International organisation for standardisation

6.10.6NAL unit header

6.10.6.1.1.1.1.1.1JCTVC-F463 On NAL unit header [Y.-K. Wang (Huawei)]

This document presented the following proposals:

1) Dropping forbidden_zero_bit from the NAL unit header;

Remark: It was also envisioned as an error flag that could be set when a problem was detected to flag error conditions for the decoder.

Remark: Define a special "bad NAL unit type" to indicate errors?

Remark: Making it reserved would let us use it later if desirable.

Remark: That had something to do with distinguishing from MPEG-2 Systems start codes.

The group agreed to keep it zero because of MPEG-2 Systems.

2) Changing nal_ref_idc (2 bits) in the NAL unit header to nal_ref_flag (1 bit);

Decision: Adopt (and absorb it into the nal_unit_type).

3) Unification of the NAL unit header to be a fixed-length 2-byte header;

Further study of this was encouraged.

4) A sub-bitstream extraction process utilizing the unified NAL unit header.

Further study of this was encouraged.