Merge branch '8-ambisonics-evs-reference-conditions-add-option-to-process-foa-only' into 'main' (95fb0a2b) · Commits · IVAS Codec Public Collaboration / IVAS Processing Scripts

ivas_processing_scripts/audiotools/audiofile.py

+14 −3

Original line number	Diff line number	Diff line
		@@ -34,11 +34,13 @@ import logging
		import struct
		from pathlib import Path
		from typing import Optional, Tuple, Union
		from warnings import warn

		import numpy as np
		import scipy.io.wavfile as wav

		from .audioarray import trim, window
		from .constants import VERT_HOA_CHANNELS_ACN

		logger = logging.getLogger("__main__")
		logger.setLevel(logging.DEBUG)
		@@ -326,6 +328,8 @@ def split_channels(
		in_file: str,
		out_filenames: list,
		in_nchans: int,
		out_nchans: int,
		is_planar: Optional[bool] = False,
		in_fs: Optional[int] = 48000,
		) -> None:
		"""
		@@ -339,6 +343,10 @@ def split_channels(
		List of output file names (.pcm, .raw or .wav)
		in_nchans: int
		Input number of channels
		out_nchans: int
		Number of channels to be split
		is_planar: Optional[bool]
		If true vertical SBA channels are set to zero
		in_fs: Optional[int] = 48000
		Input sampling rate, default 48000 Hz

		@@ -350,14 +358,17 @@ def split_channels(
		# validation
		if in_nchans is None:
		raise ValueError("Number of channels to split must be specified!")
		if in_nchans != len(out_filenames):
		print(
		if out_nchans != len(out_filenames):
		warn(
		"Split: Mismatch between number of channels and output filenames length. Truncating output filenames list."
		)
		out_filenames = out_filenames[:in_nchans]
		out_filenames = out_filenames[:out_nchans]

		x, in_fs = read(in_file, nchannels=in_nchans, fs=in_fs)

		if is_planar:
		x[:, VERT_HOA_CHANNELS_ACN[VERT_HOA_CHANNELS_ACN < in_nchans]] = 0

		# Write output files
		for idx, out_file in enumerate(out_filenames):
		# extract correct channel

+2 −0

Original line number	Diff line number	Diff line
		@@ -300,11 +300,13 @@ def get_processing_chain(
		"tx": tx_cfg,
		"preamble": preamble,
		"evs_lfe_9k6bps_nb": evs_lfe_9k6bps_nb,
		"sba_format": cond_cfg.get("sba_format", tmp_in_fmt),
		}
		)
		)
		# update values to reflect decoder output
		tmp_in_fs = dec_cfg.get("fs", tmp_in_fs)
		tmp_in_fmt = cond_cfg.get("sba_format", cfg.input["fmt"])

		elif cond_cfg["type"] == "ivas":
		cod_cfg = cond_cfg["cod"]

+20 −3

Original line number	Diff line number	Diff line
		@@ -167,16 +167,33 @@ class EVS(Processing):
		if not self.out_fs:
		self.out_fs = self.in_fs

		# consider SBA format for condition
		condition_fmt = audio.fromtype(self.sba_format)
		if isinstance(self.in_fmt, audio.SceneBasedAudio):
		if not isinstance(condition_fmt, audio.SceneBasedAudio) or (
		condition_fmt.num_channels > self.in_fmt.num_channels
		):
		raise ValueError(
		"SBA format has to be SBA with lower order than input signal"
		)
		is_planar = condition_fmt.is_planar
		else:
		if self.in_fmt.name != condition_fmt.name:
		raise ValueError("Cannot specify SBA format for non SBA input formats")
		is_planar = False

		# Split the channels to prepare for multi-mono coding
		split_chan_files = [
		out_file.with_suffix(f".{i}.pcm") for i in range(self.in_fmt.num_channels)
		out_file.with_suffix(f".{i}.pcm") for i in range(condition_fmt.num_channels)
		]

		split_channels(
		in_file,
		split_chan_files,
		self.in_fmt.num_channels,
		self.in_fs,
		in_nchans=self.in_fmt.num_channels,
		out_nchans=condition_fmt.num_channels,
		in_fs=self.in_fs,
		is_planar=is_planar,
		)

		# run processing

+1 −0

Original line number	Diff line number	Diff line
		@@ -265,6 +265,7 @@ conditions_to_generate:
		### For possible arguments see overall bitstream modification
		type: "JBM"
		error_profile: 3
		sba_format: "PLANARFOA"

		################################################
		### Post-processing