Jezar Wakefield has been involved with audio technology for over 25 years. He worked on the initial development of Epoc32, the forerunner of Symbian OS. In this interview Jezar talks to Richard Bloor about the past, present, and future of music on Symbian OS.
Richard: For any of our readers who are not familiar with the Symbian Developer Network ambassadors program, could you explain what it is about?
Jezar: I'm somebody who has worked both inside and outside Symbian. As a third-party developer, I missed having access to somebody who has really got their hands dirty in the core OS code. From my perspective, the Symbian Developer Ambassador program is therefore about providing an opportunity for all developers to access people who are familiar with the inner workings of the OS. And while my area of expertise is audio, I have been involved with Symbian OS from its inception: So even if I don't know a specific area well enough to answer a question, usually I will know the right person to ask.
Richard: So how do developers tap into this knowledge?
Jezar: Audio queries from developers who contact SDN++ get passed to me. Part of my role is to review SDN++ message boards and I will answer queries posted there. Symbian Developer Ambassadors also attend industry and Symbian events, where developers are welcome to talk with us.
Richard: So what has brought you to be the SDN++ expert in music and audio technology on Symbian OS?
Jezar: I've been involved in audio technology for 25 years. I started in recording and mixing, working on a number of chart songs in the 1980s. I then worked at Pinewood Studios, before becoming involved in developing commercial karaoke systems. It was while doing this, in 1994, that I met David Wood and he invited me to join Psion. This was just as the first lines of Symbian OS, or Epoc32 as it was known then, were being written.
More recently I've been lecturing in music technology at Kendal Performing Arts College and writing for one of the U.K.'s best-known music magazines. So, when I came back to London it was fortunate that I was a natural fit for the audio team at Symbian.
Richard: What was it like returning to Symbian?
Jezar: It was certainly interesting: To see how the company has changed. The first surprise was just how big the company had become and how formal the engineering processes are now. When we did the original development work at Psion in the 90s the team was small enough so that everything was done in informal discussion groups: An interesting contrast to the full engineering systems Symbian uses today.
So, I now find myself as team leader for the section of the new architecture that deals with the transportation of audio data.
Richard: Could you explain what you mean by transportation of audio data?
Jezar: The new architecture we announced at the last Smartphone Show borrows heavily from the FreeWay (IP networking) architecture, to create an architecture that can be easily configured to route audio according to policies defined by a Symbian OS licensee.
It might be helpful to compare audio on a smartphone with audio on a Hi-Fi. On a Hi-Fi, plugging in headphones disconnects the speakers. On a phone however, that would prevent ringtones from being heard, and so a smartphone's speakers and headphones are always switched on - and it's left up to the software to route audio to the correct destination.
We could even have to deal with situations where an alarm goes off at the same time as an incoming video call is ringing, whilst the user is listening to music and playing a Java game. To handle these complex scenarios you need a sophisticated audio data transport and routing mechanism, which is exactly what Symbian OS provides.
The sub-systems I work with are responsible for responding to audio policies and ensuring that audio data gets processed and transported, according to its priority, to the correct output hardware.
It's also worth noting, that in conjunction with FreeWay, the new audio architecture provides a highly efficient data path from the internet straight out through a smartphone's speakers when streaming audio from the web.
Richard: What impact does this have on third-party developers?
Jezar: Initially it need not have an impact. There is a compatibility layer that sits on top of the new architecture, so that all existing programming interfaces to the audio system will work.
However, for developers who want to delve a little deeper there are new features that enable extremely sophisticated signal processing using third-party plug-in components. These components are wrapped in such a way that they can be controlled with very simple commands. So developers will get very powerful signal processing, for very little coding effort.
Richard: What will this practically mean for developers?
Jezar: Take a developer creating a sophisticated music application, one that mixes synthesised sounds with the playback of pre-recorded material. This developer would be able to gain access to mixing devices on the smartphone, rather than having to do the signal processing in their application. The same might apply to accessing equalisers or effects such as echo or chorus. Developers would simply call on the plug-ins and they would do their work automatically, behind-the-scenes.
Richard: When you talk about mixing devices and these other signal processors, presumably they are going to be dependent on hardware implementations or will they be software features in the OS itself?
Jezar: It could be either, software processors in the OS or a specific device, or hardware accelerated processing. Importantly, the new architecture means the implementation will be transparent to the developer.
Richard: How does a developer know what features are available, surely features like echo won't be come a standard part of the OS?
Jezar: There will be a baseline of features included in the operating system. This is because the new architecture incorporates the OpenMAX standard from the Khronos Group. Various signal processors are delivered as part of the implementation. Developers are then able to query OpenMAX to determine what additional processors are available on specific device models.
Richard: Given that these features come from an open standard, what advantages does Symbian OS offer, given that other platforms will also include the same APIs?
Jezar: It is important for us to be able to offer an open standard. It would be naive to assume that every company with a signal processing technology is going to train staff in Symbian OS, to enable them to create a Symbian OS port. Hence the reason for supporting OpenMAX. However, if a developer is willing to put in a little additional effort and do some Symbian specific work - which could be done by a third party - they can achieve a much tighter integration between their component and our New Multimedia Architecture. In fact they should be able to create an instance of their component and plumb it into the system with just four lines of Symbian specific code - and one of those lines is the deletion when the component is finished with.
Richard: What effect do you see this technology having on mobile audio?
Jezar: We have already seen a trend in members of the public moving from consumer to producers of content, right now this is mainly in images and video. My view is that this technology will be the catalyst for users to move from mobile music consumers to mobile music producers.
There are already a number of applications available for desktop computers and with OpenMAX and our new architecture it will be trivial for companies to port their effects processors to Symbian OS. As a result I expect to see some very interesting and highly creative music applications coming to Symbian smartphones.
