[smitelli]

I never understood why they encode stuff that way. Most streams that I've looked at devote literally 98-99% of the bits to the video, and the audio stream is just scraps.

The film's sound department worked really hard on that stuff! It doesn't take much -- 500 kbit/sec can sound amazing if they encode it well.

[archon]

I would guess that the number of people listening to this media on anything higher quality than the built-in speakers on a 65-inch TV is minuscule. They’re optimizing for sound on an iPad, not a full surround sound setup.

[TylerE]

Even iPads, especially new the newer ones, actually have fairly competent speakers. Much better than most tvs.

[crazygringo]

The thing is, if you watch using AirPods on your iPad with spatial audio, it's actually far better than the vast majority of people's actual surround sound setups.

I never cared much about surround sound until the AirPods got spatial audio. Now it's like I can't live without it.

[griomnib]

Yup. That’s my guess as well, but on the other hand there is a pretty large industry serving the many people who do have a decent home theater system (and even some where it’s more “theater” than “home”).

As somebody who has a very nice (and expensive) 5.1 set up, I’m legitimately a bit worried about this trend.

[NewJazz]

Then make it opt in

[bobdvb]

A significant part is that there's very poor data about how many people have surround sound systems or systems that can make use of such quality.

Sending it speculatively adds to the cost of delivery, but for a percentage of the audience it pushes their video quality down to the next resolution down. And for a percentage of the audience that'll be a more noticeable impact.

Here's another oddity: there's no great ways to measure audio quality subjectively. It's kind of been done for voice telecommunications but for perceptual codecs and media sound? The tools are terrible. So, quantifying decisions about how much bandwidth to allocate are hard. Most companies still depend on trained individuals ("golden ears") to test audio quality and for independent testing you need A/B testing with a listener panel. For video quality we have accepted tools to measure quality. They're not perfect but comparatively, any time you see an audio quality test tool you'll see a substantial professional audience that will happily dismiss it.

All increases in quality, audio or video, are subject to the law of diminishing returns. In audio the argument in favour of higher quality is far weaker than it is for something like HDR.