interview Fraunhofer & LC3
An off-hand comment about a new Bluetooth codec brought us an invitation to talk with audio engineers at Fraunhofer, the world's leading applied research organisation. We took them up on the offer to talk about their new pair of codecs, while also discovering that they think high-resolution audio, and lossless high-res in particular, is mainly marketing spiel — a waste of the audio consumer's energy and money.
This interview is long, and needs context. The previous piece was an editorial comment about the new version of Bluetooth, Bluetooth LE (for Low Energy). Bluetooth LE will allow some useful things, like sending a signal to each earbud individually (currently Bluetooth goes to one earbud and has to be rebroadcast to the other). You may also be able to play to two or more pairs of headphones at once, and to select an ‘Auracast’ stream from, say, multiple monitors in an airport or a gym.
But as the ‘LE’ indicates, the priority is low energy use — as it always has been for Bluetooth, where the bulk of applications are for things other than audio, where range and power consumption are more important than higher bit rates for audio quality. And sure enough, the maximum bit rate allocated to audio under LE is lower than for existing Bluetooth.
Other things being equal, lower audio bit rates mean lower audio quality; that's hardly rocket science. But with the various announcements came rather chirpy news about a new codec, LC3, which is to be the ‘mandated’ codec inside Bluetooth LE, just as the SBC codec long has been within Bluetooth Classic.
In that earlier article, I made a slightly offhand comment about the new LC3 codec coming from Fraunhofer, “the makers of MP3”, but I also noted that they had a higher-level codec, ‘LC3plus’, which seemed not intended for Bluetooth, partly because it had been granted the Japanese Audio Society's logo for Hi Res Audio Wireless. And regular readers will know that we regularly point out that high-res audio cannot pass through a Bluetooth pipe, because the pipe is just too small for the bit rates required for lossless high-res.
Fraunhofer dropped me a line, very polite, noting that LC3plus was, in fact, now being used for Bluetooth (this had developed since the earlier literature I had read), and indeed Fraunhofer's LC3plus had been partly developed specifically to minimise THD+N for its high-res audio customers. Would I like to chat?
Of course. And a long chat, it turned out, with Alexander Tschekalinskij (Alex), Fraunhofer's Senior Engineer in Audio for Communications, and Alfonso Carrera, owner/CEO of Naradiss Media Technologies in Spain and a Senior Consultant at Fraunhofer. They later turned up in Sydney at a Bluetooth event, so I got to meet them, but this interview was conducted internationally over Team Viewer in Q3 2023.
76 research institutes
Even if you haven't heard of Fraunhofer, I can pretty much guarantee that some of its technologies are in your home, and almost certainly in your phone. Fraunhofer-Gesellschaft, headquartered in Germany, claims to be the world's leading applied research organisation, with 76 institutes and research units around Germany, the largest of which is the Fraunhofer Institute for Integrated Circuits IIS in Erlangen. This is home to the audio and media technology research in which we're interested, including the development of MP3 and AAC.
We talked so long that I originally intended to excerpt only a section or two as an interview, in particular the discussion around what constitutes high-res audio. But Fraunhofer was clearly barely interested in high-res and far more keen to explain how LC3 (but actually LC3plus, it turns out) might be better than I had implied. And that turned out to be interesting too, if with sufficient acroynms and initialisms to require the glossary included here, as well as knocking on the ceiling of my relatively minimal electronic engineering education (B.Sc. Hons. Nottingham 1985, but only a 2.2).
So to cover both ends of the topics required, let's instead go through the conversation in more length, with most of what Alex and Alfonso said, transcribed and edited only for clarity, or to avoid repetition, and to strip out some detail on, for example, hearing-aid profiles and other areas too far outside our hi-fi focus. The presentation part of their talk included more slides than Wet'n’Wild — we've rejigged only a few of the more useful charts here, but detailed literature on LC3plus is available online if you need more.
If you do glaze over on the technical stuff early on, move forward, as it gets interesting. They were nice guys, keen to get the facts out, and happy to help. Here's the conversation.
ALEXANDER TSCHEKALINSKIJ
a new codec
(ALEX): I think it makes sense to give you an intro — how did we develop LC3? How did we develop LC3plus from that? Why did we develop LC3plus? Where is it standardised and where is it used, etc? And then we can also answer the questions that you sent that were mostly related, I guess, to the bit rate and to the high-res logo from the Japanese Audio Society. These were the main things?
JEZ: Yes that's sort of where we came from, but you never know where it'll go. So go ahead.
ALEX: Exactly. So, LC3 for Bluetooth — we started our work, I think, sometime in 2016. This was when the Bluetooth SIG* started looking for a new audio codec. So the Bluetooth Hearing Aid Group at that time made an official call that they are looking for a new audio codec. Fraunhofer and Ericsson replied to this call and said OK, we are going to develop a new Bluetooth codec for you. So basically this is how LC3 development started.
JEZ: OK I've got a question already — so the Bluetooth SIG wants a new Bluetooth codec. What — there's nothing suitable out there? They couldn't have adopted LDAC, which has been made open source anyway? Why do they need a new one?
Well they were looking for a codec that is basically royalty-free, right? Because they do not want their members to pay royalties for a codec. They were looking for a codec that fits — I mean, at that time they were looking for a codec that fits mostly the hearing aid use cases. I guess if you're referring to LDAC — LDAC is a high-resolution audio codec for music streaming from Sony, and this is, I guess, a bit too much for a hearing aid. Basically they were looking
