-
Notifications
You must be signed in to change notification settings - Fork 165
Commit
This commit does not belong to any branch on this repository, and may belong to a fork outside of the repository.
- Loading branch information
Showing
1 changed file
with
33 additions
and
7 deletions.
There are no files selected for viewing
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Original file line number | Diff line number | Diff line change |
---|---|---|
@@ -1,13 +1,39 @@ | ||
|
||
## 3Blue1Brown captions | ||
|
||
This repository contains captions for videos on [3blue1brown](https://www.youtube.com/3blue1brown). | ||
The files of the form year/video-id/language/sentence_translations.json are used to automatically create subtitles, and the goal is to use them to create automatic dubbing as well. | ||
As a proof of concept for the aim here, on [this video](https://youtu.be/cy8r7WSuT1I), click the gear icon, and change the audio track to "Korean", or take a look a [this one](https://youtu.be/YtkIWDE36qU), and try the audio tracks in French, Greman, Hindi or Spanish. | ||
This repository contains translations for videos on [3blue1brown](https://www.youtube.com/3blue1brown). To review a given translation, or to submit corrections, visit [translate.3blue1brown.com](https://translate.3blue1brown.com/). | ||
|
||
The goal is to use these translations to create subtitles, and for ones which have undergone a full review, to make automatic dubbings. As a proof of concept for autodubbing, take a look at [this video](https://youtu.be/cy8r7WSuT1I), click the gear icon, and change the audio track to "Korean", or take a look a [this one](https://youtu.be/YtkIWDE36qU), and try the audio tracks in French, Greman, Hindi or Spanish. | ||
|
||
Most of this repo was formed with a script that first used Whisper to transcribe video narration into English with punctuation and then used a translation API to translate that script sentence-by-sentence, recording the timings. | ||
Machine translation is often far from perfect, so I'll only feel comfortable with the final results here if those translations have been looked over by a native speaker. | ||
This repo was initially formed with a script that first used Whisper to transcribe video narration into English, recording the time stamps of each word, and then a first pass for translations was made using either Google translate or DeepL. Machine translation is often far from perfect, so I'll only feel comfortable with the final results here if those translations have been looked over by a native speaker. | ||
|
||
### How to help | ||
|
||
We're figuring out the best way to incorporate community contributions. | ||
If you'd like to help out, it might be best to hold off until a proper system is in place, and note your interest in this [Google form](https://forms.gle/AR8YQPL1USxhM5989), or to join [this discord](https://discord.gg/MSKzxDgTGE). | ||
- Go to [translate.3blue1brown.com](https://translate.3blue1brown.com/). | ||
- Navigate to a particular video and language you would like to edit. | ||
- For each sentence, you can either edit the translation, or click the thumbs up to indicate that it sounds natural as it is. | ||
- Optionally, if you ensure translated sentences that are marked as too long are abreviated to fit within the time constraint, this will help to ensure an autodubbing can be created smoothly. | ||
- If you see a "legacy translation" below a line, this comes from community contributions back in the days when YouTube allowed people to submit subtitles. The timings often don't align perfectly, but hopefully they offer helpful suggestions. | ||
- When you've finished with the edits you wanted to make, enter your name, and click submit. | ||
|
||
#### What happens after submitting? | ||
|
||
A pull request is created on this repository with your changes that will be reviewed. If you're unfamiliar with GitHub and pull requestes, don't worry! You don't have to do anything more. If you _are_ familiar with pull requests, another very way to contribute is to look over the ones here in your language and offer a review or feedback. | ||
|
||
Once your edits are reviewed and approved, a script will be run to turn it into a captions file and upload it to YouTube, giving you attribution in the description for your contribution. | ||
|
||
If a video is considered completely reviewed, separate scripts will be run to create the automatic dubbing, and upload it to YouTube. _Note, we're still iterating on this, so it may take a little time to work out the kinks._ | ||
|
||
A video is considered completely reviewed and ready for autodubbing if: | ||
- Each translated sentence has been either edited, or given a thumbs up | ||
- Each translated sentence fits within the time that the original english sentence occupies. | ||
|
||
|
||
### Principles | ||
- **Bias to what's natural**. Don’t worry about translations being word-for-word reflections of the original. Instead, it’s best to follow the spirit of the original, but ensure the result sounds natural and fluent. If an English idiom or turn of phrase doesn't make sense in a given language, don't hesitate to change things completely. | ||
- **Briefer is better**. Because creating an auto-dubbing requires that translations can be spoken in a fixed time interval, so as to match the original video, the default translation length may often be too long. Feel free to cut out words and phrases that are not essential, or to rephrase it entirely to capture the spirit of the original in a different way. | ||
- **Informal > formal**. For languages with alternate formality in the second person, or similar grammatical constructs, it's better to stay informal. The feeling should be one of a personal and friendly conversation between | ||
|
||
If you run into problems, or want to request the creation of a translation file for a new language on a particular video, feel free to create an [issue on this repository](https://github.com/3b1b/captions/issues). | ||
|
||
To discuss other nuances, and to coordinate with your language's community, you can join [this discord](https://discord.gg/MSKzxDgTGE). |