JVET-R0338 AHG9: A summary of proposals on chroma deblocking t_c and β offsets signalling [Y.-K. Wang (Bytedance)]

Discussed in AHG Session 1.1 Monday 6 April at 1315 UTC (chaired by GJS & YKW).

This contribution intends to provide a summary of the 12 proposals on signalling of chroma deblocking tc and β offsets submitted to this JVET meeting by the 3 April 2020 submission deadline.

It is suggested that this summary, in terms of a list of design questions, is used for the reviewing of these proposals, such that the discussions can be in a more structured and efficient manner.

1. Skip the signalling of the chroma t_c and β deblocking offset syntax elements (SEs) in the PPS when the chroma format is ( 4:0:0 or (4:4:4 and the separate color plane coding mode is in use) ) and/or when the parameter values for chroma are the same as for luma?
   1. Yes, (JVET-R0077, JVET-R0078, JVET-R0095, JVET-R0106, JVET-R0152, JVET-R0172, JVET-R0206, JVET-R0218, JVET-R0232), and condition the SEs on
      1. A new PPS flag for controlling the presence of chroma deblocking parameters (JVET-R0077, JVET-R0078, JVET-R0081, JVET-R0106, JVET-R0206)
      2. The existing pps_chroma_tool_offsets_present_flag currently for controlling the presence of the QP offsets in the PPS (JVET-R0078, JVET-R0095, JVET-R0106, JVET-R0152, JVET-R0172, JVET-R0206, JVET-R0218, JVET-R0232) AHG Recommendation (cleanup): Recommended by AHG. (The editor may also consider renaming the flag.) If the flag is zero, the chroma offsets (if needed) are inferred from the luma offsets.
      3. ChromaArrayType (JVET-R0172)
   2. No. (JVET-R0048, JVET-R0079, JVET-R0081, JVET-R0232)
      1. Impose semantics constraints that values shall be equal to 0 when ChromaArrayType is equal to 0. (JVET-R0079, JVET-R0081, JVET-R0232)
2. Skip the signalling of the chroma t_c and β deblocking offset syntax elements (SEs) in the PH and the SH when the chroma format is (4:0:0 or (4:4:4 and the separate color plane coding mode is in use) ) and/or when the parameter values for chroma are the same for luma?
   1. Yes, (all the 12 contributions), and condition the SEs on
      1. ChromaArrayType (JVET-R0048, JVET-R0078, JVET-R0079, JVET-R0081, JVET-R0095, JVET-R0106, JVET-R0152, JVET-R0172, JVET-R0206, JVET-R0218, JVET-R0232)
      2. A new PPS flag for controlling the presence of chroma deblocking parameters (JVET-R0077, JVET-R0078, JVET-R0106)
      3. The existing pps_chroma_tool_offsets_present_flag currently for controlling the presence of the QP offsets in the PPS (JVET-R0078, JVET-R0152, JVET-R0232) AHG Recommendation (cleanup): Recommended by AHG. (The editor may also consider renaming the luma beta and tc offset control syntax elements.) If the flag is zero, the chroma offsets (if needed) are inferred from the luma offsets.
      4. A new PH flag and a new SH flag (JVET-R0081, JVET-R0206)
   2. No, but impose semantics constraints that values shall be equal to 0 when ChromaArrayType is equal to 0. (JVET-R0079, JVET-R0232)