arcusmaximus / YTSubConverter

A tool for creating styled YouTube subtitles
MIT License
767 stars 55 forks source link

Videos with furigana/karaoke style subtitles #91

Open jam-madrigal opened 6 months ago

jam-madrigal commented 6 months ago

So, I followed the steps here and downloaded the subtitle files in srv3 format with yt-dlp, then converted them to .ass format with YTSubConverter. The English subs work great, whether i use the --visual option in command line or not, however, I can't seem to get subtitles with furigana to display properly. The furigana shows up like so

image

When It should look like this image

Is there a way to correct this? Either by correcting the placement or even if that means removing the furigana if it comes down to it.

I'm working with this video https://youtu.be/HOmS6Kn0bn8

frezvent commented 5 months ago

Ruby Text can only be displayed after conversion by the program, and will only be displayed as [.../..] in AegiSub.

jam-madrigal commented 5 months ago

Ruby Text can only be displayed after conversion by the program, and will only be displayed as [.../..] in AegiSub.

Alright. Well, do you know of a way to take the existing srv3 file and being able to use it in a media player with the ruby text and karaoke timings similarly displayed as it is on youtube?

frezvent commented 5 months ago

Basically, the format for the ruby text feature is different from the default of Aegisub itself, and the one supported by this program. The program uses special override tags {yt...} including the ruby text tag {ytruby}. If you want to display ruby text in Aegisub that has been converted from srv3, then you must change the entire subtitle line to follow Aegisub's default format. You can read this tutorial.

arcusmaximus commented 5 months ago

Basically, the format for the ruby text feature is different from the default of Aegisub itself, and the one supported by this program. The program uses special override tags {yt...} including the ruby text tag {ytruby}

There's some confusion here. When creating Aegisub subtitles for conversion to SRV3, you indeed use {\ytruby} and the [心/しん] notation. When you perform a regular reverse conversion from SRV3 back to Aegisub, you again get that same notation. However, this ticket is about the visual reverse conversion, where the intention is very much to visually replicate the subtitles as they appear on YouTube, without requiring any manual actions by the user.

In short, this report is perfectly valid, and the only reason furigana isn't handled yet is that it's tricky to do. I'll see if I can work something out.