Skip to content
Transcription

How to translate audio to English or another language

Updated 11 min read

Short answer

To translate audio, transcribe it first, then translate the text. In FilmScribe, upload the audio or video: we transcribe it in 90+ languages and, on Studio and Agency, translate the transcript and captions into 9 languages. You download the translated text and the SRT. The result is text, not a dubbed voice or a live interpreter.

Step by step

  1. 01 Upload the audio, the video or a link An audio or video file, or a YouTube link, with no conversion first. We transcribe it in the language of the recording, detected automatically.
  2. 02 Review the original transcript Fix names, acronyms, numbers and jargon in the original text. The translation starts from the corrected text.
  3. 03 Pick a language and translate On Studio and Agency, choose one of 9 languages. Close the window if you like: we email you when it's done.
  4. 04 Read and download the translated text Switch languages in the selector, read the translated transcript with the same speakers and timestamps, and download it as TXT or PDF.
  5. 05 Download the translated captions Download the translated SRT, with the same timing as the original captions, or the FCPXML for Final Cut.
In this guide
  1. In short
  2. What does it mean to translate audio?
  3. How to translate audio in FilmScribe, step by step
  4. Which route should you take to translate audio?
  5. How do I translate foreign-language audio to English?
  6. Is translating audio to text the same as transcribing?
  7. How do I translate a WhatsApp voice message?
  8. Can I translate the audio of a video or a YouTube link?
  9. Can FilmScribe dub the audio or translate it live?
  10. How do I review translated audio?
  11. What can you do with the translated text and captions?
  12. Can Google Translate translate audio? Other routes
  13. Common mistakes
  14. Final checklist
  15. Frequently asked questions
  16. Sources

Translating audio is almost never a single step. To get from one language to another and trust the result, the speech has to become text first. That’s why the route that works has two stages: transcribe, then translate. Skip the first and you end up with a translation nobody can check.

In FilmScribe, you upload the audio, the video or a YouTube link. We transcribe it in the language it was recorded in and, on Studio and Agency, take the transcript and the captions into 9 languages. That’s how you translate audio to English, or to another language, and walk away with text to read and an SRT to publish.

One thing up front: what you get is translated text. We don’t replace the voice in the recording, the way dubbing does, and we don’t translate a live conversation, the way an interpreter does. If you already have a subtitle file and only need it in another language, the right guide is how to translate subtitles. If the audio is a church service, see how to translate a sermon.

In short

  • Translating audio takes two stages: transcribe in the original language, then translate the text. A translation is never better than the transcript it came from.
  • FilmScribe transcribes 90+ languages and detects the language for you. Translation takes the text and the captions into 9 languages, on Studio and Agency.
  • You download the translated transcript as TXT or PDF and the translated captions as an SRT, with the same timing as the original.
  • The result is text. We don’t do dubbing or live translation.
  • Review the original before you translate: a wrong name in the transcript becomes a wrong name in every language.

What does it mean to translate audio?

One search hides several very different jobs. Before you pick a tool, work out which one is yours.

What you wantWhat it’s calledDoes FilmScribe do it?
Read in English what was said in another languageTranslated transcriptYes, on Studio and Agency
Publish the video with captions in another languageTranslated captions, as SRT or burned inYes, on Studio and Agency
Turn the audio into text, in the same languageTranscriptionYes, on every plan
Hear the recording in another voice and languageDubbingNo
Talk on the spot with someone who speaks another languageInterpreting, or live voice translationNo

The first three rows start the same way: with the text of what was said. That’s why transcription comes before everything else. With every word written down and timed, translation stops being a guess about sound and becomes text you can read, fix and check.

The last two are a different kind of work. They come up further down, along with the routes that exist for each.

How to translate audio in FilmScribe, step by step

The flow is the same for an interview, a lecture, a recorded meeting or a podcast episode.

Audio as MP3, WAV, M4A, AAC, OGG, OPUS or WEBM. Video as MP4, MOV, AVI, MKV or WEBM. Or a YouTube link. The file goes in as it is, with no conversion first, up to 10 GB. If only part of it matters, mark the start and end before processing: only that range is transcribed and counts toward your plan’s hours.

2. We transcribe in the original language

You don’t have to tell us what language the recording is in: we detect it, across 90+ languages. The transcript comes back word by word, with every word timed and each speaker separated. Everything after this is built on it.

3. Review the original before you translate

Read the transcript and listen to the hard parts: clicking a line’s timestamp takes the player to that point. Fix names of people and companies, acronyms, numbers and jargon right in the text. The fix carries into the captions and the exports, and the translation starts from the corrected text. Ten minutes here saves you the same fix in every language.

4. Pick a language and translate

In the language selector, choose one of the 9 target languages: English, Spanish, French, German, Italian, Portuguese (Brazil), Japanese, Korean and Chinese (Simplified). Translate the captions, the content or both. Content means the full transcript, the summary, the chapter titles and the hooks. Translation runs in the background: close the window if you like, and we’ll email you when it’s ready. Translating doesn’t use your plan’s hours, and one file can hold several languages.

5. Read and download the translated transcript

Switch languages in the selector and read the whole thing translated, with the same lines, the same speakers and the same timestamps. The names you gave each person stay in place, so a translated interview still reads like a conversation. To take the text elsewhere, download it as TXT or PDF, with or without speaker names and the time of each line.

6. Download the translated captions as an SRT

Translated captions follow the original timing: every block comes in and goes out at the same point. Download them as an SRT, with TV, Netflix and YouTube presets and your project’s frame rate, or as FCPXML for Final Cut. With video, translated captions can also be burned into a clip, on their own or next to the original as bilingual captions.

An example: a podcast team records a 50-minute interview with a guest who speaks Spanish. The audio goes into FilmScribe and the transcript comes back in Spanish, with host and guest separated. The team fixes the name of the guest’s company and two acronyms, translates to English and downloads the PDF to write the show notes. The English SRT goes on the video version on YouTube.

Translation is on Studio and Agency, and in their trial: 3 days with 1 hour of media, with a card. In the trial, each file gets 2 translations, counting captions and content together. Pro transcribes in 90+ languages but doesn’t translate. See pricing and more on Translation, Transcription and Languages.

Which route should you take to translate audio?

SituationRouteWatch out for
A recorded interview, lecture, meeting or podcastFilmScribe: transcribe and translate the textReview the transcript before translating
A video you’ll publish in another languageFilmScribe: translated captions, SRT or in a clipRead the captions with the video playing
You only want the audio as text, same languageFilmScribe: the transcript aloneNo translation needed, any plan works
You already have a finished subtitle fileTranslate the SRT, block by blockDon’t touch block numbers or timestamps
A short phrase or a conversation on the spotA voice translation app on your phoneNo file with speakers and timestamps
The audience needs to listen, not readDubbing, done outside FilmScribeAnother voice, another cost, another review

How do I translate foreign-language audio to English?

Whether it’s a voice memo, an interview or a full lecture, the order never changes: upload the audio in its original language, let the transcript finish in that language, review it, and only then translate to English. What changes is the care each source language asks for.

  • Idioms. Spanish “me da igual” isn’t “it gives me equal”: it’s “I don’t mind”. Automatic translation tends to get the common ones and trip on the rare ones.
  • False friends. Spanish “actualmente” means “currently”, not “actually”. Portuguese “pretender” means “to intend”, not “to pretend”. Read the whole sentence, not the word.
  • Numbers, dates and units. “03/04” is April 3 in most of the world and March 4 in the US. Kilometers, kilos and degrees Celsius need converting when the number matters.
  • Names. People, companies and products keep their names. Check that none of them turned into an ordinary word.

Here’s the difference between translating word for word and translating the meaning:

TimeOriginal (Spanish)Word for wordThe meaning
00:02:14”Sacamos la primera versión en seis semanas.""We took out the first version in six weeks.""We shipped the first version in six weeks.”
00:02:19”La verdad, estaba todo prendido con alfileres.""The truth, it was all fastened with pins.""Honestly, it was held together with duct tape.”

Automatic translation is a solid first draft, and the review is what makes it ready to publish. The other direction works the same way: a recording in English becomes text and captions in Spanish or Portuguese for an audience abroad.

What if the language you need isn’t one of the 9? You still get the transcript in the original language, as TXT or PDF, and can take the text to the translator of your choice. We won’t promise what we don’t do: translated captions with matching timing exist only for the languages on the list.

Is translating audio to text the same as transcribing?

Most of the time, yes. Plenty of people search for “translate audio to text” when what they want is the audio written out in the same language. That’s transcription, not translation.

The difference matters when you choose a plan. Transcription is on every FilmScribe plan, with timestamps, speakers separated and TXT, PDF and SRT downloads. Translation, which takes that text into another language, is on Studio and Agency.

If text is all you need, the walkthrough is in how to transcribe audio to text, and the page to start from is free audio to text. To go straight to a subtitle file, see MP3 to SRT.

How do I translate a WhatsApp voice message?

A voice message in another language is a common case: the supplier who answers in Portuguese, the client who sends audio in Spanish, the relative who lives abroad.

  1. Save the audio as a file. On your phone, press and hold the message and use Share. In WhatsApp Web or the desktop app, open the message menu and download it.
  2. Upload it to FilmScribe without converting. OPUS and OGG, the formats WhatsApp uses, go in as they are.
  3. Translate the text. The transcript comes back in the spoken language. On Studio and Agency, pick a language and read the translation.

If someone sent five messages in a row, merge them into one file with the free audio joiner before you upload. The page built for this case is WhatsApp audio to text, which takes the file and explains translation too. The details for each device are in how to transcribe WhatsApp audio.

Yes. An uploaded video and a YouTube link go through the same flow: a transcript in the original language, then translation of the text and the captions. You don’t need to extract the audio first.

Video adds one more output. When you download a clip, choose original, translated or bilingual captions, with both together, and apply one of 235 animated caption styles. Japanese, Korean and Chinese translate but can’t be burned into the video: for those, use the SRT.

Two limits worth knowing ahead of time. Links work for YouTube only: a video from another platform goes in as a file. And text already burned into the picture isn’t read: we start from the speech. The walkthrough for links is in how to transcribe a YouTube video.

Can FilmScribe dub the audio or translate it live?

No. FilmScribe delivers text: the transcript and the captions in the new language. The voice in the recording stays the original.

Dubbing is a different product: a new audio track, in another voice, in the target language. If you publish on YouTube, the platform has automatic dubbing. According to YouTube Help (checked in November 2026), it’s enabled for eligible creators and generates translated audio tracks in different languages, and the video is labeled as auto-dubbed in the description. The same page lists the limits: you can’t edit automatic dubs, they may contain pronunciation errors and mistranslated proper nouns, idioms and jargon, and a video longer than 120 minutes isn’t eligible.

Live translation isn’t what we do either. We work from the recording: the meeting, class or talk happens, and the file is uploaded afterward. For a conversation on the spot, the route is an interpreter or a voice translation app.

It’s worth asking what you really need. Captions cover a lot of what people expect from dubbing: the audience hears the voice and the emotion of the person who spoke and reads along in their own language. And because the text can be reviewed line by line, a translation mistake shows up before it goes out.

How do I review translated audio?

The translation inherits the transcript, and the transcript inherits the audio. Review in that order.

First, the audio. A recording with one person speaking at a time, close to the microphone and without loud music takes less work. If you haven’t recorded yet, the free online voice recorder gives you a file that’s ready to upload.

Then, the original. This is where review pays off most. Search the text for every proper name and acronym and fix each one once.

Last, the translation. Read the translated text with the audio playing, at least for the parts that matter. These are the spots where automatic translation slips most often:

Where it slipsExampleWhat to do
Proper names”Sra. Flores” becoming “Mrs. Flowers”Fix it in the original and check the translation
Idioms”Estar en las nubes” becoming “be in the clouds”Swap in the matching idiom
Jargon and acronyms”PYME”, “IVA”Decide to translate or keep, and stay consistent
Numbers, dates and units”03/04”, kilometers, kilosCheck against the audio and convert if needed
Noisy parts and people talking over each otherA wrong word already in the transcriptListen to the part and fix the original
Tone and formality”Usted” flattened into a casual “you”Adjust to how your audience speaks

If the error came from the original and the translation was already done, fix the transcript, delete that language’s translation on the Translations page and translate again. Final style tweaks are easy to make in the downloaded file, in your text editor. For legal, medical or contract use, automatic translation is a draft: the final text should be signed off by a professional translator.

What can you do with the translated text and captions?

  • YouTube. In YouTube Studio, open Subtitles, pick the video, click Add language and upload the translated SRT. Each language becomes a track viewers choose in the player.
  • Show notes, minutes and reports. The translated TXT or PDF, with speaker and time, is the starting point for the write-up of an interview or the notes for a podcast.
  • Research and reporting. Keep the original and the translation side by side. Matching timestamps make it easy to go back to the audio and verify a quote. See how to transcribe an interview.
  • Your video editor. Import the translated SRT onto the timeline, or the FCPXML in Final Cut. The format is explained in how to create an SRT file.
  • Reels and TikTok. There’s no reliable place to upload an SRT there. Translated captions get burned into the clip.

Before you publish translated captions, check the reading speed with the characters per second checker: translated text often runs longer than the original.

Can Google Translate translate audio? Other routes

It translates the speech your microphone picks up. Google Translate’s help pages (checked in November 2026) describe two uses. In the first, you tap Speak and say the phrase: if your device has a microphone, you can translate spoken words and phrases. In the second, Transcribe translates someone speaking another language in near real time, a lecture or a speech for example, and lets you save the transcript at the end. Those pages don’t describe uploading a recorded audio file.

For a phrase on a trip or a quick conversation, that’s the shortest route. Other options, each with its place:

  • Your phone’s own translation app. It works the same way: speak into the microphone, read the translation on screen.
  • WhatsApp’s own transcripts. According to the WhatsApp Help Center (checked in November 2026), on Android they are text versions of voice messages, available only in English, Portuguese, Spanish and Russian. The page doesn’t mention translation: you read the message in the language it was spoken in.
  • Transcribe in one place, translate in another. Fine for short texts. You paste the transcript into a text translator and lose the timestamps and the speakers along the way.
  • Video editors. Many generate automatic captions inside a project, and some translate them. That makes sense when the video is already being edited there.
  • A professional translator. For contracts, reports and official documents, it’s the only route. A timestamped transcript gives them a head start.

FilmScribe earns its place when the recording is long, when more than one person is speaking, or when the text has to become a file: a transcript with speakers and times, translation in the same place, and captions that come out already in sync.

Common mistakes

  • Translating without reviewing the original. A transcript error travels into every language.
  • Expecting new audio. Translation gives you text and captions, not a voice in another language.
  • Mixing up transcribing and translating. If the output language is the same as the recording’s, the transcript is all you need.
  • Mixing languages in one file. Each file has one detected language. If a recording switches between two, we don’t promise each part in its own language: split it with the audio cutter and upload one part at a time.
  • Publishing without reading along with the audio. A swapped name or a word-for-word idiom only shows up on a careful read.
  • Treating automatic translation as an official document. It’s a draft for the professional translator.

Final checklist

  • Did you upload the whole recording, or only the part that matters?
  • Names, acronyms and numbers fixed in the original transcript?
  • Is the target language one of the 9 on the list?
  • Did you read the translation with the audio playing, at least for the key parts?
  • Text downloaded as TXT or PDF and captions as SRT, in the right language?
  • Does whoever receives it know it’s a reviewed automatic translation?

Frequently asked questions

How do I translate audio to English?

Upload the audio to FilmScribe as it is. We transcribe it in the language it was spoken in, detected automatically across 90+ languages. Once you've fixed names and terms, pick English in the language selector, on Studio or Agency. You read the translated transcript on screen and download it as TXT, PDF or SRT.

How do I translate a voice recording to English?

Upload the file from your phone's recorder app as it is: M4A, MP3 and WAV go in with no conversion, and a WhatsApp voice message goes in as OPUS or OGG. We transcribe it in the spoken language and, on Studio and Agency, you translate the text into 9 languages, English included. You get translated text, not a new recording.

Is translating audio to text the same as transcribing it?

Most of the time, yes. People who search for an audio to text translator usually want a transcript: the audio written out in the language it was spoken in. FilmScribe does that on every plan. Translation is the next step, taking that text into another language, and it's on Studio and Agency.

Does FilmScribe translate audio into a dubbed voice?

No. We deliver translated text: a transcript to read and captions as an SRT or burned into a clip. The voice in the recording stays the original, and we don't generate an audio track in another language. For videos published on YouTube, the platform has its own automatic dubbing, with the limits listed in its help pages.

Can I translate audio in real time?

Not in FilmScribe. We work from the recording: you upload the file after the conversation, class or meeting has happened, and get the text in the other language. To talk with someone on the spot, use an interpreter or a voice translation app on your phone, which translates phrases spoken into the microphone.

Is there a free audio translator?

FilmScribe has no free plan. The 3-day free trial, with 1 hour of media, asks for a card, and translation is included in the Studio and Agency trial, with 2 translations per file. At no cost, voice translation apps on your phone handle short phrases spoken into the microphone, not a long recording.

Which languages can I translate audio into?

The audio can be in any of the 90+ languages we transcribe. Translation takes the text and captions into 9 languages: English, Spanish, French, German, Italian, Portuguese (Brazil), Japanese, Korean and Simplified Chinese. Outside that list, you can still download the transcript in the original language.

Sources

The official documentation behind the numbers and rules in this guide.

  1. 01Translate by speech(opens in a new tab) · Google Translate Help
  2. 02Transcribe in Google Translate(opens in a new tab) · Google Translate Help
  3. 03Use automatic dubbing(opens in a new tab) · YouTube Help
  4. 04Add subtitles and captions(opens in a new tab) · YouTube Help
  5. 05How to turn voice message transcripts on or off(opens in a new tab) · WhatsApp Help Center

Try it on your own recording. See the result before adding a card.

Upload a file or paste a YouTube link. Transcript, chapters, clips and captions are ready in minutes.