Launch
EarDub
Visit
Example Image

EarDub

Hear any video in your language, in sync with the speaker

Visit

EarDub is a Chrome extension for people who would rather listen to a foreign-language YouTube video than read subtitles for the whole thing. It uses the captions the video already has (auto-generated ones work too), joins the short caption fragments back into complete sentences, translates them with context, and plays a neural voice-over in your language that stays in sync with the video.

The original audio keeps playing quietly underneath, so you still hear the speaker's tone and pauses. Free with an account for 30 minutes of dubbing a day; paid plans start at $9.99/month. It runs on youtube.com in desktop Chrome, and a video needs captions to be dubbed.

Example Image
Example Image
Example Image
Example Image

Comments

But we can ask claude as well to do this right?

good application .good to try

custom-img
Product manager at EarDub, a Chrome exte...

Hi Fazier, I'm the product manager at EarDub. EarDub is for the moment you find the perfect YouTube talk or tutorial, but it's in a language you'd have to read subtitles for the whole time. It takes the captions the video already has, merges the short caption fragments into full sentences, translates them with context, and plays a synced AI voice-over in your language, with the original audio kept quietly underneath. Honest limits: it needs captions to work, and it runs on youtube.com in desktop Chrome for now. I'd really like feedback on how the timing feels in your language, especially on fast speakers, and which videos break it.

Keeping the original audio quietly underneath is a smart choice — full dubbing tends to flatten the speaker's tone and timing. Thirty minutes a day free is enough to test it on a lecture or two. Does it work on embedded players like Vimeo or course platforms, or only where captions are exposed in the page?

Very authentic and top notch

Joining caption fragments into full sentences before translating is the right call. Translating fragment by fragment is what makes most auto-dubs sound broken. When auto-captions mishear a word, can I correct the line before the voice reads it?

Joining caption fragments into full sentences before translating is the smart part. Translating 2–3 word caption chunks is where most auto-dubbing sounds wrong. Does it support Urdu as a target language? Many of the gadget reviews our team in Lahore watches are in Chinese or Korean, and an Urdu voice-over would stand out.

Keeping the original audio playing underneath is a good call, that's what makes dubs feel less robotic. Question about Spanish: is the voice-over Latin American or Spain Spanish, and can you choose? For people in Mexico the accent makes a big difference in whether it sounds natural.

Keeping the original audio quietly underneath is a nice touch. Tone and pauses carry a lot of meaning, especially in Korean videos. Does it handle auto-generated Korean captions with no punctuation well when it joins fragments into sentences?

custom-img
Indie maker building Keyloro, a free way...

Joining caption fragments back into full sentences before translating is the right call — auto captions split mid-clause and that's where most dubbing tools sound robotic. How do you handle timing when the translated sentence runs longer than the original, e.g. English to German? Do you speed up the voice or let it drift and resync at the next sentence?

Using the existing captions instead of transcribing from scratch is smart, that is where most dubbing tools fall over on names and fast speech. 30 free minutes a day is enough to actually test it properly. How well does it handle two people talking over each other?

Premium Products
View all
Example Image
Awards
View all
Example Image
custom-img
Product manager at EarDub, a C...
Makers
custom-img
Product manager at EarDub, a C...

Comments

But we can ask claude as well to do this right?

good application .good to try

custom-img
Product manager at EarDub, a Chrome exte...

Hi Fazier, I'm the product manager at EarDub. EarDub is for the moment you find the perfect YouTube talk or tutorial, but it's in a language you'd have to read subtitles for the whole time. It takes the captions the video already has, merges the short caption fragments into full sentences, translates them with context, and plays a synced AI voice-over in your language, with the original audio kept quietly underneath. Honest limits: it needs captions to work, and it runs on youtube.com in desktop Chrome for now. I'd really like feedback on how the timing feels in your language, especially on fast speakers, and which videos break it.

Keeping the original audio quietly underneath is a smart choice — full dubbing tends to flatten the speaker's tone and timing. Thirty minutes a day free is enough to test it on a lecture or two. Does it work on embedded players like Vimeo or course platforms, or only where captions are exposed in the page?

Very authentic and top notch

Joining caption fragments into full sentences before translating is the right call. Translating fragment by fragment is what makes most auto-dubs sound broken. When auto-captions mishear a word, can I correct the line before the voice reads it?

Joining caption fragments into full sentences before translating is the smart part. Translating 2–3 word caption chunks is where most auto-dubbing sounds wrong. Does it support Urdu as a target language? Many of the gadget reviews our team in Lahore watches are in Chinese or Korean, and an Urdu voice-over would stand out.

Keeping the original audio playing underneath is a good call, that's what makes dubs feel less robotic. Question about Spanish: is the voice-over Latin American or Spain Spanish, and can you choose? For people in Mexico the accent makes a big difference in whether it sounds natural.

Keeping the original audio quietly underneath is a nice touch. Tone and pauses carry a lot of meaning, especially in Korean videos. Does it handle auto-generated Korean captions with no punctuation well when it joins fragments into sentences?

custom-img
Indie maker building Keyloro, a free way...

Joining caption fragments back into full sentences before translating is the right call — auto captions split mid-clause and that's where most dubbing tools sound robotic. How do you handle timing when the translated sentence runs longer than the original, e.g. English to German? Do you speed up the voice or let it drift and resync at the next sentence?

Using the existing captions instead of transcribing from scratch is smart, that is where most dubbing tools fall over on names and fast speech. 30 free minutes a day is enough to actually test it properly. How well does it handle two people talking over each other?

Premium Products