Accent Correction in Post-Production: When Language Coaching Hits Its Limits

If there's a villain in post, it's not the on-set mic or the air‑con, but the accent that still sounds like the actor flew in from L.A., like yesterday. Accent issues are one of the most common reasons productions schedule ADR, even after actors spend months with dialect coaches.
On The Brutalist (2024), director Brady Corbet has said the goal wasn’t to replace Adrien Brody or Felicity Jones, but to make the Hungarian sound truly native to Hungarian listeners — turning to Respeecher for an ethical, tightly scoped layer of accent correction in post‑production.
This article is about that move: how to fix accent in post‑production when ADR isn’t the right answer, and where AI voice correction steps in after the dialect coach has done everything they can.
Key Takeaways
-
Dialect coaches solve accent work in pre‑production and on set. Anything left over after footage is locked becomes a post-production problem.
-
ADR is brilliant when you need to rebuild whole chunks of dialogue, but it’s a sledgehammer solution when the problem is one rogue vowel in a complicated language like Hungarian.
-
AI voice correction is the quiet tool that fixes accent and pronunciation inside the take everyone loves — same actor with the same emotion, minus fewer winces from native speakers.
What Dialect Coaches Actually Do (and What They Can't Fix)
A dialect coach is the person who takes an actor with a very normal voice and teaches them to sound like they grew up in Budapest, Marseille, or Kyoto. They're a highly technical consultant who lives in the weeds of phonetic accuracy, break every line into sounds, drill stress and melody, and feed the actor a steady diet of reference recordings until the accent stops falling apart under pressure.
Most of that work happens in pre‑production. The coach comes in weeks (or months) before the shoot, marks up the script, runs one‑on‑one sessions, and sometimes stands just off camera during takes to nudge a vowel or consonant till "this passes for native." If the schedule allows, they might also show up for ADR sessions and try to rebuild the performance in a quiet studio.
The real dialect coach film limitations show up after the shoot wraps: they can’t fix the phonetic accuracy of a performance once the scene is already shot and locked. If editorial discovers, three months later, that the strongest emotional take includes one very non‑Hungarian "r," the usual choices are a reshoot or a full ADR call. Both expensive, both dependent on an actor who may already be on another continent.
That's why languages with unforgiving phonetics like Hungarian, Polish, Arabic, Mandarin, and Japanese cause so much "trouble" — native speakers hear every wobble instantly. The actor can do months with a coach, the test screening goes well, and then local audiences latch onto tiny pronunciation slips that nobody on the English‑speaking crew even noticed.
When Traditional ADR Isn’t the Answer
ADR exists to rescue dialogue after the fact: you bring the actor into a studio, re‑record the lines, and drop the new audio into the finished cut so it sounds like it was always there. It's brilliant for fixing bad location sound, rustling mics, or a plane that decided to fly over the one perfect take.
But if an actor can’t reliably hit the right sounds in a booth, you usually get the same imperfect accent you had on set, only cleaner. When the problem is performance, doing another pass of the same performance doesn’t suddenly deliver native‑level phonetic accuracy.
Sometimes studios go much further. In Elysium (2013), for instance, Jodie Foster’s original French‑accented English was swapped for a more neutral American delivery after poor test‑screening feedback. The accent issue went away, but a new one appeared: in the finished film there are shots where the ADR doesn’t quite sit on the lips, so picture and performance feel slightly out of sync.
Then there’s the logistics. You realise during the mix that a line isn’t working, and the actor is halfway through another shoot, or literally on another continent, and nobody wants to blow a day for one line. For many accent correction film problems, the real task is microscopic: sharpen one word, tidy one sound, rather than spin up a full ADR session with studio, engineer, and talent.
Instead of booking new ADR, Respeecher lets you correct the accent and pronunciation inside the take you already love. See how pronunciation correction in post-production works →
How AI Voice Correction Fixes What Coaches Can't
At a technical level, AI voice correction with Respeecher works in three main steps:
- We analyse the actor’s original pronunciation in the scene, down to individual phonemes and timing.
- We build or load a custom model of that specific voice, so the system “knows” how this actor sounds across a range of lines.
- We apply targeted phonetic adjustments to the recorded line, correcting accent and pronunciation while keeping the original performance.
Crucially, this isn’t a voice swap – we’re not dropping in a totally new performance. We’re nudging individual sounds inside the footage you already shot. The character still “speaks” with the same voice audiences know, but the accent and pronunciation slide closer to what a native ear expects — ideal for post-production pronunciation correction and those subtle AI accent correction fixes that only locals notice.
Our synthetic speech artists sit in the same loop as the re‑recording mixer and the native‑speaker coach, listening back to lines in context and nudging them until the fixes stop drawing attention to themselves. In the workflow, speech enhancement in post-production through speech‑to‑speech (STS) becomes another tool on the sound stage that sits somewhere between editing, ADR, and localization.
Case Study: The Brutalist (2024)
The Brutalist is Brady Corbet's drama about Hungarian architect‑émigré László Tóth and his wife Erzsébet building a new life in post‑war America. Adrien Brody and Felicity Jones both went through intensive work with dialect coaches to play Hungarian characters in English and Hungarian — and still, some of the phonetic details in a language often listed among the hardest for English speakers needed help after the shoot.
When the team realised in post that some lines were just shy of native‑level phonetic accuracy, they wanted to fix the accent without dragging the actors back into ADR or tearing up performances that were already working emotionally.
Corbet’s team brought in Respeecher to work directly on Brody’s and Jones’s original takes, training voice models from the material they’d already recorded and then tightening specific words and sounds in post. The grain of the voice, the pacing, the emotional arc all stayed where they were; what shifted was the accent, inching toward what the Hungarian language consultants were happy to sign off as truly native.
The result was more than an internal win: The Brutalist went on to become an awards‑season heavyweight, with over 160 nominations worldwide and major wins including three Academy Awards, BAFTA Best Director and Best Actor, and Golden Globe Best Motion Picture. For many filmmakers, it’s now the go‑to reference for how pronunciation correction in post-production can live inside a standard studio pipeline without compromising performance or ethics.
Dialect Coach vs ADR vs AI Voice Correction
|
Parameter |
Dialect Coach |
Traditional ADR |
AI Voice Correction (Respeecher) |
|
When it’s used |
Pre‑production and on set during the shoot |
After the shoot, if the actor is available for a session |
After the shoot, even when the actor is busy or unavailable |
|
Can it fix already‑shot material? |
No – only future takes, not locked scenes |
Partly – you can replace lines, but only by re‑recording with talent |
Yes – accent fix without re‑recording inside existing takes |
|
Keeps the original emotion/performance |
Yes |
At risk – depends how well ADR matches the on‑set performance |
Yes – the same take stays, only pronunciation and accent are corrected |
|
Difficult languages (e.g. Hungarian, Arabic) |
Can help, but may not reach native‑level under time pressure |
Can still fall short if the actor can’t reproduce native sounds in the booth |
Designed for precise, native‑level pronunciation correction in post‑production |
|
Actor presence required |
Required – all work is done with the actor present |
Required – you need the actor in the ADR studio |
Not required – models work even if the actor can’t come back |
When to Use Each Approach
The dialect coach vs ADR vs AI voice correction decision usually plays out as a sequence across the whole production.
-
A dialect coach is non‑negotiable in pre‑production for any role with a foreign accent. You bring them in early to shape the character’s sound, drill lines, and get the accent close enough that you’re not relying on post to rescue every scene.
-
Traditional ADR makes sense when you catch the problem early, the actor is available, and you’re talking about full scenes or big chunks of dialogue rather than one rogue word. It’s still the standard fix for bad production sound, missing lines, or performance changes — as long as you can get the talent back in front of a mic.
-
AI voice correction is the tool you reach for when the usual options break down: the show is locked, the actor is unavailable, or the issue is microscopic — a name, a consonant cluster, a handful of vowels. It’s also where accent correction in post‑production shines on languages, where even strong work with a coach and ADR may not quite hit native‑speaker level.
In reality, how to fix accent in post‑production isn't an either/or question. You let the dialect coach shape the performance, you fall back on ADR when you have to rebuild bigger chunks, and you keep AI accent correction film work for the surgical fixes in post — the words and sounds that aren’t worth a full day of pickups.
Final Thoughts
By the time the credits roll, it turns out nobody needed a cage match between the dialect coach and the AI. One gets the actor ready to go on day one, the other cleans up whatever slips through once the picture is locked and everyone has gone home.
The Brutalist is the case study everyone now name‑drops at festivals: a foreign‑language accent film where AI voice correction just helped Hungarian sound like Hungarian. No secret replacement, no “AI actors,” but a very patient tech doing pronunciation work that would have been painful to fix with ADR.
If your next project lives in that space — awards‑season dramas, festival titles, or global streaming shows — it’s worth deciding early what your last‑line plan is. Learn more →
FAQ
In practice, it’s a mix of pronunciation correction post‑production and subtle timing tweaks, often using AI voice technology to fix specific words or sounds while keeping the original performance.
When it’s used for AI accent correction, the goal is to leave the performance alone and only edit how sounds are formed. Because the system works inside the existing take, the emotional arc, and voice quality stay the same. What changes is the phonetic layer, which functions more like very targeted speech enhancement post‑production for accents than a full re‑performance.
Languages with very precise vowel systems and stress patterns tend to be the toughest: Hungarian, Polish, Arabic, Mandarin, Japanese, and similar. Native speakers notice even small deviations, so accent correction in post‑production often means iterating closely with language consultants to make sure every change still feels like a real, not overly‑processed, local voice.
AI accent correction adjusts pronunciation and accent while keeping their tone, timing, and emotional delivery. AI dubbing accent workflows are closer to traditional dubbing, where you replace the voice entirely (often in another language) and match it to picture, which is great for localization but a very different tool than post‑production accent correction


.png?width=477&height=264&name=How%20Much%20Does%20ADR%20Cost%20(1).png)

