Thorsten: Our sister brand AMR was one of the first premium Hi-Fi brands to adopt xCORE multicore microcontrollers back in 2011, and quickly realised that XMOS USB performance levels were far higher than other vendors. Since then we’ve used xCORE processors and their USB audio solutions across our range of award-winning, high-end audio products including Nano and Micro iDSD as well as the recently launched Retro Stereo 50.

XMOS: What is the benefit to iFi Audio using XMOS technology?

Thorsten: At iFi we believe in delivering the best audio experience possible. For us that means delivering bit-perfect, ‘ultra-resolution’ digital audio whatever the source.

nano iDSD

Take the Nano and Micro iDSD which turn any Smartdevice into a portable ‘ultra-resolution’ system. Many competitive systems support both native PCM and DSD files, and the ‘best-in-class’ today offers dual-speed DSD (DSD128 or 2xDSD) and PCM384. But since August 2014, we have been shipping state-of-the-art DSD512/ PCM768.

The Nano iDSD supports quad-speed DSD(DSD256 or 4xDSD) and the Micro iDSD supports octa-speed DSD(DSD512 or 8xDSD) with asynchronous audio clocking enabled by xCORE technology. The result is advanced digital-to-audio conversion leaving recording studios to play ‘catch up.’

Thanks to the XMOS architecture, we will continue to push the DSD envelope beyond the assumed capacity. Offering DSD512, bit-perfect, asynchronous USB audio allows us to front-line consumer ‘ultra-fidelity’ audio. We will continue to break through audio barriers with XMOS on future generations aimed at discerning consumers around the world.

micro iDSD

XMOS: How would you sum up your experience of designing and developing products with XMOS?

Thorsten: Both iFi and XMOS strive to exceed customer’s expectations and push the boundaries of what is considered possible, resulting in the delivery of ground breaking products which please the listener and change the way they experience music. This shared understanding makes working together incredibly easy as we are striving for the same goals.

XMOS’ USB audio solutions are based on the xCORE range of multicore microcontrollers. Programmable in ‘C/C++’ our software designers can rapidly develop new features while leveraging XMOS’ feature rich reference software to ensure that iFi continues to innovate and push the envelope of what is thought possible.

Find out more about iFi Audio and their products at http://ifi-audio.com/.

For more information on how XMOS makes audio sound better go to the High Resolution Audio page.

For those readers not subscribed to our weekly newsletter here is a taste of what you are missing.

Here are our top picks from the last 7 days of Hi-Res Audio news. There is definitely a streaming theme to this week's round-up.

The biggest story of the week has to be Apple Music!

The streaming music announcement at the Apple World Wide Developers Conference, said nothing about support for high resolution audio, but we're sure it's going to change the way we consume music. If you haven't listened to the announcement yet it starts at around 103 mins into the video.

Do you want Hi-Res Audio streaming on your smartphone?

What Hi-Fi report that Qobuz will be the first streaming service to bring true high resolution audio streaming to Android smartphones. Sublime subscribers will be able to stream 24-bit Hi-Res tracks on compatible Android devices from 15 June.

Do you want Hi-Res Audio streaming on your smartphone

The online music store HIRESAUDIO has announced VirtualVault, which allows customers to stream their audio files to mobile devices, desktop web browsers and car web browsers. The company uses Adaptive Audio Streaming to deliver the best possible sound quality. Click to find out more.

NAD C 510 Direct Digital Preamp DAC

On the hardware front, Audiostream get their hands on the NAD C510 digital pre-amplifier. Does it deliver the precise presentation of the music captured in the recording studio? Find out now.

Want to be kept up-to-date with our latest posts? Follow our Twitter and Facebook pages so that you are notified the moment a new piece of content goes live.

Sign up now to receive future newsletters straight to your inbox

Did you know we publish a weekly Hi-Res Audio newsletter covering the latest XMOS and industry news? For those not subscribed to the newsletter here is a typical example from June 17, 2015.

Here are our top picks from the last 7 days of Hi-Res Audio news.

Crowdfunding is a great way to bring new products to market, including DACs/headphone amplifiers. Two market leading designers are launching products through INDIEGOGO.

DACport HD: CEntrance

Michael Goodman and CEntrance invented the 'dongle DAC' back in 2011. The new DAC/headphone amplifier, DACPort HD, has a real volume knob, low/high gain settings, DSD decoding and can handle headphones from large over-the-head cans to sensitive IEMs.

Geek Out V2: LH Labs

Following customer feedback on the first Geek Out DAC, LH Labs have designed V2 with support for real 32 bit/384 kHz PCM audio and native DSD128. The Geek Out V2 amplifier is up to 10 times more powerful than standard headphone amplifiers but still remains crystal clear. Find out more here.

iFi Audio discuss why XMOS is one of the special ingredients in their award winning products

iFi Audio designs and manufacturers portable and desktop ‘ultra-fidelity’ audio products. As part of the Abbingdon Group Limited (AGL) they were one of the first premium Hi-Fi brands to adopt xCORE multicore microcontrollers back in 2011, and continue to use the technology to push the envelope of digital audio beyond what is thought possible. Find out why they use XMOS.

Emotiva launches Ego USB DACs

Emotiva have announced the Big Ego and Little Ego DACs, which play high resolution computer audio files up to 32/384 PCM data. The Big Ego and Little Ego will be manufactured in their new Tennessee facility, as part of their "Made-in-USA" initiative.

Want to be kept up-to-date with our latest posts? Follow our Twitter and Facebook pages so that you are notified the moment a new piece of content goes live.

Sign up now to receive future newsletters straight to your inbox

Did you know we publish a weekly Hi-Res Audio newsletter covering the latest XMOS and industry news? For those not subscribed to the newsletter here is a typical example from June 28, 2015.

XMOS silicon devices are used in audio products worldwide to enable Hi-Res Audio.

Here are our top picks from the last 7 days of Hi-Res Audio news.

Hi Resolution Audio is a lot older than you thought!

Steve Guttenburg for CNET asks what was the original high-res audio format? Was it Super-Audio CD, DVD-Audio or perhaps even vinyl LP? Find out Steve's answer here.

DoubleTwist's new Android app turns Dropbox into a music streaming service

Double Twist have released their Cloud Player app which will link to, and stream your music files from DropBox, OneDrive and GoogleDrive. This could be a great way to access your content, Hi fidelity sound is supported with Apple Lossless and FLAC support, Double Twist have tested with 24bit 192kHz streaming. The player is pretty unique in supporting both Chromecast and AirPlay. Read more at The Verge.

Trusted Reviews gets hands-on with Cambridge Audio CXN Network Music Player: Makes your digital music sparkle

Danny Philips of Trusted Reviews takes a look at Cambridge Audio's upscaling network music player the CXN: "...in terms of audio format support CXN leaves no stone unturned. Among the playable formats are 192kHz/24-bit FLAC, ALAC, WAV and AIFF, .... the CXN's powerful second generation ATF2 engine upsamples any source to 384kHz/24-bit to reduce jitter."

TASCAM release free Hi-Res Audio Editor

This free Hi-Res Editor software is one of the only ways to edit native DSD master recordings.

Want to be kept up-to-date with our latest posts? Follow our Twitter and Facebook pages so that you are notified the moment a new piece of content goes live.

Sign up now to receive future newsletters straight to your inbox

Back in March of this year we announced a first for XMOS … a dedicated audio processor, called xCORE-AUDIO Hi-Res 2. Well we’ve had the samples back in our lab for a while now and the device is now ready to be let out into the wild. Here at the factory we are very excited about the new development.

So what is xCORE-AUDIO Hi-Res 2?

Simply put it’s a dedicated audio processor which offers high resolution stereo (hence the 2 … clever eh?) audio interfacing for consumer audio applications such as DACs and headphone amplifiers.

Click to Enlarge

xCORE-AUDIO Hi-Res 2 provides a bit-perfect, asynchronous USB interface for audio streaming on Windows, Mac OS X, iOS and Android. The device supports the full range of high resolution files formats including PCM up to 384kHz/32-bit and DSD 2.8/5.6 MHz (aka DSD 64/128). The chip delivers a precise output with low noise/distortion for soothing, high-quality sound that allows the listener to get closer to the music.

With eight GPIO (General Purpose Input/Output) and a further eight GPI (General Purpose Input), product designers can easily configure the device to meet the exact needs of their HMI (Human Machine Interfaces).

For high volume customers in the DAC/headphone amplifier space, we’ve developed a reference design that actually implements a high resolution portable DAC/HPA, which you can use to evaluate xCORE-AUDIO Hi-Res 2. If you’re just browsing at this point, then download the datasheet and schematics for xCORE-AUDIO Hi-Res 2 from www.xmos.com/products/silicon/xcore-audio/hires.

To program or not to program?

Some of you reading this blog may already be familiar with XMOS for our programmable multicore microcontrollers or maybe even our market leading USB2.0 audio solutions and software. You may also be wondering where xCORE-AUDIO fits in our product portfolio.

In developing the xCORE-AUDIO Hi-Res 2 we leveraged all the knowledge and understanding that we’ve gained developing and delivering world beating UAC2.0 solutions, then packaged into a more accessible format.

Remember that we still provide our software reference design free of charge, and development boards like the eXplorerKIT, for developers who want more features and are happy to get stuck into the code.

So that’s xCORE-AUDIO Hi-Res 2, the first member of our new xCORE-AUDIO family of devices, which free product designers from the challenges of implementing a bit-perfect digital audio interface, allowing them to focus on the analog and acoustic design, user interfaces and the total product experience. Safe in the knowledge that the digital processing and interfacing is being delivered by the market leading vendor of USB 2.0 audio solutions.

If you are want access to the most responsive, flexible microcontroller on the market, but don’t have the time or resources to deploy on developing a ground-up solution, maybe the xCORE-AUDIO Hi-Res 2 device is the processor for you. Why not take a look and find out?

Welcome to our latest guest blog, where XMOS customers discuss their experiences of using xCORE multicore microcontrollers.

Joining us this time is Dominic Baker, Technical Director at Cambridge Audio.

For more than 40 years, UK-based Cambridge Audio has been at the forefront of hi-fi innovation, designing and building a range of award winning products for music lovers. Each product is designed to deliver the best performance and sound quality from a range of sources.

XMOS: How would describe Cambridge Audio’s DNA?

Dominic: Simple really. Sound first. Music should always sound amazing. This philosophy drives all our product developments as we strive to create dynamic, powerful products that delight people who really love music.

Everyone at Cambridge Audio loves music, and this passion enables us to design innovative and sublimely engineered products that deliver an unparalleled sound.

XMOS: When did Cambridge Audio start using xCORE processors?

DacMagic 100

Dominic: Being a British based company we were one of the first audio equipment vendors to adopt XMOS’ revolutionary technology. We first started using xCORE multicore microcontrollers in 2011 on our DacMagic Plus high performance DAC product, and very quickly realised that XMOS’ USB solutions were far superior to other vendors offering a richer feature set and higher performance Asynchronous mode. Since then we’ve used xCORE processors across our rage of award-winning audio products including DacMagic 100 as well as our latest hi-fi separates 851N and CXN/CXA products. Of course it helps that they are British too!

XMOS: What is the benefit to Cambridge Audio using XMOS technology?

CX series

Dominic: At Cambridge Audio we believe nothing should come between the listener and the music. Our goal is to deliver bit-perfect audio whatever the source. Thanks to the inherent flexibility in XMOS’ audio solutions, realising this goal is made easier and quicker.

XMOS delivers us the flexibility to support a range of audio formats and sample rates, from 32kHz/16-bit all the way up to 192kHz/24-bit and beyond. They also enable us to natively support both PCM and DSD files without any conversion or pre-processing. Not only does this simplify our design processes it also means that we can use xCORE processors in standalone mode, making it an ideal solution for a compact headphone amplifier such as the DacMagic XS.

DacMagic XS: Small but mighty

For my engineering team perhaps the biggest benefit that XMOS has provided is in simplifying our development flow. Now we have a single platform that can be used across the product range, our development cycles are shorter since we spend less time switching between the different design methodologies used by the various chip vendors.

XMOS: How would you sum up your experience of designing and developing products with XMOS?>

Dominic: Both Cambridge Audio and XMOS want to bring the listener as close to the music as possible, almost as if they are in the same room as the artist. This means developing products which deliver precise reproductions with astonishing clarity.

XMOS USB audio solutions based on xCORE multicore microcontrollers are fully programmable in ‘C/C++’ making it an easy task for our engineers to implement bit-perfect, asynchronous USB audio streaming and deliver innovative, award winning products that remain true to our philosophy of “sound first”.

Find out more about Cambridge Audio and their products at http://www.cambridgeaudio.com/.

For more information on how XMOS makes audio sound better go to the High Resolution Audio page.

Did you know we publish a weekly Hi-Res Audio newsletter covering the latest XMOS and industry news? For those not subscribed to the newsletter here is a typical example from July 07, 2015.

XMOS silicon devices are used in audio products worldwide to enable Hi-Res Audio.

Here are our top picks from the last 7 days of Hi-Res Audio news.

Best of Both? Vinyl and Hi-Res

With the rise of vinyl sales and streaming services, British manufacturer Musical Fidelity are looking to cash in with the Merlin 1. Trusted Reviews take a look at the system which promises the best of new and old audio worlds in a single device.

iRig HD/PRO users get upgrade to studio quality

IK Multimedia announced a free firmware upgrade for iRig HD and iRig Pro audio interfaces for existing users. The update delivers audio at 24bit/96kHz and a promise of "even more clean, nuanced and transparent sound" from their ultra-compact mobile interfaces.

Foo Fighters join Sony for the Hi-Res battle

The Foo Fighters, who are renowned for an uncompromising commitment to audio quality, have announced a partnership with Sony to promote Hi-Res Audio worldwide. Find out how they are going to convert us - One By One.

Hi-Res Audio, as clean as mountain air

With a design inspired by Switzerland's Matterhorn, the Astell and Kern AK500N reproduces your music with as little distortion as possible - it even uses a battery to keep it clear of noise from the mains wiring. Jeremy White, The Telegraph, is impressed.

Want to be kept up-to-date with our latest posts? Follow our Twitter and Facebook pages so that you are notified the moment a new piece of content goes live.

Sign up now to receive future newsletters straight to your inbox

Did you know we publish a weekly Hi-Res Audio newsletter covering the latest XMOS and industry news? For those not subscribed to the newsletter here is a typical example from July 16, 2015.

XMOS silicon devices are used in audio products worldwide to enable Hi-Res Audio.

Here are our top picks from the last 7 days of Hi-Res Audio news.

xCORE-AUDIO Hi-Res 2 now available

Our latest audio processor for high resolution consumer applications is now available. Offering stereo audio interfacing, the xCORE-AUDIO Hi-Res 2 supports the full range of high resolution formats including PCM 384kHz/32bit and DSD 2.8/5.6MHz.

Sony's SRS-X88 isn't any ordinary wireless speaker

According to Adrian Weckler, Technology Editor for the Irish Independant, Sony's support for high resolution audio and the ability to upscale low resolution files is leading the market.

Hi-res streaming will soon be the standard

"We will be streaming Hi-Res audio as a normal way of accessing music in a couple of years" - that's the view of Pete Downton, 7digital's Chief Commercial Officer.

Audient iD14 wins Future Music award

iD14 was recently reviewed by Future Music Magazine's Jono Buchanan and he liked it so much he gave it the much coveted Platinum Award. Read the full review at Future Music's sister platform Music Radar.

---

Like this news round-up? Why not sign up for free and get it delivered straight to your inbox?

Sign up now for free

Did you know we publish a weekly Hi-Res Audio newsletter covering the latest XMOS and industry news? For those not subscribed to the newsletter here is a typical example from July 23, 2015.

XMOS silicon devices are used in audio products worldwide to enable Hi-Res Audio.

Here are our top picks from the last 7 days of Hi-Res Audio news.

Turn it up to eleven !

Marshall will launch their first Android phone called the 'London' next month that is targeted at the hi-res music lover, it even has two headphone jacks so you can listen with a friend.

Fly me to the Moon !

Simaudio have introduced the MOON Neo-230HAD a DAC/Headphone amplifier which has 4 digital inputs including S/PDIF, Toslink and USB, supporting up to 32bit @ 384kHz PCM or DSD256.

Italian emotional appeal

What Hi-Fi take a look at the Pathos Acoustics Converto DAC, none of the usual valves but an uncluttered performer. "An engaging performance that’s lovely to listen to" gets it 4 stars.

Superior strength and agility

The Audeze Deckard finally lands for retail in the UK, this DAC and amplifer can drive any headphones you throw at it, and takes 32bit 384kHz in its stride.

---

Like this news round-up? Get it delivered straight to your inbox.

Sign up for Hi-Res Audio newsletter

Did you know we publish a weekly Hi-Res Audio newsletter covering the latest XMOS and industry news? For those not subscribed to the newsletter here is a typical example from July 30, 2015.

XMOS silicon devices are used in audio products worldwide to enable Hi-Res Audio.

Here are our top picks from the last 7 days of Hi-Res Audio news.

LH Labs pairing delivers a clean precise sound

Audiostream's Michael Lavorgna puts LH Labs Geek Pulse Xfi and LPS4 power supply through their paces and awards the pairing a coveted "Greatest Bits" award.

NuPrime DAC-10H

NuPrime, emerged from Optoma's acquisition of NuForce in late 2014. The DAC-10H is one of the first products to emerge from the new brand. An all-in-one DAC, pre-amp and headphone amp, the DAC-10H supports PCM up to 384kHz/32b and DSD256. Find out more in Inner Fidelity's review.

xCORE-AUDIO Hi-Res 2, our first audio processor

Earlier this year we announced the xCORE-AUDIO Hi-Res 2 family. In her latest blog Jacqui Adams gives a brief introduction to this exciting new product from XMOS.

AKG N90Q headphones

Sound Guys get hands on with the latest collaboration from legendary brand AKG and producer Quincy Jones. They may cost $1500 but you do get a built-in DAC which by-passes the one on your computer or Android device and the convenience of an all-in-one offering.

---

Like this news round-up? Get it delivered straight to your inbox.

Sign up for Hi-Res Audio newsletter

Did you know we publish a weekly Hi-Res Audio newsletter covering the latest XMOS and industry news? For those not subscribed to the newsletter here is a typical example from August 06, 2015.

XMOS silicon devices are used in audio products worldwide to enable Hi-Res Audio.

Here are our top picks from the last 7 days of Hi-Res Audio news.

Proof that Hi-Res Audio is better?

The debate surrounding whether there is a measurable difference between Hi-Res Audio files and lower resolution formats may have finally been answered by an ABX test carried out by Arny Kruger. Audiostream summarises the test results and asks if the debate is finally over.

Cambridge Audio's Dominic Baker talks to XMOS

For more than 40 years, UK-based Cambridge Audio has been at the forefront of hi-fi innovation, with a reputation for producing sublimely engineered products delivering a clean, crisp sound. Find out why they use XMOS in their award winning products.

Leema Acoustics: Libra DAC

Announced at the Bristol show earllier this year, the Libra DAC is now shipping. The Libra supports high-resolution audio up to 384kHz PCM and DSD 64/128 via USB as well as offering coax, optical and I2S inputs, the latter being via a RJ45 connector.

Hi-Res Audio player streams your music library

Radius' high resolution audio player is now available to download from the Apple App Store. Launched in Japan in April and rolled out globally last week this latest offering connects to any DLNA enabled device to playback audio up to 192kHz/24-bit on your iPhone.

---

Like this news round-up? Get it delivered straight to your inbox.

Sign up for Hi-Res Audio newsletter

Designing electronic products today is a highly complicated challenge that goes well beyond the old question of "How many MIPS does the microprocessor have?" Designers must solve the fundamental requirements of their products, while meeting strict international standards, and integrating innovative features to support IoT (Internet of Things) technologies. And consumer products also need to be fashionable as well.

The traditional approach is to use combinations of heterogeneous microprocessors to handle the each task, and ‘glue’ all the devices together. But this approach gets more difficult each time you have to add a new feature, with more complexity and opportunities for unexpected events to happen within the system. Combinations of silicon chips also increase the product BOM and development cost, power consumption, product size, when management want to reduce each cost while getting products to market quicker.

Let’s consider a new type of product that’s starting to appear in our homes – the Home Hub. The Amazon Echo is a simple example of a home hub or Virtual Digital Assistant (VDA); it has to wait in sleep mode, recognize keywords that wake it up, decode voice instructions into a PCM audio stream, optimize audio levels, manage physical input signals from buttons, communicate with a web server, receive an audio signal back and output it as an analog stream - quite a lot of tasks all-in-all. Analyzing these requirements, however, it’s not as bad as it initially appears. Most of the requirements come down to high quality analog and digital audio input/output coupled to some digital signal processing (DSP) for managing the microphone arrays, audio compression/decompression, and noise reduction. And finally it needs a robust connection to the Cloud services over your WiFi hub.

xCORE microcontrollers provide a uniquely optimized solution to these requirements. The multi-processor architecture and flexible I/O allow developers to configure the xCORE I/O ports to handle the audio, IoT and HMI requirements in software (XMOS provides libraries for Ethernet, USB, I2C, I2S, PDM etc). The I/O can be integrated with separate processing cores that run DSP algorithms and application code written in C/C++. xCORE devices are built around a 32-bit fixed point architecture, which includes a 64-bit accumulator for enhanced DSP code, an architecture that delivers the best available trade-off between performance / power consumption / cost.

Next time you design a product that needs to integrate high quality audio with DSP and IoT capabilities, try xCORE. It’s a great choice for conference microphones, voice-enabled gaming headsets, acoustic sensors for industry and security applications, home hubs and VDAs, as well as headphone amplifiers/DACs and DJ mixing decks. All the key technology requirements are integrated into a single deterministic device that removes most issues related to multichip systems while saving you development and BOM costs.

For products of today and tomorrow, integration is more important than the MIPS.

Accurate, accessible and economical Voice User Interface (VUI) design has been slow to emerge. Our ability to interpret dialogue of varying pronunciation, accent, speed and grammatical correctness is something that we humans take for granted from an early age, but that has proved elusive for machines. Nonetheless, VUI has become a mainstay of our predicted future - who can forget HAL 9000 or the "Talkie toaster" (immortalized by Grant Naylor in the BBC TV series "Red Dwarf"? http://www.youtube.com/watch?v=LRq_SAuQDec)

Capturing that level of understanding in a machine is not trivial - even with today's multi-core, multi-gigahertz applications processors. Recognition of words, and interpretation of the meaning of dialogue is a challenge that is beyond existing embedded processors and probably also the architectures upon which they are based. But the rise of server-based voice activated computing means that the adoption of VUIs is now a realistic solution for many applications.

In search of the perfect toast…

Let's take a quick look behind the scenes of a conversation you might have with a VUI-enabled toaster:

You: "Toaster!"

• Firstly, your toaster must identify a keyword "toaster". This is to ensure that your user interface is only activated when you want it. It sounds trivial, but your VUI must do this reliably, regardless of where you are in a room, and of any other noise services (both organic and non-organic).
• This activation opens a window during which further speech is captured.

Toaster: "Yes"

• This could be a standard response, but we clearly have both a microphone and a speaker in our Toaster. That implies that we need echo cancellation to ensure that we don't create a kitchen full of howling appliances. We also don't want our Toaster's response to be fed back and interpreted as our own speech.

You: "Make me some toast"

• This speech is captured for interpretation as part of an identified sentence. Interpretation consists of two stages; identifying words and identifying meaning. Once we have this, we can formulate a response. In a limited context, like making toast, this may be one of the easiest parts, although ...

Toaster: "No Dave, I recall that you are on a diet and avoiding white carbohydrates"

• Busted! Your toaster also has speaker recognition – it knows who you are, and has interpreted your command in a context. In a more virtuous moment, you have informed your toaster of your weight-loss plans.

This tongue-in-cheek example demonstrates the sort of features that might be offered by a Voice-User-Interface that is genuinely useful. The question remains – how to deliver it economically?

Enter IoT and the cloud

Consider a VUI system that comprises a VUI client (the toaster) and a VUI server, which exists in the cloud. These applications are highly cost sensitive, so a careful partition must be made, to ensure that the client is as 'thin' (low cost) as possible.

Fortunately, the requirements of a 'thin VUI client' are actually quite limited:

• A high quality microphone, capable of far-field, directional sound capture
• Keyword recognition, to ensure that your user-interface is only activated when you want it
• Delineation of sentences
• Echo cancellation, to eliminate unwanted signals in the captured speech
• Secure transmission to and from the cloud
• Audio decoder. Why? Because your synthesized response will probably be delivered as encoded audio – just like your music
• Audio amplification
• Local controls, which will be specific to the appliance

The rest of the functions are delivered by cloud services. As a result, your 'thin' VUI client can now be very low-cost indeed – so low in fact that you might pause to consider whether your toaster really should have one.

Conclusions

The prospect of voice based user interfaces dominating our interactions with machines is now real and imminent. Thus-far, widespread adoption has been hampered by the economics of embedding the required compute performance in very cost sensitive end equipment.

Fast, accurate natural language interpretation in many common languages is already available, but at a price - the cost of the resources required would certainly spoil your breakfast. So what to do? The answer may already be upon us – place the facility in the cloud, where the cost of infrastructure can be shared amongst an internet full of users.

The ability to leverage widespread internet-based learning and intelligence through 'thin VUI clients' presents a solution to this problem, enabling high quality, low cost VUI.

The cloud delivers access to practically limitless compute resources, at the expense of some communications that you may have thought you never needed. "Ha! Why does my dishwasher need to connect to the internet?"– I am sure many of us have said it, probably in response to one campaign or other espousing the Internet of Things. The answer might just be "So that you can talk to it".

DO YOU WANT A TALKIE TOASTER?
TALK TO XMOS AT 139th AES CONVENTION, NEW YORK 2015