3 ms·
Giant middle finger to Linux users by using a new proprietary Fraunhofer codec. We have finally gotten support for some of the other proprietary codecs in Linux
by rektide 7y ago
Giant middle finger to Linux users by using a new proprietary Fraunhofer codec. We have finally gotten support for some of the other proprietary codecs in Linux Bluetooth land, then Bluetooth SIG goes & picks a new proprietary codec owned by the very very old guard. Another decade of mp3 licensing fees & stacks of legal paperwork, just to encode or decode audio.
There's so many better ways this could have gone.
- layoutIfNeeded 7y agoWhy didn’t they choose Opus?
- makomk 7y agoBecause there's no juicy licensing money to be made from getting Opus into the Bluetooth standard? Seriously, there's no real technical advantage to this weird proprietary codec as far as I can tell - the performance and computational complexity seem to be pretty much on a par with Opus, even based on benchmarks and hardware that the new codec was designed to do well on.
- ohazi 7y agoI mean, you're right. But on the other hand, this is ten years overdue, so I'll take it. Do we even have a real licensing story yet? I can see this going one of two ways... the ARM way (chipmaker pays the license fee) or the video codec way (system integrator / set-top box manufacturer / media player software vendor pays the fee). The ARM way has worked really well for ARM -- they only have to deal with a small handful of big chip companies, and there's no real way for any of them to get out of paying the fee, even in China. This being Fraunhofer, the codec way seems more likely, but it's not a particularly good business model. There are more companies for them to interact with, more chances for many of them to not pay the fee (none of the Chinese system integrators are going to pay, end users definitely aren't going to pay, etc.). Obviously everything would be easier if the BLE 5.2 chipmaker paid the fee, and then software can do whatever it wants, but I guess this is unlikely. Oh well. There'll be a usable implementation in `non-free` that everyone will use, Fraunhofer won't get their licensing fees, and it'll be too difficult / impractical for them to go after anyone.