diff --git a/Makefile b/Makefile index 8020cf7..ad276bb 100644 --- a/Makefile +++ b/Makefile @@ -1,14 +1,20 @@ ORIGINAL_SUBS=autogen-subs -VIDEO=1ere-introcut.mp4 +VIDEO=1ere.mkv +# VIDEO=1ere-introcut.mp4 + +test: getvideo + ffmpeg -i $(VIDEO) -i mangen-subs.fr.vtt -vcodec copy -acodec copy -c:s mov_text test.mp4 -y + +getvideo: + yt-dlp https://www.youtube.com/watch?v=WRq2197FlMw -o $(VIDEO) cleansubs: python scripts/parse-subs.py $(ORIGINAL_SUBS).fr.vtt -# ffmpeg -i 1ere.mkv -i output.vtt -vcodec copy -acodec copy -c:s mov_text test.mp4 -y getsubs: yt-dlp --write-auto-subs --sub-lang=fr --skip-download https://www.youtube.com/watch?v=WRq2197FlMw -o autogen-subs # omit the music -clip: - ffmpeg -ss 3:18 -i 1ere-combined.mp4 -vcodec copy -acodec copy $(VIDEO) +# clip: +# ffmpeg -ss 3:18 -i 1ere-combined.mp4 -vcodec copy -acodec copy 1ere-introcut.mp4 diff --git a/README.md b/README.md index c189254..5786d68 100644 --- a/README.md +++ b/README.md @@ -9,14 +9,18 @@ The original video and the auto-generated captions in French can be sourced from The automatically generated French transcript needs to be checked over by a human. The French transcript then needs to be translated to English, also with human oversight. -File `mangen-subs.fr.vtt` contains the current draft of the human-edited auto-generated transcript. +* Verify and edit the file `mangen-subs.fr.vtt` which contains the current draft of the human-edited auto-generated transcript. +* Run `make test` to add subtitles to the video and ensure the format is correct. ### Contributors * tuxmain, [https://txmn.tk/](https://txmn.tk/) +* 1 anonymous contributor(s) ### Getting the original auto-generated transcript +*To see how much a southern accent and speech quirks can throw off an AI.* + Run the following to download the auto-generated transcript from YouTube to a new file named `autogen-subs.fr.vtt`. yt-dlp --write-auto-subs --sub-lang=fr --skip-download https://www.youtube.com/watch?v=WRq2197FlMw -o autogen-subs