fix(mux): address embedded audio by the file's real mkvmerge track IDs

A TS file can hold its audio first, which put the audio name on the video.
This commit is contained in:
imSp4rky
2026-10-24 12:45:25 -06:00
parent ecf17baffe
commit 1036528fa3
6 changed files with 127 additions and 62 deletions
+6 -1
View File
@@ -1012,6 +1012,11 @@ class Track:
"""The language tag written to the output file."""
return self.language
def embedded_audio_name(self) -> str:
"""Name the audio a service keeps inside this track's file, by its language when the service set none."""
lang = self.data["audio_language"]
return self.data.get("audio_language_name") or language_name(Language.get(lang)) or lang
def get_track_name(self) -> Optional[str]:
"""Get the Track Name, or the name of its language when the service set none."""
return self.name if self.name is not None else language_name(self.mux_language)
@@ -1276,7 +1281,7 @@ class Track:
if hasattr(self, "data") and self.data.get("audio_language"):
audio_lang = self.data["audio_language"]
audio_name = self.data.get("audio_language_name", audio_lang)
audio_name = self.embedded_audio_name()
args.extend(
[
"-metadata:s:a:0",
+32 -17
View File
@@ -1,5 +1,6 @@
from __future__ import annotations
import json
import logging
import os
import re
@@ -28,7 +29,6 @@ from unshackle.core.tracks.track import Track, has_dts_uhd_sample_entry, strip_d
from unshackle.core.tracks.video import Video
from unshackle.core.utilities import (
is_close_match,
language_name,
log_event,
matching_languages,
original_tag,
@@ -47,6 +47,21 @@ def at_resolution(video: Video, res: int) -> bool:
return video.height == res or bool(video.width and int(video.width * 9 / 16) == res)
def file_track_ids(path: Path) -> dict[str, int]:
"""Return the mkvmerge ID of the first track of each type in a file. A TS file can hold its audio first."""
res = subprocess.run(
[str(binaries.MKVToolNix), "-J", str(path)], capture_output=True, text=True, encoding="utf-8", errors="replace"
)
ids: dict[str, int] = {}
for track in json.loads(res.stdout or "{}").get("tracks") or []:
ids.setdefault(track["type"], track["id"])
if not ids:
logging.getLogger("Tracks").debug(
f"mkvmerge found no tracks in {path.name}, so it is addressed as video 0, audio 1"
)
return ids or {"video": 0, "audio": 1}
class Tracks:
"""
Video, Audio, Subtitle, Chapter, and Attachment Track Store.
@@ -766,14 +781,11 @@ class Tracks:
video_track = video_track or self.videos[0]
if video_track.language.is_valid():
lang_code = str(video_track.language)
lang_name = language_name(video_track.language) or lang_code
for video in self.videos:
if video.data.get("audio_language"):
continue
video.needs_repack = True
video.data["audio_language"] = lang_code
video.data["audio_language_name"] = lang_name
if any(at.path and at.path.exists() and has_dts_uhd_sample_entry(at.path) for at in self.audio):
return self._mux_mp4(
@@ -796,7 +808,7 @@ class Tracks:
def sample_sync(start: float) -> list[str]:
shift = round((start - sample_base) * 1000)
return ["--sync", f"0:{shift}"] if config.sample and shift else []
return ["--sync", f"-1:{shift}"] if config.sample and shift else []
if config.sample:
start, end = (seconds - sample_base for seconds in config.sample)
@@ -831,15 +843,17 @@ class Tracks:
else:
is_default = i == 0
ids = file_track_ids(vt.path)
v = ids.get("video", 0)
video_args = [
"--language",
f"0:{vt.mux_language}",
f"{v}:{vt.mux_language}",
"--default-track",
f"0:{is_default}",
f"{v}:{is_default}",
"--original-flag",
f"0:{vt.is_original_lang}",
f"{v}:{vt.is_original_lang}",
"--compression",
"0:none",
f"{v}:none",
*sample_sync(vt.sample_start),
]
@@ -848,9 +862,9 @@ class Tracks:
video_args.extend(
[
"--default-duration",
f"0:{vt.fps}fps" if isinstance(vt.fps, str) else f"0:{vt.fps:.3f}fps",
f"{v}:{vt.fps}fps" if isinstance(vt.fps, str) else f"{v}:{vt.fps:.3f}fps",
"--fix-bitstream-timing-information",
"0:1",
f"{v}:1",
]
)
@@ -858,19 +872,20 @@ class Tracks:
video_args.extend(
[
"--color-transfer-characteristics",
"0:18", # ARIB STD-B67 (HLG)
f"{v}:18", # ARIB STD-B67 (HLG)
]
)
if hasattr(vt, "data") and vt.data.get("audio_language"):
audio_lang = vt.data["audio_language"]
audio_name = vt.data.get("audio_language_name", audio_lang)
a = ids.get("audio")
if vt.data.get("audio_language") and a is not None:
video_args.extend(
[
"--language",
f"1:{audio_lang}",
f"{a}:{vt.data['audio_language']}",
"--track-name",
f"1:{audio_name}",
f"{a}:{vt.embedded_audio_name()}",
"--original-flag",
f"{a}:{vt.is_original_lang}",
]
)