Apparently lossless streaming cannot be supported on the newest bluetooth spec, and will probably require some kind of wired headphones for the near future.
Bluetooth reminds me a lot of the F-35: Designed to do everything, does none of it particularly well, hard to build/maintain, and frankly would have been better off as three or four competing standards that did their "thing" well and nothing else.
Might be time to retire it, do a clean sheet audio-only standard, that is easy enough to implement/re-implement with actual tests for all the functionality.
Hard disagree. Bluetooth has an important disadvantage, in that its quality is bad. But it's also got an important advantage: it works, at all.
Basically anything made from early 00s onwards will be able to connect to a modern pair of Bluetooth headphones. Interop is the selling feature.
I wonder if there's any way for them to increase the bandwidth 10x or so. It would increase the use cases a ton.
I.e. I connect my Windows laptop to my Android phone directly through WiFI, in a very simple manner.
In theory, sure. But in practice? Not in my experience.
Yeah, it can probably connect, but will it stay connected and will it reconnect without repairing later? Often not.
LE is a cleanroom specification developed by Nokia and delivered wholesale to the Bluetooth SIG in 2010. It's much more limited in scope, limited in bandwidth and designed to fill a niche around low-power, low-bandwidth devices instead of F-35'ing with WiFi-type capabilities.
It'll get faster in time - each new revision brings with it a faster PHY without compromising on power consumption. It's up to about 2Mbits/sec now, as of 5.0. It's not quite up to where it needs to be to support streaming high-fidelity lossless audio, but it'll be there shortly. It's the right tool for the job - for once - we just need to let it mature a bit.
[edit] Double-checking how AirPods connect, but I'm almost confident it's analogous to LE Audio spec in 5.1
[1] https://www.bluetooth.com/blog/exploring-bluetooth-5-how-fas...
Maybe, just maybe, LE Audio will not actually suck for audio recording. Maybe.
There‘s a good chance I‘m wrong here, but my knowledge is that LE Audio builds on Bluetooth Core 5.2 and no devices have shipped yet.
The F-35 is not one airplane. It's three. The fact that these airplanes share technology and components is a good thing.
The claim that the F-35 is bad because it's a multi-role aircraft is unfounded. Multi role combat aircraft are the standard. Here's a partial list:
https://en.wikipedia.org/wiki/Multirole_combat_aircraft#Acti...
The claim that the F-35 is inferior to single role aircraft like the F-22 is also unfounded. None of us knows the specifics about the two airplanes but we do know that the F-35 has superior weapon systems, durability, communications, and spatial awareness.
There are planned upgrades to make the F-22 compatible with the max off boresight capability of the AIM-9X, with its future LOAL capabilities, and there are plans to coat it with more durable radar reflective "paint". The F-35 has all of those today.
https://en.wikipedia.org/wiki/Lockheed_Martin_F-22_Raptor#Up...
But the most important proof that the F-35 is a good airplane is that there are so many customers lining up to buy it.
https://en.wikipedia.org/wiki/Lockheed_Martin_F-35_Lightning...
You lost me there in an otherwise fine comment. The reason is at least as much that the big bully in the schoolyard wants you to buy his product and most don't dare not to. In many countries that did buy it there have been complaints from the other bidders on the contract that the contact was made specifically so no one else could win so they couldn't even bid properly (and lots of evidence says they are absolutely correct in stating this). This doesn't happen if the product was as good as you claim. Then it would have just won fairly. Personally I haven't heard of a single fair win but I don't follow it that close.
France makes its own combat airplanes, some of which like the Mirage 2000 and the Rafale are in operation.
Sweden chose to make the Gripen, the UK, Germany, Spain, and Italy chose to make the Eurofighter Typhoon, and Japan, well that's a weaker argument because their native fighter is based on an American jet fighter.
Edit: Having some knowledge about how the US operates in Europe I am not surprised. I have personally seen lots of examples of exercises in Europe where foreign nations participate and behave very well with the host. I have only ever seen the US navy and the US army just show up using radio frequencies they damn well feel like, unlike every other nation ever. Everyone else reserve frequencies beforehand. The US not so much, no matter what they might interfere with. The US is a bully that behaves like it damn well feels like.
I'll skip the whole debate about how good perceptual lossy encodings actually are these days and whether any person could reasonably expect to hear that measured difference, but just wanted to clarify that the sample rate of the headphones' built-in ADC shouldn't be conflated with the lossless codec.
I haven't seen anywhere whether AirPods can do 192kHz AAC over Bluetooth, which in my opinion would be more interesting. I assume not if all of the DSP stuff is done at 48kHz.
We have the tech to solve all these issues. But that wouldn't be as profitable.
It's plain as day why certain design decisions are made, and calling it a tradeoff might not be wrong, but it's clearly too kind to these practices.
Max bandwidth for regular Bluetooth is around 3 Mbit/s, to my knowledge. Bluetooth LE is lower at 2 Mbit/s.
https://www.novelbits.io/bluetooth-5-speed-maximum-throughpu...
In any event 300kbps is more than sufficient for transparent compression of audio and, if the source is lossless, then you shouldn't have recompression artifacts. Doing so in real-time does have costs (low latency compression uses both more CPU time and ends up with higher bitrates than when you allow more latency)
A funny thing I‘d do if I had more time and energy to dedicate to the topic of Bluetooth telephony (HSP and HFP are garbage for arbitrary reasons) would be opening a network and doing reasonable quality telephony over it to demonstrate it‘s possible.
In anecdotal data, I noticed that my Bluetooth headphones had latency so bad they were not usable for gaming (CS:GO, Valorant). Switching the codec to aptX reduced latency noticable but also dramatically reduced quality to somewhere similar to a phone call (flat, noisy).
So that's my assumption.
Playing audio on a Homepod used to have huge latency, we're talking about 2 seconds lag, and it still had the sales. Sure thing, they have improved this with Airplay 2 but nonetheless... I'm not sure how big of the concern latency is for the music streaming (not for the video, calls or games).
The headphones should have gigs of buffer and be able to manage lossless with a great UX to support this.
People are already used to buffer waits for streaming video and this is already a niche use case.
Assuming even mediocre speeds, an average album in ALAC is likely around 350 MB. That’s about 32 minutes at 1.4Mbps.
They could start listening around the 17 minute mark and remain uninterrupted.
I suspect most APM owners would happily load the music to the headphones in advance to perform this playback. This seems so obvious to me, that I suspect it is held back by licensing issues more than cost reasons.
The playing device would need to do a good job with the interface but this is not a big ask—-people have been buffering for media they want since RealAudio.
AirPods are headphones, not a music player. A device that you have to "load music" onto isn't a set of headphones anymore -- and the features you're describing would add significantly to the cost of an already expensive device.