Add support for multiple audio tracks

[Theta-Dev]

What is it?

• Bugfix (user facing)
• Feature (user facing)
• Codebase improvement (dev facing)
• Meta improvement to the project (dev facing)

Description of the changes in your PR

• Adds an audio language selector next to the video quality selector
• Default audio track is chosen by the configured content language. If no track matches, it defaults to English.
• New settings option: Content > Prefer descriptive audio (choose the descriptive audio track by default)

Before/After Screenshots/Screen Record

Fixes the following issue(s)

• Fixes [STOP OPENING DUPLICATES] Support multiple audio tracks #5854

APK testing

• V1 (208887d53) https://github.com/TeamNewPipe/NewPipe/suites/11651391410/artifacts/605129298
• V2 (ed06f559a) https://github.com/TeamNewPipe/NewPipe/suites/11688854704/artifacts/607796221
• V3 https://github.com/TeamNewPipe/NewPipe/suites/11709640891/artifacts/609309922
• V4 (39a5c8b) https://github.com/TeamNewPipe/NewPipe/suites/11887030011/artifacts/621997850

Due diligence

• I read the contribution guidelines.

[AudricV]

Thanks a lot! Here are some changes that I would suggest to do and that you need to do:

• For the audio settings you introduced, I would move the descriptive audio setting to Video and audio.

• For the language fallback you use, I would use the default locale instead of the English one, unless it creates some issues.

• For the audio track label, I am not sure if we should use the label set by the streaming service, as it creates localization issues. Track names provided by YouTube provides whether an audio track is an original or not, but that's only on recent videos, so we can't reliably extract this information on the label (and that's why I didn't added this property on the extractor, as I am not sure whether we should add it).

  Instead, I would use a string with the audio language translated in the app UI language and include the descriptive audio property if it is true with a string format (%1s (%2s) with %1s as the audio language and %2s as descriptive).

You need to fix the following behaviors:

• the background players (external and in-app) and video downloader ignore the settings you introduced;

• the downloader do not list the audio language for audio downloads;

• you should use the default audio language or the fallback language for video downloads and external players at least, or better, if you want to do this in your PR and if other team members agree, add the audio selector below the video one for downloads and create an audio streams dialog for external players: this would give flexibility to users. Someone else could always do the latter part :)

  This audio selector must be only present for demuxed streams and a message should be shown for muxed streams saying something like An audio track should be already present in this stream.
  The introduction of this message also raises the question to show all video streams and not merge video only and muxed video streams in the same list, as for instance, downloading a Mr Beast video in 720p with French audio track would not be possible in this case.

[AudricV]

Bug: the audio selection algorithm doesn't seem to take into account anymore the preferred audio format.

When selecting an audio track of latest Mr Beast's video for an external audio player with the M4A codec, I got the URL of a WEBM format (itag 251) where I should have got the itag 140.

Also, I don't know if that's intended or if you did changes since my comment about that, but no audio language is still shown in the downloader audio list.

[Theta-Dev]

Hi @AudricV,
I am still working on the downloader, so I have not pushed it yet.
I'll look into the codec selection issue.

[Theta-Dev]

Alright, I have everything implemented.

Do you know why the android tests time out?

[Stypox]

Things that should be tested before merging, on videos with various combinations of formats and languages. @Theta-Dev it would be great if you could provide a list of videos that have different features (e.g. no audio stream at all (i.e. only video streams with embedded audio), only original audio, descriptive audio, dubbed audio, all audio types in multiple languages, possibly not only from YouTube but from other platforms, too).

• downloading streams
• external player
• audio track menu both in the play queue activity and in the player

Thanks :-)

[Theta-Dev]

Videos for testing:
no external audio: any Peertube or CCC video
original+descriptive audio: https://www.youtube.com/watch?v=TjxC-evzxdk
original+dub: https://www.youtube.com/watch?v=_8W2LIfl5RE
original+dub+descriptive: https://www.youtube.com/watch?v=Kn56bMZ9OE8
multilanguage (no audio stream types): https://www.youtube.com/watch?v=tVWWp1PqDus
multilanguage (with audio stream types): https://www.youtube.com/watch?v=YLt73w6criQ

[Theta-Dev]

Okay, I applied the changes and updated the PR.

[AudricV]

Thanks a lot!

Please address the following changes and try to avoid in the future that your IDE or yourself does formatting changes which make reviews harder.

[Stypox]

Thank you! I tested everything in multiple ways, as said, and everything worked smoothly. The only small inconsistency I found is that the external audio player dialog does not have a "Open in browser" button and opens the audio directly if there is only one audio (which is a different behavior wrt videos). If you agree with me on changing the external audio player dialog to make it consistent with the video one, please do so ;-)

[sonarqubecloud]

Kudos, SonarCloud Quality Gate passed!   

0 Bugs
0 Vulnerabilities
0 Security Hotspots
2 Code Smells

0.0% Coverage
0.0% Duplication

[DebugKirby]

All the testing apk links are 404'd
How do I get this apk?

[ShareASmile]

As this code has been merged into dev branch. You can Try NewPipe Nightly builds:
https://github.com/TeamNewPipe/NewPipe-nightly/releases

[TobiGr]

I found a regression in the latest nightly which was most likely introduced by this PR: #10180