[fix] scaling of ISM and SBA rendered output before combining for OSBA (335d6dc3) · Commits · IVAS Codec Public Collaboration / IVAS Processing Scripts

ivas_processing_scripts/audiotools/convert/osba.py

+6 −5

Original line number	Diff line number	Diff line
		@@ -81,14 +81,14 @@ def convert_osba(

		# combine results
		try:
		out.audio = out_ism.audio + out_sba.audio
		out.audio = 0.5 * (out_ism.audio + out_sba.audio)
		except ValueError as e:
		if (
		out_ism.audio.shape[0] != out_sba.audio.shape[0]
		and out_sba.audio.shape[0] == osba.audio.shape[0]
		):
		out.audio = out_sba.audio
		out.audio[: out_ism.audio.shape[0], :] += out_ism.audio
		out.audio = 0.5 * out_sba.audio
		out.audio[: out_ism.audio.shape[0], :] += 0.5 * out_ism.audio
		else:
		raise e

		@@ -104,7 +104,7 @@ def convert_osba(
		render_sba_to_cba(sba, out_sba)

		# combine results
		out.audio = out_ism.audio + out_sba.audio
		out.audio = 0.5 * (out_ism.audio + out_sba.audio)

		# OSBA -> SBA
		elif isinstance(out, audio.SceneBasedAudio):
		@@ -118,7 +118,7 @@ def convert_osba(
		render_sba_to_sba(sba, out_sba)

		# combine results
		out.audio = out_ism.audio + out_sba.audio
		out.audio = 0.5 * (out_ism.audio + out_sba.audio)

		# OSBA -> OMASA
		elif isinstance(out, audio.OMASAAudio):
		@@ -136,6 +136,7 @@ def convert_osba(
		out.audio = np.concatenate(
		(osba.audio[:, : osba.num_ism_channels], out_sba.audio), axis=1
		)
		out.audio *= 0.5

		# OSBA -> OSBA
		elif isinstance(out, audio.OSBAAudio):