Skip to content

extractors

Eac3to dataclass

Bases: Extractor

Extracts audio files using eac3to

Parameters:

Name Type Description Default
track int

Relative audio track number

 0
preserve_delay bool

Will preserve existing container delay

 True
output PathLike | None

Custom output. Can be a dir or a file. Do not specify an extension unless you know what you're doing.

 None
append str

Specify a string of args you can pass to Eac3to

 ''
Source code in muxtools/audio/extractors.py 
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
@dataclass
class Eac3to(Extractor):
    """
    Extracts audio files using eac3to

    :param track:               Relative audio track number
    :param preserve_delay:      Will preserve existing container delay
    :param output:              Custom output. Can be a dir or a file.
                                Do not specify an extension unless you know what you're doing.
    :param append:              Specify a string of args you can pass to Eac3to
    """

    track: int = 0
    preserve_delay: bool = True
    output: PathLike | None = None
    append: str = ""

    def extract_audio(self, input: PathLike, quiet: bool = True) -> AudioFile:
        eac3to = get_executable("eac3to")
        input = ensure_path_exists(input, self)
        track = get_absolute_track(input, self.track, TrackType.AUDIO, self)
        form = format_from_track(track)
        if not form:
            error(f"Unrecognized format: {track.format}", self.extract_audio)
            warn("Will extract as wav instead.", self.extract_audio)
            extension = "wav"
        else:
            extension = form.ext

        out = make_output(input, extension, f"extracted_{self.track}", self.output)
        p = run_commandline(f'"{eac3to}" "{input}" {track.track_id+1}: "{out}" {self.append}', quiet)
        if p == 0:
            if not out.exists():
                pattern_str = rf"{re.escape(out.stem)} DELAY.*\.{extension}"
                pattern = re.compile(pattern_str, re.IGNORECASE)
                for f in os.listdir(out.parent):
                    if pattern.match(f):
                        f = Path(out.parent, f)
                        out = f.rename(f.with_stem(out.stem))
                        break

            return AudioFile(
                out, getattr(track, "delay_relative_to_video", 0) if self.preserve_delay else 0, input, duration=duration_from_file(input, self.track)
            )
        else:
            raise error(f"eac3to failed to extract audio track {self.track} from '{input}'", self.extract_audio)

FFMpeg

Bases: HasExtractor, HasTrimmer

Source code in muxtools/audio/extractors.py 
 78
 79
 80
 81
 82
 83
 84
 85
 86
 87
 88
 89
 90
 91
 92
 93
 94
 95
 96
 97
 98
 99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
class FFMpeg(HasExtractor, HasTrimmer):
    @dataclass
    class Extractor(Extractor):
        """
        Extracts audio files using FFMPEG

        :param track:               Relative audio track number
        :param preserve_delay:      Will preserve existing container delay
        :param output:              Custom output. Can be a dir or a file.
                                    Do not specify an extension unless you know what you're doing.
        :param append:              Specify a string of args you can pass to Eac3to
        """

        track: int = 0
        preserve_delay: bool = True
        output: PathLike | None = None

        def extract_audio(self, input: PathLike, quiet: bool = True) -> AudioFile:
            ffmpeg = get_executable("ffmpeg")
            input = ensure_path_exists(input, self)
            track = get_absolute_track(input, self.track, TrackType.AUDIO, self)
            form = format_from_track(track)
            debug(f"Extracting audio track from '{input.stem}' using ffmpeg...", self)
            if not form:
                ainfo = parse_audioinfo(input, self.track, self)
                lossy = getattr(track, "compression_mode", "lossless").lower() == "lossy"
                if lossy:
                    raise error(f"Unrecognized lossy format found: {track.format}", self)
                else:
                    extension = "wav"
                    warn("Unrecognized format: {track.format}\nWill extract as wav instead.", self, 2)
                    out = make_output(input, extension, f"extracted_{self.track}", self.output)
            else:
                ainfo = parse_audioinfo(input, self.track, self, form.ext == "thd")
                lossy = form.lossy
                extension = form.ext
                out = make_output(input, extension, f"extracted_{self.track}", self.output)

            args = [ffmpeg, "-hide_banner", "-i", str(input.resolve()), "-map_chapters", "-1", "-map", f"0:a:{self.track}"]

            specified_depth = getattr(track, "bit_depth", 16)
            if str(specified_depth) not in ainfo.stats.bit_depth and not lossy and not is_fancy_codec(track) and specified_depth is not None:
                actual_depth = int(ainfo.stats.bit_depth) if "/" not in ainfo.stats.bit_depth else int(ainfo.stats.bit_depth.split("/")[0])
                debug(f"Detected fake/padded {specified_depth} bit. Actual depth is {actual_depth} bit.", self)
                if specified_depth - actual_depth > 4:
                    debug("Track will be outputted as flac and truncated to 16 bit instead.", self)
                    out = make_output(input, "flac", f"extracted_{self.track}", self.output)
                    args.extend(["-c:a", "flac", "-sample_fmt", "s16"])
            else:
                if extension == "wav":
                    args.extend(["-c:a", "pcm_s16le" if specified_depth == 16 else "pcm_s24le", "-rf64", "auto"])
                else:
                    args.extend(["-c:a", "copy"])
                    if extension == "dtshd" or extension == "dts":
                        # FFMPEG screams about dtshd not being a known output format but ffmpeg -formats lists it....
                        args.extend(["-f", "dts"])
            args.append(str(out))
            duration = duration_from_file(input, self.track)
            if not run_cmd_pb(args, quiet, ProgressBarConfig("Extracting...", duration)):
                return AudioFile(out, getattr(track, "delay_relative_to_video", 0) if self.preserve_delay else 0, input, ainfo, duration)
            else:
                raise error("Failed to extract audio track using ffmpeg", self)

    @dataclass
    class Trimmer(Trimmer):
        """
        Trims audio files using FFMPEG.
        If you're working with lossless files it is strongly recommended to use SoX instead.

        :param trim:                Can be a single trim or a sequence of trims.
        :param trim_use_ms:         Will use milliseconds instead of frame numbers
        :param fps:                 Fps fraction that will be used for the conversion. Also accepts a timecode (v2) file.
        :param preserve_delay:      Will preserve existing container delay
        :param num_frames:          Total number of frames used for calculations
        :param output:              Custom output. Can be a dir or a file.
                                    Do not specify an extension unless you know what you're doing.
        """

        trim: Trim | list[Trim] | None = None
        preserve_delay: bool = False
        trim_use_ms: bool = False
        fps: Fraction | PathLike = Fraction(24000, 1001)
        num_frames: int = 0
        output: PathLike | None = None

        def _escape_name(self, s: str) -> str:
            """Makes filepaths suitable for ffmpeg concat files"""
            return s.replace("\\", "\\\\").replace("'", "\\'").replace(" ", "\\ ")

        def _calc_delay(self, delay: int = 0, num_samples: int = 0, sample_rate: int = 48000) -> int:
            """
            Calculates the delay needed to fix the remaining sync for lossy audio.
            """
            from math import ceil, floor

            frame = num_samples * 1000 / sample_rate
            leftover = (round(delay / frame) * frame) - delay
            return ceil(leftover) if leftover > 0 else floor(leftover)

        def _targs(self, trim: Trim) -> str:
            """
            Converts trim to ffmpeg seek args.
            """
            arg = ""
            if trim[1] and trim[1] < 0 and not self.trim_use_ms:
                raise error("Negative input is not allowed for ms based trims.", FFMpeg())

            if trim[0] is not None and trim[0] > 0:
                if self.trim_use_ms:
                    arg += f" -ss {format_timedelta(timedelta(milliseconds=trim[0]))}"
                else:
                    arg += f" -ss {format_timedelta(frame_to_timedelta(trim[0], self.fps))}"
            if trim[1] is not None and trim[1] != 0:
                end_frame = self.num_frames + trim[1] if trim[1] < 0 else trim[1]
                if self.trim_use_ms:
                    arg += f" -to {format_timedelta(timedelta(milliseconds=trim[1]))}"
                else:
                    arg += f" -to {format_timedelta(frame_to_timedelta(end_frame, self.fps))}"
            return arg

        def trim_audio(self, input: AudioFile, quiet: bool = True) -> AudioFile:
            if not isinstance(input, AudioFile):
                input = AudioFile.from_file(input, self)
            self.trim = sanitize_trims(self.trim, self.num_frames, not self.trim_use_ms, caller=self)
            minfo = input.get_mediainfo()
            form = format_from_track(minfo)
            lossy = input.is_lossy()
            if not form and lossy:
                raise error(f"Unrecognized lossy format: {minfo.format}", self)

            args = [get_executable("ffmpeg"), "-hide_banner", "-i", str(input.file.resolve()), "-map", "0:a:0"]
            if lossy or is_fancy_codec(minfo):
                args.extend(["-c:a", "copy"])
                extension = form.ext
            else:
                args.extend(["-c:a", "flac", "-compression_level", "0"])
                extension = "flac"

            out = make_output(input.file, extension, "trimmed", self.output)
            ainfo = parse_audioinfo(input.file, caller=self) if not input.info else input.info

            if len(self.trim) == 1:
                debug(f"Trimming '{input.file.stem}' with ffmpeg...", self)
                tr = self.trim[0]
                if lossy:
                    args[2:1] = splitcommand(self._targs(tr))
                else:
                    args.extend(splitcommand(self._targs(tr)))
                args.append(str(out.resolve()))
                if not run_commandline(args, quiet):
                    if tr[0] and lossy:
                        ms = tr[0] if self.trim_use_ms else frame_to_ms(tr[0], self.fps)
                        cont_delay = self._calc_delay(ms, ainfo.num_samples(), getattr(minfo, "sampling_rate", 48000))
                        debug(f"Additional delay of {cont_delay} ms will be applied to fix remaining sync", self)
                        if self.preserve_delay:
                            cont_delay += input.container_delay
                    else:
                        cont_delay = input.container_delay if self.preserve_delay else 0

                    debug("Done", self)
                    return AudioFile(out, cont_delay, input.source)
                else:
                    raise error("Failed to trim audio using FFMPEG!", self)
            else:
                debug(f"Generating trimmed tracks for '{input.file.stem}'...", self)
                concat: list = []
                first = True
                for i, tr in enumerate(self.trim):
                    nArgs = args.copy()
                    if lossy:
                        nArgs[2:1] = splitcommand(self._targs(tr))
                        if first:
                            cont_delay = self._calc_delay(ms, ainfo.num_samples(), getattr(minfo, "sampling_rate", 48000))
                            debug(f"Additional delay of {cont_delay} ms will be applied to fix remaining sync", self)
                            first = False
                    else:
                        nArgs.extend(splitcommand(self._targs(tr)))
                        if first:
                            cont_delay = input.container_delay if self.preserve_delay else 0
                            first = False
                    nout = os.path.join(get_temp_workdir(), f"{input.file.stem}_part{i}.{extension}")
                    nArgs.append(nout)
                    if not run_commandline(nArgs, quiet):
                        concat.append(nout)
                    else:
                        raise error("Failed to trim audio using FFMPEG!", self)
                debug("Concatenating the tracks...", self)
                concat_f = os.path.join(get_temp_workdir(), "concat.txt")
                with open(concat_f, "w") as f:
                    f.writelines([f"file {self._escape_name(c)}\n" for c in concat])

                args[3] = concat_f
                args[2:1] = ["-f", "concat", "-safe", "0"]
                args.append(str(out.resolve()))

                if not run_commandline(args, quiet):
                    debug("Done", self)
                    clean_temp_files()
                    return AudioFile(out, cont_delay, input.source)
                else:
                    clean_temp_files()
                    raise error("Failed to trim audio using FFMPEG!", self)

Extractor dataclass

Bases: Extractor

Extracts audio files using FFMPEG

Parameters:

Name Type Description Default
track int

Relative audio track number

 0
preserve_delay bool

Will preserve existing container delay

 True
output PathLike | None

Custom output. Can be a dir or a file. Do not specify an extension unless you know what you're doing.

 None
append

Specify a string of args you can pass to Eac3to

 required
Source code in muxtools/audio/extractors.py 
 79
 80
 81
 82
 83
 84
 85
 86
 87
 88
 89
 90
 91
 92
 93
 94
 95
 96
 97
 98
 99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
@dataclass
class Extractor(Extractor):
    """
    Extracts audio files using FFMPEG

    :param track:               Relative audio track number
    :param preserve_delay:      Will preserve existing container delay
    :param output:              Custom output. Can be a dir or a file.
                                Do not specify an extension unless you know what you're doing.
    :param append:              Specify a string of args you can pass to Eac3to
    """

    track: int = 0
    preserve_delay: bool = True
    output: PathLike | None = None

    def extract_audio(self, input: PathLike, quiet: bool = True) -> AudioFile:
        ffmpeg = get_executable("ffmpeg")
        input = ensure_path_exists(input, self)
        track = get_absolute_track(input, self.track, TrackType.AUDIO, self)
        form = format_from_track(track)
        debug(f"Extracting audio track from '{input.stem}' using ffmpeg...", self)
        if not form:
            ainfo = parse_audioinfo(input, self.track, self)
            lossy = getattr(track, "compression_mode", "lossless").lower() == "lossy"
            if lossy:
                raise error(f"Unrecognized lossy format found: {track.format}", self)
            else:
                extension = "wav"
                warn("Unrecognized format: {track.format}\nWill extract as wav instead.", self, 2)
                out = make_output(input, extension, f"extracted_{self.track}", self.output)
        else:
            ainfo = parse_audioinfo(input, self.track, self, form.ext == "thd")
            lossy = form.lossy
            extension = form.ext
            out = make_output(input, extension, f"extracted_{self.track}", self.output)

        args = [ffmpeg, "-hide_banner", "-i", str(input.resolve()), "-map_chapters", "-1", "-map", f"0:a:{self.track}"]

        specified_depth = getattr(track, "bit_depth", 16)
        if str(specified_depth) not in ainfo.stats.bit_depth and not lossy and not is_fancy_codec(track) and specified_depth is not None:
            actual_depth = int(ainfo.stats.bit_depth) if "/" not in ainfo.stats.bit_depth else int(ainfo.stats.bit_depth.split("/")[0])
            debug(f"Detected fake/padded {specified_depth} bit. Actual depth is {actual_depth} bit.", self)
            if specified_depth - actual_depth > 4:
                debug("Track will be outputted as flac and truncated to 16 bit instead.", self)
                out = make_output(input, "flac", f"extracted_{self.track}", self.output)
                args.extend(["-c:a", "flac", "-sample_fmt", "s16"])
        else:
            if extension == "wav":
                args.extend(["-c:a", "pcm_s16le" if specified_depth == 16 else "pcm_s24le", "-rf64", "auto"])
            else:
                args.extend(["-c:a", "copy"])
                if extension == "dtshd" or extension == "dts":
                    # FFMPEG screams about dtshd not being a known output format but ffmpeg -formats lists it....
                    args.extend(["-f", "dts"])
        args.append(str(out))
        duration = duration_from_file(input, self.track)
        if not run_cmd_pb(args, quiet, ProgressBarConfig("Extracting...", duration)):
            return AudioFile(out, getattr(track, "delay_relative_to_video", 0) if self.preserve_delay else 0, input, ainfo, duration)
        else:
            raise error("Failed to extract audio track using ffmpeg", self)

Trimmer dataclass

Bases: Trimmer

Trims audio files using FFMPEG. If you're working with lossless files it is strongly recommended to use SoX instead.

Parameters:

Name Type Description Default
trim Trim | list[Trim] | None

Can be a single trim or a sequence of trims.

 None
trim_use_ms bool

Will use milliseconds instead of frame numbers

 False
fps Fraction | PathLike

Fps fraction that will be used for the conversion. Also accepts a timecode (v2) file.

 Fraction(24000, 1001)
preserve_delay bool

Will preserve existing container delay

 False
num_frames int

Total number of frames used for calculations

 0
output PathLike | None

Custom output. Can be a dir or a file. Do not specify an extension unless you know what you're doing.

 None
Source code in muxtools/audio/extractors.py 
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
@dataclass
class Trimmer(Trimmer):
    """
    Trims audio files using FFMPEG.
    If you're working with lossless files it is strongly recommended to use SoX instead.

    :param trim:                Can be a single trim or a sequence of trims.
    :param trim_use_ms:         Will use milliseconds instead of frame numbers
    :param fps:                 Fps fraction that will be used for the conversion. Also accepts a timecode (v2) file.
    :param preserve_delay:      Will preserve existing container delay
    :param num_frames:          Total number of frames used for calculations
    :param output:              Custom output. Can be a dir or a file.
                                Do not specify an extension unless you know what you're doing.
    """

    trim: Trim | list[Trim] | None = None
    preserve_delay: bool = False
    trim_use_ms: bool = False
    fps: Fraction | PathLike = Fraction(24000, 1001)
    num_frames: int = 0
    output: PathLike | None = None

    def _escape_name(self, s: str) -> str:
        """Makes filepaths suitable for ffmpeg concat files"""
        return s.replace("\\", "\\\\").replace("'", "\\'").replace(" ", "\\ ")

    def _calc_delay(self, delay: int = 0, num_samples: int = 0, sample_rate: int = 48000) -> int:
        """
        Calculates the delay needed to fix the remaining sync for lossy audio.
        """
        from math import ceil, floor

        frame = num_samples * 1000 / sample_rate
        leftover = (round(delay / frame) * frame) - delay
        return ceil(leftover) if leftover > 0 else floor(leftover)

    def _targs(self, trim: Trim) -> str:
        """
        Converts trim to ffmpeg seek args.
        """
        arg = ""
        if trim[1] and trim[1] < 0 and not self.trim_use_ms:
            raise error("Negative input is not allowed for ms based trims.", FFMpeg())

        if trim[0] is not None and trim[0] > 0:
            if self.trim_use_ms:
                arg += f" -ss {format_timedelta(timedelta(milliseconds=trim[0]))}"
            else:
                arg += f" -ss {format_timedelta(frame_to_timedelta(trim[0], self.fps))}"
        if trim[1] is not None and trim[1] != 0:
            end_frame = self.num_frames + trim[1] if trim[1] < 0 else trim[1]
            if self.trim_use_ms:
                arg += f" -to {format_timedelta(timedelta(milliseconds=trim[1]))}"
            else:
                arg += f" -to {format_timedelta(frame_to_timedelta(end_frame, self.fps))}"
        return arg

    def trim_audio(self, input: AudioFile, quiet: bool = True) -> AudioFile:
        if not isinstance(input, AudioFile):
            input = AudioFile.from_file(input, self)
        self.trim = sanitize_trims(self.trim, self.num_frames, not self.trim_use_ms, caller=self)
        minfo = input.get_mediainfo()
        form = format_from_track(minfo)
        lossy = input.is_lossy()
        if not form and lossy:
            raise error(f"Unrecognized lossy format: {minfo.format}", self)

        args = [get_executable("ffmpeg"), "-hide_banner", "-i", str(input.file.resolve()), "-map", "0:a:0"]
        if lossy or is_fancy_codec(minfo):
            args.extend(["-c:a", "copy"])
            extension = form.ext
        else:
            args.extend(["-c:a", "flac", "-compression_level", "0"])
            extension = "flac"

        out = make_output(input.file, extension, "trimmed", self.output)
        ainfo = parse_audioinfo(input.file, caller=self) if not input.info else input.info

        if len(self.trim) == 1:
            debug(f"Trimming '{input.file.stem}' with ffmpeg...", self)
            tr = self.trim[0]
            if lossy:
                args[2:1] = splitcommand(self._targs(tr))
            else:
                args.extend(splitcommand(self._targs(tr)))
            args.append(str(out.resolve()))
            if not run_commandline(args, quiet):
                if tr[0] and lossy:
                    ms = tr[0] if self.trim_use_ms else frame_to_ms(tr[0], self.fps)
                    cont_delay = self._calc_delay(ms, ainfo.num_samples(), getattr(minfo, "sampling_rate", 48000))
                    debug(f"Additional delay of {cont_delay} ms will be applied to fix remaining sync", self)
                    if self.preserve_delay:
                        cont_delay += input.container_delay
                else:
                    cont_delay = input.container_delay if self.preserve_delay else 0

                debug("Done", self)
                return AudioFile(out, cont_delay, input.source)
            else:
                raise error("Failed to trim audio using FFMPEG!", self)
        else:
            debug(f"Generating trimmed tracks for '{input.file.stem}'...", self)
            concat: list = []
            first = True
            for i, tr in enumerate(self.trim):
                nArgs = args.copy()
                if lossy:
                    nArgs[2:1] = splitcommand(self._targs(tr))
                    if first:
                        cont_delay = self._calc_delay(ms, ainfo.num_samples(), getattr(minfo, "sampling_rate", 48000))
                        debug(f"Additional delay of {cont_delay} ms will be applied to fix remaining sync", self)
                        first = False
                else:
                    nArgs.extend(splitcommand(self._targs(tr)))
                    if first:
                        cont_delay = input.container_delay if self.preserve_delay else 0
                        first = False
                nout = os.path.join(get_temp_workdir(), f"{input.file.stem}_part{i}.{extension}")
                nArgs.append(nout)
                if not run_commandline(nArgs, quiet):
                    concat.append(nout)
                else:
                    raise error("Failed to trim audio using FFMPEG!", self)
            debug("Concatenating the tracks...", self)
            concat_f = os.path.join(get_temp_workdir(), "concat.txt")
            with open(concat_f, "w") as f:
                f.writelines([f"file {self._escape_name(c)}\n" for c in concat])

            args[3] = concat_f
            args[2:1] = ["-f", "concat", "-safe", "0"]
            args.append(str(out.resolve()))

            if not run_commandline(args, quiet):
                debug("Done", self)
                clean_temp_files()
                return AudioFile(out, cont_delay, input.source)
            else:
                clean_temp_files()
                raise error("Failed to trim audio using FFMPEG!", self)

Sox dataclass

Bases: Trimmer

Trim lossless audio using SoX.

Parameters:

Name Type Description Default
trim Trim | list[Trim] | None

List of Trims or a single Trim, which is a Tuple of two frame numbers or milliseconds

 None
preserve_delay bool

Keeps existing container delay if True

 False
trim_use_ms bool

Will use milliseconds instead of frame numbers

 False
fps Fraction | PathLike

The fps fraction used for the calculations. Also accepts a timecode (v2) file.

 Fraction(24000, 1001)
num_frames int

Total number of frames used for calculations

 0
output PathLike | None

Custom output. Can be a dir or a file. Do not specify an extension unless you know what you're doing.

 None
Source code in muxtools/audio/extractors.py 
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
@dataclass
class Sox(Trimmer):
    """
    Trim lossless audio using SoX.

    :param trim:                List of Trims or a single Trim, which is a Tuple of two frame numbers or milliseconds
    :param preserve_delay:      Keeps existing container delay if True
    :param trim_use_ms:         Will use milliseconds instead of frame numbers
    :param fps:                 The fps fraction used for the calculations. Also accepts a timecode (v2) file.
    :param num_frames:          Total number of frames used for calculations
    :param output:              Custom output. Can be a dir or a file.
                                Do not specify an extension unless you know what you're doing.
    """

    trim: Trim | list[Trim] | None = None
    preserve_delay: bool = False
    trim_use_ms: bool = False
    fps: Fraction | PathLike = Fraction(24000, 1001)
    num_frames: int = 0
    output: PathLike | None = None

    def _conv(self, val: int | None):
        if val is None:
            return None

        if self.trim_use_ms:
            return abs(val) / 1000
        else:
            return frame_to_timedelta(abs(val), self.fps).total_seconds()

    def trim_audio(self, input: AudioFile, quiet: bool = True) -> AudioFile:
        import sox

        if not isinstance(input, AudioFile):
            input = AudioFile.from_file(input, self)
        out = make_output(input.file, "flac", "trimmed", self.output)
        self.trim = sanitize_trims(self.trim, self.num_frames, not self.trim_use_ms, allow_negative_start=True, caller=self)
        source = ensure_valid_in(input, caller=self, supports_pipe=False)

        if len(self.trim) > 1:
            files_to_concat = []
            first = True
            debug(f"Generating trimmed tracks for '{input.file.stem}'...", self)
            for i, t in enumerate(self.trim):
                soxr = sox.Transformer()
                soxr.set_globals(multithread=True, verbosity=0 if quiet else 1)
                if t[0] < 0 and first:
                    soxr.trim(0, self._conv(t[1]))
                    soxr.pad(self._conv(t[0]))
                else:
                    soxr.trim(self._conv(t[0]), self._conv(t[1]))
                first = False
                tout = os.path.join(get_temp_workdir(), f"{input.file.stem}_trimmed_part{i}.wav")
                soxr.build(str(source.file.resolve()), tout)
                files_to_concat.append(tout)

            debug("Concatenating the tracks...", self)
            soxr = sox.Combiner()
            soxr.set_globals(multithread=True, verbosity=0 if quiet else 1)
            formats = ["wav" for file in files_to_concat]
            soxr.set_input_format(file_type=formats)
            soxr.build(files_to_concat, str(out.resolve()), "concatenate")
            debug("Done", self)
        else:
            soxr = sox.Transformer()
            soxr.set_globals(multithread=True, verbosity=0 if quiet else 1)
            debug(f"Applying trim to '{input.file.stem}'", self)
            t = self.trim[0]
            if t[0] < 0:
                soxr.trim(0, self._conv(t[1]))
                soxr.pad(self._conv(t[0]))
            else:
                soxr.trim(self._conv(t[0]), self._conv(t[1]))
            soxr.build(str(source.file), str(out.resolve()))
            debug("Done", self)

        clean_temp_files()
        return AudioFile(out.resolve(), input.container_delay if self.preserve_delay else 0, input.source)