With the launch of Skype for Windows 4.0, Skype delivered its first client incorporating Skype’s new superwideband SILK codec providing a new user experience in voice calling. For example, Alec Saunders in “Skype 4.0 audio: smooth as SILK”:
Perhaps the biggest improvement, though, is audio quality. We all thought that Skype audio was great, right? Skype’s internally developed SILK codec slipped into the final release, despite not having been in prior betas. A wideband codec that delivers the goods at half the bitrate of prior codecs, SILK makes talking on Skype a pure pleasure.
The SILK codec will be rolled out with the final release of Skype for Mac 2.8 (currently in beta but without the SILK codec) in April and in the next Linux release (no date given). The coding was designed to allow easy embedding into hardware devices, such as sound cards, headphones and Skype phones. But this still limits the SILK experience to Skype-to-Skype calls. So they looked at the question of how can Skype most readily advance the deployment of this high quality codec and the resulting user experience at billions of additional voice conversation end points?
At eComm 2009 this afternoon Skype announced, during Jonathan Christensen’s keynote presentation, that the SILK codec will be licensable, royalty-free, to device manufacturers and others; it has been designed independent from Skype signaling protocols such that it can be used in voice streams incorporating the SIP protocol. Jonathan mentioned that there is no requirement to use the SILK codec in conjunction with a Skype implementation.
From an interview with Jonathan Christensen, Skype’s GM for Audio and Video, with Alec Saunders in Skype’s SILKen bear hug: “Further, SILK media streams pass unchanged across Skype’s own media gateways, meaning that SIP endpoints have suddenly become much more compelling in the Skype world”. Alec goes on to discuss two reasons for Skype’s strategy: (i) to spread superwideband audio everywhere and (ii) to give existing carriers a bear hug. Read the post for details.
In his post on this announcement, Jonathan pointed out what the three year Skype SILK development process focused on:
- improving audio bandwidth going from 8 kHz to 12 kHz, meaning that a SILK conversation sounds like you are in the same room as the person you are speaking with
- providing real-time bandwidth scalability to deal with degraded network conditions
- balancing codec optimization between voice, music and background noise, each of which can have an impact on the overall user experience
- delivering a robust solution that delivers a more consistent audio experience, regardless of network conditions and an individual user’s voice signature
I have watched as a wireless carrier attempts at new video calling services are introduced and subsequently fail due to a lack of adoption – the primary reason being there are too few devices out there supporting video calling and thus, there’s nobody to whom one can make a wireless video call. Skype’s announcement today overcomes a major barrier to allowing a much broader user base to take advantage of superwideband audio. It’s a first, but key, step towards a broader goal than “Skype Everywhere”, namely, superwideband audio on all voice calling.
Finally, this afternoon I had a chance to meet many members of the SILK development team who attended Jonathan’s presentation. They deserve kudos and congratulations for all their ground breaking efforts and success in setting new standards in voice call quality.
Developers can find more information at Skype’s SILK website.
Phil Wolff, Skype Journal comments on the SILK performance slides shown at the presentation: SILK performs better
Update: The transcript of Jonathan Christensen’s Codec Evolution and Industry Proposal presentation.