formatting (f4c03b80) · Commits · IVAS Codec Public Collaboration / IVAS Processing Scripts

ivas_processing_scripts/init.py

+8 −3

Original line number	Diff line number	Diff line
		@@ -50,7 +50,7 @@ from ivas_processing_scripts.processing.processing import (
		preprocess_2,
		process_item,
		reorder_items_list,
		reverse_process_2
		reverse_process_2,
		)
		from ivas_processing_scripts.utils import DirManager, apply_func_parallel

		@@ -98,8 +98,13 @@ def main(args):

		# Re-ordering items based on concatenation order
		if hasattr(cfg, "preprocessing_2"):
		if cfg.preprocessing_2.get("concatenate_input") and cfg.preprocessing_2.get("concatenation_order", None) is not None:
		cfg.items_list = reorder_items_list(cfg.items_list, cfg.preprocessing_2["concatenation_order"])
		if (
		cfg.preprocessing_2.get("concatenate_input")
		and cfg.preprocessing_2.get("concatenation_order", None) is not None
		):
		cfg.items_list = reorder_items_list(
		cfg.items_list, cfg.preprocessing_2["concatenation_order"]
		)

		# check for ISM metadata
		if cfg.input["fmt"].startswith("ISM"):

ivas_processing_scripts/audiotools/audiofile.py

+0 −1

Original line number	Diff line number	Diff line
		@@ -431,4 +431,3 @@ def parse_wave_header(
		"ext_param_size": ext_param_size,
		"ext_param": ext_param,
		}

ivas_processing_scripts/processing/preprocessing_2.py

+45 −13

Original line number	Diff line number	Diff line
		@@ -32,16 +32,23 @@

		import logging
		from pathlib import Path
		import numpy as np
		from warnings import warn

		from ivas_processing_scripts.processing.processing import Processing
		from ivas_processing_scripts.audiotools.audiofile import read, write
		from ivas_processing_scripts.audiotools.audioarray import trim, delay
		import numpy as np

		from ivas_processing_scripts.audiotools import audio
		from ivas_processing_scripts.audiotools.metadata import add_remove_preamble, write_ISM_metadata_in_file
		from ivas_processing_scripts.audiotools.wrappers.bs1770 import get_loudness, loudness_norm
		from ivas_processing_scripts.audiotools.audioarray import delay, trim
		from ivas_processing_scripts.audiotools.audiofile import read, write
		from ivas_processing_scripts.audiotools.metadata import (
		add_remove_preamble,
		write_ISM_metadata_in_file,
		)
		from ivas_processing_scripts.audiotools.wrappers.bs1770 import (
		get_loudness,
		loudness_norm,
		)
		from ivas_processing_scripts.audiotools.wrappers.random_seed import random_seed
		from ivas_processing_scripts.processing.processing import Processing


		class Preprocessing2(Processing):
		@@ -54,7 +61,9 @@ class Preprocessing2(Processing):
		logger.debug(f"Preprocessing2 {in_file.absolute()} -> {out_file.absolute()}")

		# load in file
		audio_object = audio.fromfile(self.in_fmt, in_file, fs=self.in_fs, in_meta=in_meta)
		audio_object = audio.fromfile(
		self.in_fmt, in_file, fs=self.in_fs, in_meta=in_meta
		)

		# add preamble
		if self.preamble:
		@@ -75,7 +84,12 @@ class Preprocessing2(Processing):
		audio_object.obect_pos = metadata

		# add preamble to actual signal
		audio_object.audio = trim(audio_object.audio, audio_object.fs, (-self.preamble, 0), self.pad_noise_preamble)
		audio_object.audio = trim(
		audio_object.audio,
		audio_object.fs,
		(-self.preamble, 0),
		self.pad_noise_preamble,
		)

		# add background noise
		if self.background_noise:
		@@ -91,17 +105,28 @@ class Preprocessing2(Processing):
		range_delay = (1, 2400000)

		# load background noise
		noise_object = audio.fromfile(self.in_fmt, self.background_noise["background_noise_path"], fs=self.in_fs, in_meta=in_meta)
		noise_object = audio.fromfile(
		self.in_fmt,
		self.background_noise["background_noise_path"],
		fs=self.in_fs,
		in_meta=in_meta,
		)

		# if noise is too short raise error
		if len(noise_object.audio) < len(audio_object.audio):
		raise ValueError("Background noise too short for audio signal")
		if len(noise_object.audio) - range_delay[1] < len(audio_object.audio):
		warn("Background noise may be to short for audio signal when considering the random delay")
		warn(
		"Background noise may be to short for audio signal when considering the random delay"
		)

		# measure loudness of audio signal based on output format
		tmp_object = audio.fromtype(self.out_fmt)
		if isinstance(tmp_object, audio.ObjectBasedAudio) or isinstance(tmp_object, audio.SceneBasedAudio) or isinstance(tmp_object, audio.MetadataAssistedSpatialAudio):
		if (
		isinstance(tmp_object, audio.ObjectBasedAudio)
		or isinstance(tmp_object, audio.SceneBasedAudio)
		or isinstance(tmp_object, audio.MetadataAssistedSpatialAudio)
		):
		out_format = None
		else:
		out_format = self.out_fmt
		@@ -112,8 +137,15 @@ class Preprocessing2(Processing):
		loudness_noise = loudness_signal - self.background_noise["snr"]

		# apply random delay and cut signal
		rand_delay = random_seed(range=range_delay, master_seed=self.background_noise["master_seed"], prerun_seed=self.background_noise["seed_delay"], hexa=False)
		noise_object.audio = delay(noise_object.audio, delay=-rand_delay, samples=True, fs=noise_object.fs)[:len(audio_object.audio)]
		rand_delay = random_seed(
		range=range_delay,
		master_seed=self.background_noise["master_seed"],
		prerun_seed=self.background_noise["seed_delay"],
		hexa=False,
		)
		noise_object.audio = delay(
		noise_object.audio, delay=-rand_delay, samples=True, fs=noise_object.fs
		)[: len(audio_object.audio)]

		# scale background noise to desired loudness based on output format
		noise_object.audio = loudness_norm(noise_object, loudness_noise, out_format)

ivas_processing_scripts/processing/processing.py

+34 −11

Original line number	Diff line number	Diff line
		@@ -37,21 +37,28 @@ from pathlib import Path
		from shutil import copyfile
		from typing import Iterable, Union
		from warnings import warn

		import numpy as np

		from ivas_processing_scripts.audiotools import audio
		from ivas_processing_scripts.audiotools.audiofile import concat, split, read, trim, write
		from ivas_processing_scripts.audiotools.audiofile import (
		concat,
		read,
		split,
		trim,
		write,
		)
		from ivas_processing_scripts.audiotools.metadata import (
		add_remove_preamble,
		concat_meta_from_file,
		split_meta_in_file,
		metadata_search,
		add_remove_preamble,
		split_meta_in_file,
		write_ISM_metadata_in_file,
		)
		from ivas_processing_scripts.audiotools.wrappers.bs1770 import scale_files
		from ivas_processing_scripts.constants import LOGGER_DATEFMT, LOGGER_FORMAT
		from ivas_processing_scripts.processing.config import TestConfig
		from ivas_processing_scripts.utils import apply_func_parallel, list_audio, pairwise
		from ivas_processing_scripts.audiotools.wrappers.bs1770 import scale_files


		class Processing(ABC):
		@@ -159,7 +166,11 @@ def concat_teardown(cfg: TestConfig, logger: logging.Logger):
		for odir in cfg.out_dirs:
		path_input = odir / cfg.items_list[0].name
		out_paths = split(
		path_input, odir, cfg.split_names, cfg.splits, in_fs=cfg.postprocessing["fs"]
		path_input,
		odir,
		cfg.split_names,
		cfg.splits,
		in_fs=cfg.postprocessing["fs"],
		)

		logger.debug(
		@@ -177,7 +188,7 @@ def concat_teardown(cfg: TestConfig, logger: logging.Logger):
		cfg.split_names,
		cfg.splits,
		output_format,
		meta_files=cfg.metadata_path[0]
		meta_files=cfg.metadata_path[0],
		)
		out_meta.append(out_meta_paths)

		@@ -289,7 +300,11 @@ def reverse_process_2(cfg, logger):
		if cfg.postprocessing["fmt"].startswith("ISM"):
		out_meta_splits = []
		for i, condition in enumerate(out_paths_splits):
		meta_condition = metadata_search(cfg.out_dirs[i], condition, num_objects=int(cfg.postprocessing["fmt"][-1]))
		meta_condition = metadata_search(
		cfg.out_dirs[i],
		condition,
		num_objects=int(cfg.postprocessing["fmt"][-1]),
		)
		out_meta_splits.append(meta_condition)
		else:
		out_meta_splits = None
		@@ -393,19 +408,27 @@ def remove_preamble(cfg):
		if cfg.postprocessing["fmt"].startswith("ISM"):

		# search for metadata
		meta_item = metadata_search(odir, [Path(item.name)], num_objects=num_channels)
		meta_item = metadata_search(
		odir, [Path(item.name)], num_objects=num_channels
		)
		metadata_array = []
		for meta_i in meta_item:
		metadata_array.append(np.genfromtxt(meta_i, delimiter=","))

		# remove preamble
		metadata_array = add_remove_preamble(metadata_array, cfg.pre2.preamble, add=False)
		metadata_array = add_remove_preamble(
		metadata_array, cfg.pre2.preamble, add=False
		)

		# write csv files
		write_ISM_metadata_in_file(metadata_array, [path_input], automatic_naming=True)
		write_ISM_metadata_in_file(
		metadata_array, [path_input], automatic_naming=True
		)

		# read file
		x, fs = read(path_input, nchannels=num_channels, fs=cfg.postprocessing["fs"])
		x, fs = read(
		path_input, nchannels=num_channels, fs=cfg.postprocessing["fs"]
		)

		# remove preamble
		x = trim(x, fs, (cfg.pre2.preamble, 0))