Subtitles have been on by default in my house for about four years now, and I’ve stopped feeling self-conscious about it. I don’t technically need them, but I got tired of the ritual of turning the volume up for the whispered plot point, then down again before the score kicks in and rattles the patio doors.
According to AP-NORC polling from 2025, a third of American adults always or often watch with captions on, and 55% of caption users say the reason is simply wanting to catch every word. On Aug. 6, Dolby Vision 2 landed on the first shipping TVs, and most of the coverage has been about picture. The fix for the thing people actually complain about is the other half of that announcement, and it has been falling apart all year.
Dolby’s next-generation audio arrived before the picture did
AC-4 has already run a World Cup, and hardly anyone noticed
Most of the attention on Dolby’s next-generation stack has gone to Dolby Vision 2, which is understandable, because picture upgrades demo well in the shop in a way audio upgrades don’t. But when NBCUniversal announced at CES 2026 that Peacock would become the first streamer to adopt Dolby’s full technology suite, the more consequential half of that commitment was the audio codec sitting at the end of the list.
AC-4 isn’t new, mind you. ETSI standardized it in 2014, and it has been the audio system of ATSC 3.0 broadcasting for years. What it hasn’t been, until this June, is something you could actually get from a streaming service. Telemundo’s Spanish-language World Cup coverage on Peacock changed that: all 104 matches, and what Dolby calls the first commercial deployment of AC-4 by a video streamer.
If you watched any of it, you were among the first people in the world to receive streamed television in a codec built specifically to solve the problem people have up until now been solving with the subtitle button. You almost certainly didn’t notice, though, and that isn’t your equipment’s fault.
What AC-4 does that your TV’s Clear Voice mode can’t
Separating dialogue rather than guessing where it is
Every dialogue fix currently available to you works downstream of a decision already made in the mix. Your TV’s Clear Voice preset, for example, is an equalizer making an educated guess: it lifts the mid-range band where human speech lives and hopes nothing else important is sitting there. Frequently something is, from strings to engine noise, and a great deal of modern sound design occupies that same band, so the boost that rescues a line of dialogue also drags a cello up with it and leaves everything sounding thin and hard.
AC-4 doesn’t work that way. Instead, it carries dialogue as a separable element in the bitstream rather than baked into a finished mix, which means the decoder in your TV or soundbar isn’t guessing at all, but receiving speech and everything else as distinguishable things and rendering them at whatever balance you ask for. Dolby lists native dialogue enhancement, intelligent loudness, and personalization as core capabilities of the format, alongside the efficiency gains that we covered in May. The practical translation is a volume slider for the actors, one that leaves the explosions where the mixer put them.
The accessibility implications go further than that, and they’re what’s genuinely persuasive. ETSI’s release on the standard notes that service providers can deliver secondary audio for visually impaired viewers in 5.1 without doubling the bitrate, meaning audio description as a first-class element of the stream rather than a bandwidth-hungry parallel broadcast. Multiple commentary options work the same way, and it’s a better piece of engineering than the problem strictly requires.
The feature is optional, and the first deployment opted out
Dolby’s own wording gives the game away
Dolby’s June 5 announcement of the World Cup deployment says AC-4 “will also provide streamers with the option to further enhance viewing experiences through dialogue enhancement and audio personalization.” ‘Option’ is the important word in that sentence. The codec shipped, and ran a month-long tournament across 104 matches on the largest stage available to it. And it functioned, as far as any viewer could tell, as a more efficient way to deliver Dolby Atmos, which is a real benefit, but not the one anybody was asking for.
Turning on dialogue enhancement is not a firmware toggle, but requires the audio to be authored with speech kept separable, the service’s encoding pipeline configured to carry it, the playback device to decode it, and, most tediously, a user interface that surfaces the control somewhere a person might find it. Four things, four teams, and no regulatory pressure forcing any of them. The reasonable expectation is that a live sports feed built in a hurry under tournament conditions gets none of that on the first attempt. The less reassuring expectation is that it stays optional indefinitely, because it costs money and nobody has to.
Amazon has spent the wait solving it a different way
On-device separation that doesn’t care which app you’re in
It’s tempting to treat AC-4 as the only serious route to fixing this, and I did think that until I had a proper look at what Amazon has been doing. Dialogue Boost launched on Prime Video in 2022 as cloud-processed audio on a short list of Amazon Originals, but the version that runs on newer Fire TV hardware is a different animal: machine-learning separation running on the device itself, which means it works on whatever you’re watching, in whatever app. Amazon’s research blog reports that in listening tests, more than 86% of participants preferred the enhanced audio, and that approval among users with hearing loss was total.
That’s a meaningfully different bet. AC-4 requires the whole chain to cooperate and delivers a cleaner, mixer-sanctioned result. On-device separation requires nothing from the content at all and delivers a result that’s approximate but available now, everywhere. Our own coverage of the Fire TV setting is a fair guide to how well it holds up in your living room.
If Amazon’s approach keeps improving at its current rate, the argument for AC-4’s dialogue features gets narrower – not wrong, but harder to justify commercially, which is the kind of reasoning that keeps a feature forever optional.
What I’d do about it in the meantime
The workarounds are still the answer, and they’re still workarounds
Nothing written about here changes what’s available to you tonight. If your TV has a Clear Voice or Speech preset, it’s still the best and fastest thing to try, and our guides to fixing murky dialogue and the settings worth changing thoroughly cover that ground. If you’re buying hardware rather than fiddling with settings, though, I’d put a center channel above almost anything else on the list – a three-channel soundbar solves more of this than a fourth height driver ever will, as Eli’s piece on improving TV audio sets out.
What I wouldn’t do is buy anything specifically for AC-4 dialogue enhancement, as there is currently nothing to receive.
The odd thing about all of this is the sequencing. Dolby Vision 2 got an IFA unveiling, a CES roadmap, a launch partner, and a firmware rollout with a date on it, for a problem most viewers have never actually articulated. Dialogue enhancement got a clause in a press release, a checkbox nobody ticked, and 104 football matches to not appear in. The fix exists, and it has been standardized for twelve years, streaming to millions of people this summer with the useful part switched off. So, for most of us, the subtitles stayed on.
-
- Subscription with ads
-
$7.99
- Premium Subscription
-
$13.99
- Simultaneous streams
-
3
- # of profiles
-
6
- Originals
-
Yes
