Open kettenfett opened 4 years ago
The extractor already gets thoses tracks as "chapters". I'm not really into that project base, but I guess adding a new FFmpegPostProcessor
for splitting the resulting file into own tracks is a good approach?
I'm playing around with that for a bit
Instead of splitting to separate files, what about mapping the timestamps to chapters?
afaik the timestamps are already parsed into those chapters. At least thats were I got them from.
youtube-dl https://www.youtube.com/watch?v=zPuELAzJyno
No chapters, not fully sure if talking about the same thing.
We were kinda. I was talking about the chapters
raw, internal data, created by the extractors. I guess they are just not included in the local file by default. Maybe there is an extra argument for that
We were kinda. I was talking about the
chapters
raw, internal data, created by the extractors. I guess they are just not included in the local file by default. Maybe there is an extra argument for that
F3 in the main readme for chapter and timestamp did not yield any results.
being able to split a video into multiple files according to these timestamps (with the result the same as if the files had been downloaded from a playlist of those segments) would be great
I want to see the timestamps in VLC / Media Player Classic. How can I embed chapters of the video when I download it with youtube-dl -f bestaudio URL
? I only want the audio and the timestamps with titles embedded inside the file.
I've found crisbal/album-splitter which does exactly this, maybe it can be used as inspiration
For reference, here's another approach, a awk/bash script: https://github.com/Pab0/album_splitter
Syntax:
./album_splitter.sh URL [download_location]
It uses youtube-dl, ffmpeg and awk:
The youtube webpage is fetched (--get-description
) and the video's description is parsed (using 'format_titles.awk') for titles and timestamps/tracks.
Then the album file (e.g. webm) is downloaded (--get-title
), converted to audio (--extract-audio --audio-format mp3
),
and split (using 'split_audio.awk') into single tracks according to the parsed track information.
It would be helpful if the chapters could also be downloaded as a data file, so that the timestamps and the names of the chapters can be archived and used later. Perhaps it could be included in the .info.json
file that's enabled with --write-info-json
?
Please let me know if I should make a separate issue for this.
@femaref your commit looks promising, any updates on this ?
@femaref your commit looks promising, any updates on this ?
not more than is in the fork I'm afraid. I can do a PR for what's there.
yes, please
Yet another rough and simple powershell script: https://gist.github.com/SpL1Ne/54d8c8d036f58b8afb89eaa34ce77fd0
Description
It seems, youtube is now using information found in timecodes in the video description and displays them in the videoplayer like so:
It may have been like this for some time, but I saw it for the first time today (link to video).
Can this be used by youtube-dl to offer a download option, that automatically downloads the individual videos/audios according to these timecodes?
Checklist