Skip to content
This repository has been archived by the owner on Jul 4, 2024. It is now read-only.

It downloads some empty subtitles #398

Closed
shaynefortier opened this issue Apr 29, 2024 · 2 comments
Closed

It downloads some empty subtitles #398

shaynefortier opened this issue Apr 29, 2024 · 2 comments
Labels
bug Something isn't working

Comments

@shaynefortier
Copy link

shaynefortier commented Apr 29, 2024

Describe the bug
When you download certain series that have true closed captions like The Wrong Way to Use Healing Magic and Frieren: Beyond Journey's End, it downloads 2 en-US (CC) which one is empty and the other contains subtitles.

To Reproduce
Steps / command to reproduce the behavior:

$ crunchy archive -a ja-JP -a en-US -s en-US -s fr-FR -m audio --include-chapters https://www.crunchyroll.com/series/GG5H5XQX4/frieren-beyond-journeys-end

You open a video player such as VLC and under Subtitle > Sub Track we see two en-US (CC) - [English]

To even make sure that one of those English CC is empty, use the ffprobe command you will get the metadata for the empty subtitle has a duration of 0.

...
Stream #0:8(en-US): Subtitle: ass (forced)
    Metadata:
      title           : en-US (CC)
      ENCODER         : Lavc60.31.102 ssa
      DURATION        : 00:00:00.000000000
...

Expected behavior
There shouldn't be two CC subtitles with one of them empty or both contains

Screenshots
If applicable, add screenshots to help explain your problem.
Showing VLC found two English streams for English dub
Screenshot_20240429_062230
Crunchyroll has two English subtitles/Closed Captions for the dub version
Screenshot_20240429_062343
Open the inspector and we see that it loaded a VTT file containing the closed captions.
Screenshot_20240429_062513

Client (please complete the following information):

  • OS: EndevourOS (Arch Linux based)
  • Version 3.5.2 (74aaed 2024-04-24)

Additional context
I found that Crunchyroll started to add true Closed Captions in English to certain anime series such as Frieren, but the format is not in plain text nor ASS, but in VTT which we can see in the last screenshot. The issue it cause is when I need to seperate all the streams so that I can add a subtitle language that Crunchyroll doesn't provid, I can't merge them back together using ffmpeg, because the files are corrupted with a text or binary codec stream that ffmpeg doesn't support. And I can't get rid of that codec in the mp4 video even if I re-encode it.

I tested on other series like Goblin Slayer Season 1 I downloaded yesterday and there is no issue, but it doesn't have the VTT closed captions. This is why I think the issue I am having comes from that empty subtitle.

The solution may be to download it in the right codec format (VTT) or check the duration that returns 0 and ignore it.

@shaynefortier shaynefortier added the bug Something isn't working label Apr 29, 2024
@bytedream
Copy link
Member

The subtitles were already downloaded as A .ass and .vtt (hence the two CC subtitle streams). But FFmpeg had problems with the subtitles being in different formats, internally converted vtt to ass, but this didn't went well and removed all dialogs. It got fixed in v3.6.0

@shaynefortier
Copy link
Author

Thank you! It worked well.

Sign up for free to subscribe to this conversation on GitHub. Already have an account? Sign in.
Labels
bug Something isn't working
Projects
None yet
Development

No branches or pull requests

2 participants