Your browser is out-of-date!

Update your browser to view this website correctly. Update my browser now


Fraunhofer’s new partnerships

In this week's focus on research, Professor Rumsey takes a look at the new developments within the Fraunhofer Institute for Integrated Circuits

Possibly less of a household name than Dolby, but with a list of audio coding achievements as long as your arm, the Fraunhofer Institute for Integrated Circuits (IIS) is one of Europe’s leading R&D institutions. Having an ethos that lies somewhere between university and commercial outfit, the Fraunhofer institutes bring cutting edge research into the practical world. Best known for its pioneering MP3 codecs that arguably started the audio download revolution, Fraunhofer IIS is developing a number of strategic partnerships that will carry its technology to the user. One such, recently announced, is the partnership with Oxfordshire-based plug-in developer, Sonnox. Mastering engineers increasingly have to produce good-sounding audio for low bit-rate streaming and download services. At low bit rates quality can be compromised in various ways, therefore some mastering effort may be needed to get the best results. There are so many different codecs, bit rates and variables in the coding process, that it can be a time-consuming process. Encoding has often had to be carried out in non-real-time until now, making it necessary to wait to hear the results every time one changes something. Now, thanks to Sonnox’s Fraunhofer Pro-Codec, it will be possible to hear the results of coding and decoding in real-time, enabling the rapid auditioning and comparison of multiple codecs. There’s also a graphical display that shows the spectrum of the signals and a way of visualising the audibility of the codec-induced noise. Bit-stream integrity metering enables the preview of potential overloads on the decoding side – a poorly understood phenomenon that can arise when levels are maximised in some cases. All of this is packaged in a plug-in form compatible with the popular DAW platforms on Mac and PC. In a separate venture, Texas Instruments (TI) has teamed up with Fraunhofer so as to make MPEG Surround decoding an integral part of the TI network audio software development kit (nSDK). MPEG Surround bitstreams have a backward-compatible stereo core that can be decoded by legacy systems, together with surround parameters transmitted as side information that enable further channels to be reconstructed by appropriate decoders. The way in which surround information is represented – as interchannel time, level and coherence parameters referred to the stereo downmix – enables it to be streamed at very low bit rates, adding only a little to the data rate of two-channel audio. Credible surround can be delivered at 64 kbit/s, for example. The decoder can also include technology for upmixing two-channel stereo for a surround impression, as well as rendering the surround version binaurally for headphones. TI’s nSDK aims to enable rapid development of audio applications for connected entertainment systems, such as Internet radio and online streaming services, using its DA8x Aureus audio processors. By pre-integrating many of the most widely used online services as abstractions within the API (application programming interface), together with network standards and audio post-processing, developers can implement them without needing to grapple with the intricacies of each service’s unique challenges.