Bluetooth will support hearing aids, sharing, and a better audio codec
theverge.com
theverge.com
I apologise for the sarcasm. The Bluetooth situation hasn't improved over the years because the spec has. In part, yes. The spec is better. However the real difference has been large players open-sourcing parts of their bluetooth stack. However, even now, without a lot trying, you will find bugridden implementations in both hardware and software.
That bluetooth works at all is a miracle.
BLE is better. But it isn't hard to be better.
It's like a dumbed-down MPEG-1 Layer-1 audio, but with only 8 sub-bands. Like layer-1, the filterbank isn't perfect reconstruction, so the encode/decode cycle causes aliasing noise even before any quantization takes place.
(At a previous job I had to make a proprietary extension that improved the stop-band attenuation of the prototype filter, otherwise the speaker companies didn't even want to talk to us, the THD+N looked so bad.)
The only psychoacoustic model is an optional static biasing of which sub-bands gets bits allocated.
On the other hand, you can encode/decode entirely in software on a 24 MHZ arm7tdmi. (arm7 is ARMv3, not ARMv7, confusingly).
There's so many better ways this could have gone.
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.
Using bluetooth MIDI on my iPhone 11 Pro / Roland F-140R combo produces an occasional 1/3 second latency for some key-presses. Adding bluetooth audio on top produces 2 second latency in the music itself.
So close, yet so far.
I've found BLE to be much more forgiving. BUT, the BT SIG is a special group, we'll see how they mess this up.
This is not Bluetooth Classic. Your impressions (and mine) of shitty speakers that lose connection until 2am and randomly start playing something don't apply.
This is BLE now with an Audio stream. It's not the same. BT and BLE share radio, and GAP profile, maybe L2CAP, everything else is entirely different.
I'm not saying BT SIG didn't mess this up in special ways, but no, you don't know anything about this yet.
It would be nice if I could connect my hearing aids to my computer and listen to skype calls, music, youtube, etc. But so far I've only been able to connect to iOS and Android devices. Maybe this change will fix that?
The new audio codec should be of a much higher quality, though it does match what some proprietary systems have already been doing. Not just "quality" improvements, but the tinny aftertones and so on should just about vanish in all circumstances. Normalisation will work better. It'll be easier to pick out voices in a crowd, etc.
BLE should be much lower latency. Potentially anywhere between half and a third. Which should be a huge quality-of-life improvement for those using bluetooth hearing aids.
BLE is _somewhat_ simpler, and has decent adoption throughout the software industry. So connecting to Windows, macOS, Linux should be possible. (The Linux story is complicated by the new audio codec - its nonfree and patent encumbered). If they've done their job right, you'll be able to connect your hearing aids like any other output device, with a quick and simple pair from whatever Bluetooth menu.
However, BLE is also known for more frequent dropouts related to battery usage spikes. Not directly because of the spec. Because the people implementing it have done a worse job thus far. If that remains the case, then it is fairly useless as a hearing aid tool.
If you believe this, I've got a bridge to sell you.
It basically doesn't matter what "bluetooth" the absurd, bloated standard says. What matters is what implementers do, and what they'll do is hack together the cheapest, fastest version that kinda-barely works on cheapo Chinese hardware and test it only with a couple popular Apple devices.