SILK: HD Voice +

SILK.Logo .180px thumb SILK: HD Voice +If you’ve recently made a Skype-to-Skype call followed by a PSTN call, you’ve probably noticed a difference in the voice quality of the call. The Skype-to-Skype call is crisp, sharp and crystal clear with no voice distortion. Call a landline and you introduce distortion, you get the “can you repeat that, please” requests and it’s not due to hearing loss!

Skype has been a pioneer in improving the audio quality of a voice conversation. When first launched in 2003 Skype came with a wideband audio codec that certainly represented a step ahead at the time. However, Skype made a major step forward with the February 2009 launch of its SILK superwideband audio codec. When using a suitably equipped stereo headset, Skype-to-Skype calls are “in-your-head”; the other party sounds like they are in the same room. But even when using a PC’s mic and speakers, you’ll notice a difference.

So what’s behind this?

Normal human speech audio range covers 50 Hz to 14 KHz; transmitting this range over the legacy phone network was never possible due to a key parameter in the network’s design architecture. Since one could usually interpret and understand voice covering 300 Hz to 4KHz, the legacy PSTN is limited to covering this range; this limitation allowed the phone networks to carry many more conversations over a copper wire.

Wideband audio (often called “HD Voice”) expands this range to cover approximately 50 Hz to 7 KHz, resulting in a clearer audio. Beyond Skype’s initial use of wideband audio, many hardware and softphone vendors now support wideband audio with one proviso – the connection between the two end point devices must support wideband audio along the entire call path. As a result wideband audio requires a connection that totally carries the conversation via Internet packets, imposing a requirement for use of VoIP technology for these calls.

Skype’s SILK technology, however, expands the audio range to cover 20 Hz to 12 KHz resulting in crisper, crystal clear and more productive voice conversations with little impairment due to accents, poor quality microphone/speaker hardware. Other outcomes, as identified on the Wikipedia page, include:

  • Easier to recognize voices, distinguish confusing sounds and understand accented speakers
  • Ease of deciphering words that have the close sounds of ‘s’ and ‘f’ and others, often indistinguishable over telephone lines.
  • Ability to hear faint talkers and to understand double-talk (when more than one person is speaking at the same time)
  • Reduced listening effort (decreased cognitive load), resulting in increased productivity and lessened listener fatigue
  • Better understanding in the face of other impairments, such as when talkers are using a speakerphone or in the presence of background noise

The end result of Skype’s SILK superwideband audio, in addition to much crisper and clearer audio, includes a reduced need to question speech that is associated with various accents and tonal inflections. For customer service operation it means many fewer “Could you repeat that” interruptions and other misunderstandings caused by the narrower bandwidth of familiar telephone hardware. But the final proof is in the actual conversation experience.

SkypeGroupVideo.SILK .TechInfo thumb SILK: HD Voice +From Skype for Windows or Skype for Mac you can check if SILK is being used by going to the Call Technical Information screen during a call and checking the Codec parameter. The sample rate will be twice the maximum audio frequency supported; a 24000 sampling rate supports a 12KHz audio bandwidth.

Skype has invoked its SILK technology not only on Skype for Windows/Mac/Linux but also designed it to work on smartphones – it’s embedded in Skype for iPhone, Skype for iPad and Skype for Android – as well as Skype for TV.

Skype’s SILK superwideband audio technology results in a much clearer and crisper conversation not simply for any Skype-to-Skype call, but even more importantly for any multi-party voice call or Group Video call. It brings “meeting room reality” to these calls; participants feel like the other parties are in the same room. If you use a stereo headset you would say the other party(ies) is(are) almost “in your head”. The conversations become transparent to the audio and video technology, allowing participants to focus on the subject matter and agenda of the call.

Not only does SILK provide “real life” HD voice but it also has the ability to adapt to a range of Internet  connection conditions. Click here for the complete technical specifications and data sheet. In order to encourage adoption beyond Skype, the SILK codec has been made available royalty-free and has been submitted to an international standards organization for adoption as an open standard.

When making SkypeOut calls to the PSTN, Skype must compromise with other audio technology to work with the lower quality voice available on the PSTN and, even worse, on a wireless carrier’s network. This is due to limitations imposed by the audio infrastructure design parameters of the PSTN and mobile networks.

Achieving this bandwidth involves not only the sound engine in the softphone (such as the Skype client) and the sound chip in PC hardware but also ensuring the microphone and speakers can handle that bandwidth. FREETALK’s Everyman USB headset and Everyman Wireless headset are examples of Skype-certified hardware designed to support SILK superwideband audio.

For more see HD Voice: Priceless and Tom Evslin’s description and experience with a call from Vermont to Israel. Listen to SILK at work during a Skype for iPhone to Skype for iPhone call over a 3G wireless carrier connection.

 SILK: HD Voice +
No comments yet.

Leave a Reply