GitLab

Prefix SEI NAL units are now buffered until the first slice activates VPS/SPS/PPS

Patch-by: Karl Sharman

RExt: Changed types to be consistent- bool to Bool, int to Int, void to Void, double to Double, char to Char, and replaced long

From: Karl Sharman <karl.sharman@eu.sony.com>

From: Karl Sharman <karl.sharman@eu.sony.com>

RExt: Added bit stream analyser debug for decoder and new command-line-based RGB/GBR conversion mechanism.

From: Karl Sharman <karl.sharman@eu.sony.com>

apply (adapted) patch for issue #994 (Emulation prevention bytes ignored by HM decoder to calculate entry_point_offset[ i ] when entropy_coding_sync_enabled_flag is enabled.)

From: Kiran Misra <misrak@sharplabs.com>

From: David Flynn <dflynn@rim.com>

From: David Flynn <dflynn@rim.com>

Added some missing checks of TemporalId restrictions to the decoder
From Rickard Sjoberg <rickard.sjoberg@ericsson.com>

From: Karl Sharman <Karl.Sharman@eu.sony.com>

Implement the new mapping for nal unit types as defined in Table 7-1 of J1003-d7 - Macro NAL_UNIT_TYPES_J1003_D7; from Sandeep Kanumuri <skanumu@microsoft.com>

Contains only boiler-plate support (no real functionality) for the new nal unit types introduced. Going further, proponents of each nal unit type can add functionality (labeling, encoder/decoder state updates etc).

J0550: Define nal_unit_header() method. (MACRO: NAL_UNIT_HEADER); from Adarsh K. Ramasubramonian (aramasub@qualcomm.com)

J0550: Remove nal_ref_flag, allocate extra bit to reserved bits, and re-order syntax to put reserved bits after nal_unit_type (MACRO: REMOVE_NAL_REF_FLAG); from Adarsh K. Ramasubramonian (aramasub@qualcomm.com)

J0550: Signal temporal_id_plus1 instead of temporal_id in NAL unit, and change the value of reserved_one_5bits to zero. (MACRO: TEMPORAL_ID_PLUS1); from Adarsh K. Ramasubramonian (aramasub@qualcomm.com)

JCTVC-I0607: new NAL unit types CRANT, BLA, BLANT, and TFD introduced (MACRO: NEW_NAL_UNIT_TYPES); from Adarsh K. Ramasubramonian <aramasub@qualcomm.com>