JVET-AM0166 AHG9: Miscellaneous aspects in VSEI v4 [J. Nam, H. Tan, J. Lee, C. Kim, J. Lim, S. Kim (LGE)]

JVET-AM0166 also relates to the MI and PRI SEI messages

This contribution proposes miscellaneous modifications that are asserted to provide improvement to some aspects of SEI messages in VSEI v4 draft. The proposed modifications are as follows:

(Related to SPO/PON SEI messages)

4. On prefix and suffix SEI messages in PON SEI

5. On prefix and suffix SEI messages in SPO SEI

Item 4 proposes that the following two constraints to be added into the current spec text:

1. A PON SEI message that is present the bistream as a prefix SEI message shall not contain a PON-nested SEI message that can be present in a bistream as suffix SEI message only.
2. A PON SEI message that is present the bistream as a suffix SEI message shall not contain a PON-nested SEI message that can be present in a bistream as prefix SEI message only.

It was suggested that the specific wording should mention NAL units, and that the constraint should be in the coding specification interfaces rather than VSEI.

Conditionally adopt item 4 subject to review of text, with constraint in the coding spec interface text.

Item 5 proposes to add the constraint such that an SEI message that can be a suffix SEI message only and is included in a processing chain shall not have processing order that is lower than another SEI message that can be a prefix SEI message only.the SEI message included in prefix vs. suffix only affects the coding order and doesn’t necessarily affect the processing order.

No action on item 5.

Further discussion at 1755 on Monday 30 June 2025.

Adoption of item 4 confirmed.

JVET-AM0166 AHG9: Miscellaneous aspects in VSEI v4 [J. Nam, H. Tan, J. Lee, C. Kim, J. Lim, S. Kim (LGE)]

JVET-AM0166 also relates to the SPO and MI SEI messages

(Related to PRI SEI message)

3. On handling of persistence for PRI SEI

It is asserted that there is an issue with the the 3^rd bullet in the PRI persistence semantics. This text is proposed for the 3^rd bullet:

: – A picture in an AU associated with a Another packed regions information SEI message is output present in an AU that follows the current picture AU in output order.

It was suggested that the proposed text may have issues in a multi-layer case. This text was suggested to replace all 3 bullets:

pri_persistence_flag equal to 1 specifies that the packed regions information SEI message applies to the current AU and persists for all subsequent AUs in output order in the current CVS until the next AU, when present, that follows the current AU in output order and contains a packed regions information SEI message.

Agreed to modify the persistence language. Exact text delegated to the editors.

JVET-AM0166 AHG9: Miscellaneous aspects in VSEI v4 [J. Nam, H. Tan, J. Lee, C. Kim, J. Lim, S. Kim (LGE)]

JVET-AM0166 also relates to the SPO and PRI SEI messages

This contribution proposes miscellaneous modifications that are asserted to provide improvement to some aspects of SEI messages in VSEI v4 draft. The proposed modifications are as follows:

(Related to MI SEI message)

1. On signalling of min and max wavelengths in MI SEI
2. On constraint for wavelength in MI SEI

Item 1 proposes modified semantics for min and max wavelength parameters. “when present” is added to semantics. It is suggested to remove the inference values when the syntax elements are not present.

It is noted that JVET-AM0191 also has modifications to the semantics.

Further discussion requested item 1 after proposal with suggestions above is combined with JVET-AM0191.

Item 2 proposes that the max value is constrained to be larger than the min value.

Decision: adopt item 2.

Further discussed 1750 – 1755 Monday 30 June 2025.

Decision: adopt item 1.

Text description information SEI message (1)

Contributions in this area were discussed during 1830–1745 on Friday 27 June 2025 (chaired by J. Boyce).

JVET-AM0166 AHG9: Miscellaneous aspects in VSEI v4