AI video dubbing software
Turn a recording into a finished dub.
Give Tolkr™ a programme and take it back in another language: every sentence on its original, a voice for each speaker, the original under the voice.
Recorded from Tolkr™ Studio. An interview, 
Spanish to 
English.
- 01On its originalevery sentence starts where it was said
- 02A voice eachfor every speaker in the programme
- 03One foldervideo, sound and subtitles, ready to deliver
From file to folder
Six steps. You do the first and the last.
Choose the file and what you want back, then come back to a folder. In between, a card shows each step and the time left.
01
Set up
You choose the file, the languages and the files you want back.
Tolkr™ sets itself up for the workstation.
02
Write it down
You do nothing. A card shows the step and the time left.
Tolkr™ writes down what is said, in its own language.
03
Find the speakers
You do nothing, or set names later.
Tolkr™ tells the speakers apart, gives each a voice, and can suggest names.
04
Translate and voice
You do nothing.
Tolkr™ translates and voices every sentence, each starting with its original.
05
Mix
You do nothing.
Tolkr™ lowers the original under the voice and mixes to the set values.
06
Check and deliver
You listen, then open the folder.
Tolkr™ saves the files you chose.
The card during a dub: each step, and the time left.
From the report of this dub
- Speaker 1a woman, her own voice
- Speaker 2a man, his own voice
Who speaks when, sentence by sentence. Names can be suggested from the conversation, and put on subtitle lines if you choose.
The mix
The original stays in the background, lowered while the voice speaks, the way documentaries and news are voiced over. The picture is copied, never encoded again.
One folder
- VideoThe file with the dubbed sound
- SoundDubbed sound
- SoundVoice only
- SoundOriginal under the voice
- TextSubtitles, translation
- TextSubtitles, original
- TextSentence list
You choose which of these you want, before the dub or after it.
Sync
The sentence sets the time.
A dub that is not tied to the sentences drifts away from the picture. In Tolkr™ each translated sentence is placed where its original is spoken.
- StartWith its original
Each translated sentence starts when its original starts.
- EndBefore the next
It ends before the next sentence begins.
- OriginalUnder the voice
It stays in the background, lowered.
- PictureCopied as it is
It is copied, never encoded again.
Your files
Switch on what your team needs.
These are Tolkr™ Studio's own choices. A finished dub can be saved again in other formats, and nothing is dubbed again.
Sound files as
- VideoThe video with the dub, and the original as its second track
- MP3Dubbed sound
- MP3Voice only
- MP3Original under the voice
- SRTSubtitles of the translation
- SRTSubtitles of the original
- CSVSentence list
Or another folder you choose. If a file is already there, keep both or replace it.
| File | Formats |
|---|---|
| Video | Your video's own picture, with the dub as the first sound track and the original as the second |
| Sound | MP3, WAV, FLAC, M4A or Opus, at 48 or 44.1 kHz (Opus at 48 kHz) |
| Subtitles | SRT or WebVTT, of the translation or the original |
| Sentence list | CSV: each sentence with its times and both texts |
Check
Listen before you deliver.
The finished dub plays inside Tolkr™, with every sentence one click away.
- WaveformSee the whole recording and jump to any sentence.
- Captions that followThe caption of the moment shows while it plays.
- Both languagesEach translated sentence under its original, on the timeline.
- Live meterSee the sound level while it plays.
Speakers
Every speaker, a voice of their own.
Tolkr™ tells the speakers apart. Each person gets a voice, and you decide the rest in the Speakers box.
- Told apartTolkr™ tells the speakers apart.
- A voice eachTolkr™ hears whether it is a man or a woman, and you can change it. Hear a sample of each speaker.
- NamesType a name, or let Tolkr™ suggest one from what is said.
- Dub or keepChoose for each speaker whether to dub them.
- On the subtitlesSubtitle lines can carry the speaker's name.
Speakers, 2 found
- Speaker 1Woman, her own voice, dubbed
- Speaker 2Man, his own voice, dubbed
Long programmes
Long programmes, no restart from zero.
A long programme takes time to dub, so Tolkr™ shows where it is and survives an interruption.
- Time leftOn the card
A card shows each step of the work and the time left.
- RestartNot from zero
Start a stopped dub again and it continues from where it stopped.
- Keeps goingLeave the page
Go on working or leave the page: the progress stays at the top of every page.
Use cases
Who dubs with it.
- Archive backlogDub archive programmes, each sentence starting with its original, on your own systems.
- Training videosDub a training video for each branch, with subtitles in both languages, on your organisation's own systems.
- PodcastsDub a recorded podcast on your own systems. Each speaker gets a voice, with sound files and subtitles to deliver.
- LegalReview foreign language recordings in your language, on your own systems, without sending them anywhere.
Asked first.
Not answered here? Ask the people who build it
How long does a dub take?
It depends on the workstation and the length of the programme. A card shows the time left, and we run it on your own material in a demo.
Can it dub a whole film or interview?
Yes. File mode dubs long recordings, such as an interview of nearly an hour. Start a stopped dub again and it continues where it stopped.
Do I need a graphics card?
It is faster with one, and it also runs on the processor.
Is the original kept?
Yes. It stays under the voice, and the video keeps it as a second track the viewer can switch to.
Can I change the format afterwards?
Yes. Save the files again from the finished dub, without dubbing again.
Which recordings may I dub?
Only files you have the rights to.
Bring a recording.
We run Tolkr™ on a short clip of yours, so you judge it on your own material.
Request a demo