diff --git a/mythtv/libs/libmythtv/avformatdecoder.cpp b/mythtv/libs/libmythtv/avformatdecoder.cpp index 5dc65af..9fa7d00 100644 --- a/mythtv/libs/libmythtv/avformatdecoder.cpp +++ b/mythtv/libs/libmythtv/avformatdecoder.cpp @@ -3117,7 +3117,7 @@ int AvFormatDecoder::H264PreProcessPkt(AVStream *stream, AVPacket *pkt) { buf += m_h264_parser->addBytes(buf, buf_end - buf, 0); - if (m_h264_parser->stateChanged()) + if (m_h264_parser->stateChanged() && m_h264_parser->seen_SPS()) { if (m_h264_parser->FieldType() != H264Parser::FIELD_BOTTOM) { diff --git a/mythtv/libs/libmythtv/mpeg/H264Parser.cpp b/mythtv/libs/libmythtv/mpeg/H264Parser.cpp index 3e45c60..af2e137 100644 --- a/mythtv/libs/libmythtv/mpeg/H264Parser.cpp +++ b/mythtv/libs/libmythtv/mpeg/H264Parser.cpp @@ -116,14 +116,16 @@ void H264Parser::Reset(void) prev_pic_parameter_set_id = pic_parameter_set_id = -1; prev_field_pic_flag = field_pic_flag = -1; prev_bottom_field_flag = bottom_field_flag = -1; - prev_nal_ref_idc = nal_ref_idc = 0; + prev_nal_ref_idc = nal_ref_idc = 111; // != [0|1|2|3] prev_pic_order_cnt_type = pic_order_cnt_type = prev_pic_order_cnt_lsb = pic_order_cnt_lsb = 0; prev_delta_pic_order_cnt_bottom = delta_pic_order_cnt_bottom = 0; prev_delta_pic_order_cnt[0] = delta_pic_order_cnt[0] = 0; prev_delta_pic_order_cnt[1] = delta_pic_order_cnt[1] = 0; prev_nal_unit_type = nal_unit_type = UNKNOWN; - prev_idr_pic_id = idr_pic_id = 0; + + // The value of idr_pic_id shall be in the range of 0 to 65535, inclusive. + prev_idr_pic_id = idr_pic_id = 65536; log2_max_frame_num = log2_max_pic_order_cnt_lsb = 0; seq_parameter_set_id = 0; @@ -401,6 +403,7 @@ uint32_t H264Parser::addBytes(const uint8_t *bytes, const uint8_t *startP = bytes; const uint8_t *endP; bool found_start_code; + bool good_nal_unit; state_changed = false; on_frame = false; @@ -474,23 +477,49 @@ uint32_t H264Parser::addBytes(const uint8_t *bytes, nal_unit_type = sync_accumulator & 0x1f; nal_ref_idc = (sync_accumulator >> 5) & 0x3; - if (nal_unit_type == SPS || nal_unit_type == PPS || - nal_unit_type == SEI || NALisSlice(nal_unit_type)) + good_nal_unit = true; + if (nal_ref_idc) { - /* This is a NAL we need to parse. We may have the body - * of it in the part of the stream past to us this call, - * or we may get the rest in subsequent calls to addBytes. - * Either way, we set have_unfinished_NAL, so that we - * start filling the rbsp buffer */ - have_unfinished_NAL = true; + /* nal_ref_idc shall be equal to 0 for all NAL units having + * nal_unit_type equal to 6, 9, 10, 11, or 12. + */ + if (nal_unit_type == SEI || + (nal_unit_type >= AU_DELIMITER && + nal_unit_type <= FILLER_DATA)) + good_nal_unit = false; } - else if (nal_unit_type == AU_DELIMITER || - (nal_unit_type > SPS_EXT && - nal_unit_type < AUXILIARY_SLICE)) + else { - set_AU_pending(); + /* nal_ref_idc shall not be equal to 0 for NAL units with + * nal_unit_type equal to 5 + */ + if (nal_unit_type == SLICE_IDR) + good_nal_unit = false; } - } + + if (good_nal_unit) + { + if (nal_unit_type == SPS || nal_unit_type == PPS || + nal_unit_type == SEI || NALisSlice(nal_unit_type)) + { + /* This is a NAL we need to parse. We may have the body + * of it in the part of the stream past to us this call, + * or we may get the rest in subsequent calls to addBytes. + * Either way, we set have_unfinished_NAL, so that we + * start filling the rbsp buffer */ + have_unfinished_NAL = true; + } + else if (nal_unit_type == AU_DELIMITER || + (nal_unit_type > SPS_EXT && + nal_unit_type < AUXILIARY_SLICE)) + { + set_AU_pending(); + } + } + else + LOG(VB_GENERAL, LOG_ERR, + "H264Parser::addbytes: malformed NAL units"); + } //found start code } return startP - bytes; @@ -523,6 +552,9 @@ void H264Parser::processRBSP(bool rbsp_complete) set_AU_pending(); + if (!seen_sps) + SPS_offset = pkt_offset; + decode_SPS(&gb); } else if (nal_unit_type == PPS) @@ -576,6 +608,8 @@ void H264Parser::processRBSP(bool rbsp_complete) */ bool H264Parser::decode_Header(GetBitContext *gb) { + // uint first_mb_in_slice; + is_keyframe = false; if (log2_max_frame_num == 0 || pic_order_present_flag == -1) @@ -601,8 +635,7 @@ bool H264Parser::decode_Header(GetBitContext *gb) that precedes the current slice in decoding order and has the same value of colour_plane_id. */ - //uint first_mb_in_slice = get_ue_golomb(gb); - get_ue_golomb(gb); // Replaced above line + /* first_mb_in_slice = */ get_ue_golomb_long(gb); /* slice_type specifies the coding type of the slice according to @@ -611,7 +644,10 @@ bool H264Parser::decode_Header(GetBitContext *gb) When nal_unit_type is equal to 5 (IDR picture), slice_type shall be equal to 2, 4, 7, or 9 (I or SI) */ - slice_type = get_ue_golomb(gb); + slice_type = get_ue_golomb_31(gb); + + /* s->pict_type = golomb_to_pict_type[slice_type % 5]; + */ /* pic_parameter_set_id specifies the picture parameter set in @@ -642,6 +678,8 @@ bool H264Parser::decode_Header(GetBitContext *gb) bitstream.... If the current picture is an IDR picture, frame_num shall be equal to 0. + + When max_num_ref_frames is equal to 0, slice_type shall be equal to 2, 4, 7, or 9. */ frame_num = get_bits(gb, log2_max_frame_num); @@ -685,7 +723,8 @@ bool H264Parser::decode_Header(GetBitContext *gb) is_keyframe = true; } else - is_keyframe |= I_is_keyframe && isKeySlice(slice_type); + is_keyframe = (I_is_keyframe && isKeySlice(slice_type)); + /* pic_order_cnt_lsb specifies the picture order count modulo MaxPicOrderCntLsb for the top field of a coded frame or for a coded @@ -702,7 +741,7 @@ bool H264Parser::decode_Header(GetBitContext *gb) { pic_order_cnt_lsb = get_bits(gb, log2_max_pic_order_cnt_lsb); - if (pic_order_present_flag && !field_pic_flag) + if ((pic_order_present_flag == 1) && !field_pic_flag) delta_pic_order_cnt_bottom = get_se_golomb(gb); else delta_pic_order_cnt_bottom = 0; @@ -711,33 +750,44 @@ bool H264Parser::decode_Header(GetBitContext *gb) delta_pic_order_cnt_bottom = 0; /* - delta_pic_order_cnt[ 0 ] specifies the picture order count - difference from the expected picture order count for the top - field of a coded frame or for a coded field as specified in - subclause 8.2.1. The value of delta_pic_order_cnt[ 0 ] shall be - in the range of -231 to 231 - 1, inclusive. When this syntax - element is not present in the bitstream for the current slice, - it shall be inferred to be equal to 0. - - delta_pic_order_cnt[ 1 ] specifies the picture order count - difference from the expected picture order count for the bottom - field of a coded frame specified in subclause 8.2.1. The value - of delta_pic_order_cnt[ 1 ] shall be in the range of -231 to 231 - - 1, inclusive. When this syntax element is not present in the - bitstream for the current slice, it shall be inferred to be - equal to 0. - */ - if (pic_order_cnt_type == 1 && !delta_pic_order_always_zero_flag) + delta_pic_order_always_zero_flag equal to 1 specifies that + delta_pic_order_cnt[ 0 ] and delta_pic_order_cnt[ 1 ] are not + present in the slice headers of the sequence and shall be + inferred to be equal to 0. delta_pic_order_always_zero_flag + equal to 0 specifies that delta_pic_order_cnt[ 0 ] is present in + the slice headers of the sequence and delta_pic_order_cnt[ 1 ] + may be present in the slice headers of the sequence. + */ + if (delta_pic_order_always_zero_flag) + { + delta_pic_order_cnt[1] = delta_pic_order_cnt[0] = 0; + } + else if (pic_order_cnt_type == 1) { + /* + delta_pic_order_cnt[ 0 ] specifies the picture order count + difference from the expected picture order count for the top + field of a coded frame or for a coded field as specified in + subclause 8.2.1. The value of delta_pic_order_cnt[ 0 ] shall + be in the range of -2^31 to 2^31 - 1, inclusive. When this + syntax element is not present in the bitstream for the + current slice, it shall be inferred to be equal to 0. + + delta_pic_order_cnt[ 1 ] specifies the picture order count + difference from the expected picture order count for the + bottom field of a coded frame specified in subclause + 8.2.1. The value of delta_pic_order_cnt[ 1 ] shall be in the + range of -2^31 to 2^31 - 1, inclusive. When this syntax + element is not present in the bitstream for the current + slice, it shall be inferred to be equal to 0. + */ delta_pic_order_cnt[0] = get_se_golomb(gb); - if (pic_order_present_flag && !field_pic_flag) + if ((pic_order_present_flag == 1) && !field_pic_flag) delta_pic_order_cnt[1] = get_se_golomb(gb); else delta_pic_order_cnt[1] = 0; - } - else - delta_pic_order_cnt[0] = 0; + } /* redundant_pic_cnt shall be equal to 0 for slices and slice data @@ -748,7 +798,6 @@ bool H264Parser::decode_Header(GetBitContext *gb) be equal to 0. The value of redundant_pic_cnt shall be in the range of 0 to 127, inclusive. */ - redundant_pic_cnt = redundant_pic_cnt_present_flag ? get_ue_golomb(gb) : 0; return true; @@ -760,10 +809,13 @@ bool H264Parser::decode_Header(GetBitContext *gb) void H264Parser::decode_SPS(GetBitContext * gb) { int profile_idc; + int lastScale; + int nextScale; + int deltaScale; seen_sps = true; - profile_idc = get_bits(gb, 8); // profile_idc + profile_idc = get_bits(gb, 8); get_bits1(gb); // constraint_set0_flag get_bits1(gb); // constraint_set1_flag get_bits1(gb); // constraint_set2_flag @@ -772,9 +824,11 @@ void H264Parser::decode_SPS(GetBitContext * gb) get_bits(gb, 8); // level_idc get_ue_golomb(gb); // sps_id - if (profile_idc >= 100) + if (profile_idc == 100 || profile_idc == 110 || profile_idc == 122 || + profile_idc == 244 || profile_idc == 44 || profile_idc == 83 || + profile_idc == 86 || profile_idc == 118 || profile_idc == 128 ) { // high profile - if ((chroma_format_idc = get_ue_golomb(gb)) == 3) // chroma_format_idc + if ((chroma_format_idc = get_ue_golomb(gb)) == 3) separate_colour_plane_flag = (get_bits1(gb) == 1); get_ue_golomb(gb); // bit_depth_luma_minus8 @@ -785,12 +839,18 @@ void H264Parser::decode_SPS(GetBitContext * gb) { for (int idx = 0; idx < ((chroma_format_idc != 3) ? 8 : 12); ++idx) { - if (get_bits1(gb)) // Scaling list presnent + if (get_bits1(gb)) // Scaling list present { + lastScale = nextScale = 8; int sl_n = ((idx < 6) ? 16 : 64); - for(int sl_i = 0; sl_i < sl_n; sl_i++) + for(int sl_i = 0; sl_i < sl_n; ++sl_i) { - get_se_golomb(gb); + if (nextScale != 0) + { + deltaScale = get_se_golomb(gb); + nextScale = (lastScale + deltaScale + 256) % 256; + } + lastScale = (nextScale == 0) ? lastScale : nextScale; } } } @@ -809,6 +869,7 @@ void H264Parser::decode_SPS(GetBitContext * gb) int offset_for_non_ref_pic; int offset_for_top_to_bottom_field; uint tmp; + // bool gaps_in_frame_num_allowed_flag; /* pic_order_cnt_type specifies the method to decode picture order @@ -886,8 +947,7 @@ void H264Parser::decode_SPS(GetBitContext * gb) decoding process in case of an inferred gap between values of frame_num as specified in subclause 8.2.5.2. */ - //bool gaps_in_frame_num_allowed_flag = get_bits1(gb); - get_bits1(gb); // Replaced above line + /* gaps_in_frame_num_allowed_flag = */ get_bits1(gb); /* pic_width_in_mbs_minus1 plus 1 specifies the width of each @@ -1049,25 +1109,27 @@ void H264Parser::decode_SEI(GetBitContext *gb) { do { type += show_bits(gb, 8); - } while (get_bits(gb, 8) == 255); + } while (get_bits(gb, 8) == 0xFF); do { size += show_bits(gb, 8); - } while (get_bits(gb, 8) == 255); + } while (get_bits(gb, 8) == 0xFF); switch (type) { - case SEI_TYPE_RECOVERY_POINT: - recovery_frame_cnt = get_ue_golomb(gb); - exact_match_flag = get_bits1(gb); - broken_link_flag = get_bits1(gb); - changing_group_slice_idc = get_bits(gb, 2); - au_contains_keyframe_message = (recovery_frame_cnt == 0); - return; - - default: - skip_bits(gb, size * 8); - break; + case SEI_TYPE_RECOVERY_POINT: + recovery_frame_cnt = get_ue_golomb(gb); + exact_match_flag = get_bits1(gb); + broken_link_flag = get_bits1(gb); + changing_group_slice_idc = get_bits(gb, 2); + au_contains_keyframe_message = (recovery_frame_cnt == 0); + if ((size - 12) > 0) + skip_bits(gb, (size - 12) * 8); + return; + + default: + skip_bits(gb, size * 8); + break; } } diff --git a/mythtv/libs/libmythtv/mpeg/H264Parser.h b/mythtv/libs/libmythtv/mpeg/H264Parser.h index 9c984e0..eb95779 100644 --- a/mythtv/libs/libmythtv/mpeg/H264Parser.h +++ b/mythtv/libs/libmythtv/mpeg/H264Parser.h @@ -60,7 +60,7 @@ class H264Parser { // ITU-T Rec. H.264 table 7-1 enum NAL_unit_type { UNKNOWN = 0, - SLICE = 1, + SLICE = 1, // 1 - 5 are VCL NAL units SLICE_DPA = 2, SLICE_DPB = 3, SLICE_DPC = 4, @@ -73,11 +73,15 @@ class H264Parser { END_STREAM = 11, FILLER_DATA = 12, SPS_EXT = 13, - AUXILIARY_SLICE = 19 + NALU_prefix = 14, + SPS_subset = 15, + AUXILIARY_SLICE = 19, + SLICE_EXTENSION = 20 }; enum SEI_type { SEI_TYPE_PIC_TIMING = 1, + SEI_FILLER_PAYLOAD = 3, SEI_TYPE_USER_DATA_UNREGISTERED = 5, SEI_TYPE_RECOVERY_POINT = 6 }; @@ -132,7 +136,7 @@ class H264Parser { } bool onFrameStart(void) const { return on_frame; } - bool onKeyFrameStart(void) const { return on_key_frame; } + bool onKeyFrameStart(void) const { return seen_sps && on_key_frame; } uint pictureWidth(void) const { return pic_width; } uint pictureHeight(void) const { return pic_height; } @@ -147,7 +151,9 @@ class H264Parser { uint64_t frameAUstreamOffset(void) const {return frame_start_offset;} uint64_t keyframeAUstreamOffset(void) const {return keyframe_start_offset;} + uint64_t SPSstreamOffset(void) const {return SPS_offset;} + // == NAL_type AU_delimiter: primary_pic_type = 5 static int isKeySlice(uint slice_type) { return (slice_type == SLICE_I || @@ -164,6 +170,7 @@ class H264Parser { } void use_I_forKeyframes(bool val) { I_is_keyframe = val; } + bool using_I_forKeyframes(void) const { return I_is_keyframe; } uint32_t GetTimeScale(void) const { return timeScale; } @@ -172,6 +179,10 @@ class H264Parser { void parse_SPS(uint8_t *sps, uint32_t sps_size, bool& interlaced, int32_t& max_ref_frames); + bool seen_SPS(void) const { return seen_sps; } + + bool found_AU(void) const { return AU_pending; } + private: enum constants {EXTENDED_SAR = 255}; @@ -247,6 +258,7 @@ class H264Parser { bool fixedRate; uint64_t pkt_offset, AU_offset, frame_start_offset, keyframe_start_offset; + uint64_t SPS_offset; bool on_frame, on_key_frame; };