Organisation internationale de normalisation
Yüklə 5,54 Mb.
|
| MPEG Surround
Seungkwon Beack, ETRI, presented
The contribution reports on the conclusion of the MPEG-H 3D Audio Phase 2 CfP “core experiment.” The work was a collaboration between ETRI and FhG-IIS. The new technology is an extension to MPEG Surround consisting of 1-to-2 decoder boxed giving a N/2 to N channel upmix. In terms of the MPEG Surround normative architecture, the new technology specifies
It was the consensus of the Audio subgroup to take the contribution text and issue it as
And to issue a Working Draft of the associated Reference Software from this meeting, with a 1-month editing period.
SAOC Adrian Murtaza, FhG, presented the following documents
It was the consensus of the Audio subgroup to take the WD texts from the 109th MPEG meeting and issue as
Adrian Murtaza, FhG, presented the following documents
The contribution documents several errors in the SAOC text:
It was the consensus of the Audio subgroup to issue the contribution text as
USAC Max Neuendorf, FhG, presented
The presenter noted that USAC 16 kb/s streaming of stereo signals is very interesting for mobile applications having only 2G connectivity, and that such streaming might use the DASH delivery method. MPEG-H 3D Audio specifies the structure of “Immediate Playout Frame” specifically to insure seamless stream splices between any DASH adaptation sets. The contribution proposes to bring the IPF technology from MPEG-H 3D Audio core coder into MPEG-D USAC. The presenter noted that the IPF is added in a way that new bitstreams are fully compatible with USAC decoders that are not aware of IPF technology. Christof Fresch, Dolby, suggested to add a sentence indicating that “It is strongly recommended to run the dcoder with the same sampling frequency when switching e.g. bitrates.”
Toru Chinen, Sony, gave a verbal report on
In summary: 4 of 5 conformance streams were successfully cross-checked. Christof Fersch, Dolby, gave a verbal report on
In summary: Dolby generated 5 conformance streams, but recommends that the definition of the 5th stream be changed to permit a successul cross-check. Werner de Bruijn, Philips, gave a verbal report on
Philips experts worked with FhG-IIS experts so that all streams have been generated and cross-checked. Max Neuendorf, FhG, presented
The presenter thanked Sony, Dolby Philips and Samsung experts for their efforts in stream generation and cross-check. The contribution reported on open issues in USAC Conformance going forward from this meeting. The presenter requested that the issues with existing streams be captured in a defect report. It was the consensus of the Audio subgroup to edit the contribution text to make the following documents
DRC
The presenter noted that MPEG-D DRC is International Standard. This contribution notes that DRC is very flexible and supports many DRC sets for many “use profiles” and many channel or object groups. One consequence of this flexibility is that there may be a significant overhead in a bitstream for DRC sets. The contribution proposes a new mode of operation in which DRC sets are generated in the decoder. This has syntax that is backward compatible with the base standard syntax. The decoder-generated DRC sets results in a small increase (2% to 4%) in decoder complexity. Additionally it may require an additional delay in the decoder of 1 audio frame (e.g. 1024 samples), where such additional delay is signalled in the bitstream. The proposed change requires:
The presenter welcomes experts to consider and bring information on profiles and levels applicable to MPEG-D DRC. It was the consensus of the Audio subgroup to take the contribution text and make it Working Draft of MPEG-D DRC AMD X.
|