fixed some omasa metadata problems (97331089) · Commits · IVAS Codec Public Collaboration / IVAS Processing Scripts

ivas_processing_scripts/audiotools/convert/omasa.py

+16 −16

Original line number	Diff line number	Diff line
		@@ -29,7 +29,8 @@
		# the United Nations Convention on Contracts on the International Sales of Goods.
		#

		import copy
		from copy import copy, deepcopy
		import numpy as np

		from ivas_processing_scripts.audiotools import audio
		from ivas_processing_scripts.audiotools.convert.objectbased import render_oba_to_binaural, render_oba_to_cba, \
		@@ -48,19 +49,20 @@ def convert_omasa(

		# split OMASA object in ISM and MASA object
		oba = audio.fromarray("ISM" + str(omasa.num_ism_channels), omasa.audio[:, :omasa.num_ism_channels], omasa.fs)
		oba.metadata_files = omasa.metadata_files
		oba.object_pos = omasa.object_pos
		masa = audio.fromarray("MASA" + str(omasa.num_channels-omasa.num_ism_channels), omasa.audio[:, omasa.num_ism_channels:], omasa.fs)
		oba.metadata_files = copy(omasa.metadata_files)
		oba.object_pos = copy(omasa.object_pos)
		masa = audio.fromarray("MASA" + str(omasa.num_channels-omasa.num_ism_channels) + "DIR" + str(omasa.dirs), omasa.audio[:, omasa.num_ism_channels:], omasa.fs)
		masa.metadata_file = omasa.metadata_files[-1]

		# OMASA -> Binaural
		if isinstance(out, audio.BinauralAudio):
		# render MASA and ISM part separately
		# ISM
		out_ism = copy.deepcopy(out)
		out_ism = deepcopy(out)
		render_oba_to_binaural(oba, out_ism, **kwargs)

		# MASA
		out_masa = copy.deepcopy(out)
		out_masa = deepcopy(out)
		render_masa_to_binaural(masa, out_masa, **kwargs)

		# combine results
		@@ -70,11 +72,11 @@ def convert_omasa(
		elif isinstance(out, audio.ChannelBasedAudio):
		# render MASA and ISM part separately
		# ISM
		out_ism = copy.deepcopy(out)
		out_ism = deepcopy(out)
		render_oba_to_cba(oba, out_ism)

		# MASA
		out_masa = copy.deepcopy(out)
		out_masa = deepcopy(out)
		render_masa_to_cba(masa, out_masa)

		# combine results
		@@ -84,11 +86,11 @@ def convert_omasa(
		elif isinstance(out, audio.SceneBasedAudio):
		# render MASA and ISM part separately
		# ISM
		out_ism = copy.deepcopy(out)
		out_ism = deepcopy(out)
		render_oba_to_sba(oba, out_ism)

		# MASA
		out_masa = copy.deepcopy(out)
		out_masa = deepcopy(out)
		render_masa_to_sba(masa, out_masa)

		# combine results
		@@ -96,17 +98,15 @@ def convert_omasa(

		# OMASA -> OSBA
		elif isinstance(out, audio.OSBAAudio):
		# TODO (treffehn)
		# check if ism object number is the same
		if out.num_ism_channels != masa.num_ism_channels:
		if out.num_ism_channels != omasa.num_ism_channels:
		raise ValueError("OMASA to OSBA conversion only possible if number of ISM objects matches")

		# only render MASA part
		out_sba = audio.fromtype("MASA")
		render_masa_to_sba(masa, out_sba)
		out_masa = deepcopy(out)
		render_masa_to_sba(masa, out_masa)

		out.audio[:, :omasa.num_ism_channels] = omasa.audio[:, :omasa.num_ism_channels]
		out.audio[:, omasa.num_ism_channels:] = out_sba.audio
		out.audio = np.concatenate((omasa.audio[:, :omasa.num_ism_channels], out_masa.audio), axis=1)

		else:
		raise NotImplementedError(

ivas_processing_scripts/audiotools/convert/osba.py

+0 −2

Original line number	Diff line number	Diff line
		@@ -107,8 +107,6 @@ def convert_osba(
		out_sba.metadata_file = out.metadata_files[-1]
		render_sba_to_masa(sba, out_sba)

		# out.audio[:, :osba.num_ism_channels] = osba.audio[:, :osba.num_ism_channels]
		# out.audio[:, osba.num_ism_channels:] = out_sba.audio
		out.audio = np.concatenate((osba.audio[:, :osba.num_ism_channels], out_sba.audio), axis=1)

		else:

ivas_processing_scripts/processing/evs.py

+1 −1

Original line number	Diff line number	Diff line
		@@ -284,7 +284,7 @@ class EVS(Processing):

		# copy MASA metadata for MASA pass-through
		if isinstance(self.in_fmt, audio.MetadataAssistedSpatialAudio):
		# TODO: add OMASA
		# TODO treffehn: add OMASA
		md_file_in = in_file.parent / (in_file.name + ".met")
		md_file_out = out_file.parent / (out_file.name + ".met")
		copyfile(md_file_in, md_file_out)

ivas_processing_scripts/processing/processing.py

+5 −0

Original line number	Diff line number	Diff line
		@@ -36,6 +36,7 @@ from abc import ABC, abstractmethod
		from itertools import repeat
		from multiprocessing import Pool
		from pathlib import Path
		import shutil
		from shutil import copyfile
		from time import sleep
		from typing import Iterable, Union
		@@ -502,6 +503,10 @@ def process_item(
		if processing_paths_meta[-1]:
		for idx, ppm in enumerate(processing_paths_meta[-1]):
		copyfile(ppm, out_meta[idx])
		if "MASA" in chain[-1].out_fmt:
		masa_meta_file_tmp = Path(processing_paths[-1]).parent / (Path(processing_paths[-1]).name + ".met")
		masa_meta_file = Path(out_file).parent / (Path(out_file).name + ".met")
		copyfile(masa_meta_file_tmp, masa_meta_file)


		def remove_preamble(x, out_fmt, fs, repeat_signal, preamble_len_ms, meta, logger):

ivas_processing_scripts/processing/processing_splitting_scaling.py

+1 −1

Original line number	Diff line number	Diff line
		@@ -344,7 +344,7 @@ def validate_tracefile(tracefile):
		if entry[1] >= 0:
		if entry[1] == prevRtpTs:
		raise ValueError(f"Error in JBM trace file at line {j}: duplicated rtpTs found")
		# TODO: (treffehn) include RTP time stamp overflow handling?
		# TODO: include RTP time stamp overflow handling?
		# else if (entry.rtpTs + rtpTsExtension < prevRtpTs) {
		# if (entry.rtpTs + rtpTsExtension + (1LL << 32) - prevRtpTs <
		# prevRtpTs - (entry.rtpTs + rtpTsExtension)) {