【H264】x264头文件结构体详解
聂涛
/*****************************************************************************
* x264.h: x264 public header
*****************************************************************************
* Copyright (C) 2003-2016 x264 project
*
* Authors: Laurent Aimar <fenrir@via.ecp.fr>
* Loren Merritt <lorenm@u.washington.edu>
* Fiona Glaser <fiona@x264.com>
*
* This program is free software; you can redistribute it and/or modify
* it under the terms of the GNU General Public License as published by
* the Free Software Foundation; either version 2 of the License, or
* (at your option) any later version.
*
* This program is distributed in the hope that it will be useful,
* but WITHOUT ANY WARRANTY; without even the implied warranty of
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
* GNU General Public License for more details.
*
* You should have received a copy of the GNU General Public License
* along with this program; if not, write to the Free Software
* Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02111, USA.
*
* This program is also available under a commercial proprietary license.
* For more information, contact us at licensing@x264.com.
*****************************************************************************/
#ifndef X264_X264_H
#define X264_X264_H
#ifdef __cplusplus
extern "C" {
#endif
#if !defined(_STDINT_H) && !defined(_STDINT_H_) && !defined(_STDINT_H_INCLUDED) && !defined(_STDINT) &&\
!defined(_SYS_STDINT_H_) && !defined(_INTTYPES_H) && !defined(_INTTYPES_H_) && !defined(_INTTYPES)
# ifdef _MSC_VER
# pragma message("You must include stdint.h or inttypes.h before x264.h")
# else
# warning You must include stdint.h or inttypes.h before x264.h
# endif
#endif
#include <stdarg.h>
#include "x264_config.h"
#define X264_BUILD 148
/* Application developers planning to link against a shared library version of
* libx264 from a Microsoft Visual Studio or similar development environment
* will need to define X264_API_IMPORTS before including this header.
* This clause does not apply to MinGW, similar development environments, or non
* Windows platforms. */
#ifdef X264_API_IMPORTS
#define X264_API __declspec(dllimport)
#else
#define X264_API
#endif
// AVC -> Advanced video coding 高级视频编码 ,H264最早编码
// SVC -> Scalable video coding 可分级视频编码
// H.264 SVC以H.264 AVC视频编解码器标准为基础，利用了AVC编解码器的各种高效算法工具，在编码产生的编码视频时间上（帧率）、
// 空间上（分辨率）可扩展，并且是在视频质量方面可扩展的，可产生不同帧速率、分辨率或质量等级的解码视频。
/*
H.264 SVC通过在在编码码流嵌入具有相关性的多个不同分辨率的子流实现空间上的可分级。
上层所有的视频帧构成了高分辨率视频，下层所有的视频帧构成了低分辨率视频
为此，具备H.264 SVC编码的视频会议系统，在保证高效的视频压缩性能的基础上，
视频广播端可以通过一次编码产生具有不同帧率、分辨率的视频压缩码流，以适应
不同网络带宽、不同的显示屏幕和终端解码能力的应用需求，从而有效地避免了视频会议系统中MCU上复杂而昂贵的转码。
*/
/* x264_t:
* opaque handler for encoder */
typedef struct x264_t x264_t;
/****************************************************************************
* NAL structure and functions
****************************************************************************/
enum nal_unit_type_e
{
NAL_UNKNOWN = 0,
NAL_SLICE = 1,
NAL_SLICE_DPA = 2,
NAL_SLICE_DPB = 3,
NAL_SLICE_DPC = 4,
NAL_SLICE_IDR = 5, /* ref_idc != 0 */
NAL_SEI = 6, /* ref_idc == 0 */
NAL_SPS = 7,
NAL_PPS = 8,
NAL_AUD = 9,
NAL_FILLER = 12,
/* ref_idc == 0 for 6,9,10,11,12 */
};
/*
nal_unit_type_e枚举类型相关知识：
DP（数据分割/分区）
I/P/B帧分别包含的是I/P/B条带类型，就数据流来看，看到的是条带，就编码来看，看到的是帧
I条带只包括I宏块、P条带除了P宏块还可以是I宏块，B条带除了B宏块也可以是I宏块。
DP就是将条带分区，原因：
分区是为了对重要程度不同的句法元素采用不同的传输手段，比如片头的句法元素传输时可以要求对方确认，而为了实时性部分编码数据丢失就算了，不重传了；
隔离错误，如果一个条带在类似于UDP这样的非可靠传输中丢失或部分数据出现错误从而解码失败时，不影响其它条带解码。
将片头数据、片里面包括的帧内编码数据、片里面包括的帧间编码数据分成三种封装格式，并分别命名为分类2、分类3、分类4
这三类对应了
类型1：非IDR图像的编码条带（不分区），包括I(非IDR)、P、B、Bref
类型2：编码条带数据分区A（片头数据）
类型3：编码条带数据分区B（片里面包括的帧内编码数据）
类型4：编码条带数据分区C（片里面包括的帧间编码数据）
类型5：IDR图像的编码条带（不分区），只有IDR,从类型1与5可以分辨出IDR帧，尽一步跟据条带头的类型可以分辨出I帧
所以：
I 帧，可以用1封装，或者是2和3类型的组合
P/B 帧，可以用2和4类型或者2，3和4的类型组合封装
IDR 帧，5类型
*/
enum nal_priority_e
{
NAL_PRIORITY_DISPOSABLE = 0,
NAL_PRIORITY_LOW = 1,
NAL_PRIORITY_HIGH = 2,
NAL_PRIORITY_HIGHEST = 3,
};
/* The data within the payload is already NAL-encapsulated; the ref_idc and type
* are merely in the struct for easy access by the calling application.
* All data returned in an x264_nal_t, including the data in p_payload, is no longer
* valid after the next call to x264_encoder_encode. Thus it must be used or copied
* before calling x264_encoder_encode or x264_encoder_headers again. */
typedef struct x264_nal_t
{
int i_ref_idc; /* nal_priority_e */
int i_type; /* nal_unit_type_e */
int b_long_startcode;
int i_first_mb; /* If this NAL is a slice, the index of the first MB in the slice. */
int i_last_mb; /* If this NAL is a slice, the index of the last MB in the slice. */
/* Size of payload (including any padding) in bytes. */
int i_payload;
/* If param->b_annexb is set, Annex-B bytestream with startcode(前4字节0x00000001).
* Otherwise, startcode is replaced with a 4-byte size.(前4字节是编码后纯数据的长度)
* This size is the size used in mp4/similar muxing; it is equal to i_payload-4 */
uint8_t *p_payload;
/* Size of padding in bytes. */
int i_padding;//尾部补全的字节数
} x264_nal_t;
/****************************************************************************
* Encoder parameters
****************************************************************************/
/* CPU flags */
/* x86 */
#define X264_CPU_CMOV 0x0000001
#define X264_CPU_MMX 0x0000002
#define X264_CPU_MMX2 0x0000004 /* MMX2 aka MMXEXT aka ISSE */
#define X264_CPU_MMXEXT X264_CPU_MMX2
#define X264_CPU_SSE 0x0000008
#define X264_CPU_SSE2 0x0000010
#define X264_CPU_SSE3 0x0000020
#define X264_CPU_SSSE3 0x0000040
#define X264_CPU_SSE4 0x0000080 /* SSE4.1 */
#define X264_CPU_SSE42 0x0000100 /* SSE4.2 */
#define X264_CPU_LZCNT 0x0000200 /* Phenom support for "leading zero count" instruction. */
#define X264_CPU_AVX 0x0000400 /* AVX support: requires OS support even if YMM registers aren't used. */
#define X264_CPU_XOP 0x0000800 /* AMD XOP */
#define X264_CPU_FMA4 0x0001000 /* AMD FMA4 */
#define X264_CPU_FMA3 0x0002000 /* FMA3 */
#define X264_CPU_AVX2 0x0004000 /* AVX2 */
#define X264_CPU_BMI1 0x0008000 /* BMI1 */
#define X264_CPU_BMI2 0x0010000 /* BMI2 */
/* x86 modifiers */
#define X264_CPU_CACHELINE_32 0x0020000 /* avoid memory loads that span the border between two cachelines */
#define X264_CPU_CACHELINE_64 0x0040000 /* 32/64 is the size of a cacheline in bytes */
#define X264_CPU_SSE2_IS_SLOW 0x0080000 /* avoid most SSE2 functions on Athlon64 */
#define X264_CPU_SSE2_IS_FAST 0x0100000 /* a few functions are only faster on Core2 and Phenom */
#define X264_CPU_SLOW_SHUFFLE 0x0200000 /* The Conroe has a slow shuffle unit (relative to overall SSE performance) */
#define X264_CPU_STACK_MOD4 0x0400000 /* if stack is only mod4 and not mod16 */
#define X264_CPU_SLOW_CTZ 0x0800000 /* BSR/BSF x86 instructions are really slow on some CPUs */
#define X264_CPU_SLOW_ATOM 0x1000000 /* The Atom is terrible: slow SSE unaligned loads, slow
* SIMD multiplies, slow SIMD variable shifts, slow pshufb,
* cacheline split penalties -- gather everything here that
* isn't shared by other CPUs to avoid making half a dozen
* new SLOW flags. */
#define X264_CPU_SLOW_PSHUFB 0x2000000 /* such as on the Intel Atom */
#define X264_CPU_SLOW_PALIGNR 0x4000000 /* such as on the AMD Bobcat */
/* PowerPC */
#define X264_CPU_ALTIVEC 0x0000001
/* ARM and AArch64 */
#define X264_CPU_ARMV6 0x0000001
#define X264_CPU_NEON 0x0000002 /* ARM NEON */
#define X264_CPU_FAST_NEON_MRC 0x0000004 /* Transfer from NEON to ARM register is fast (Cortex-A9) */
#define X264_CPU_ARMV8 0x0000008
/* MIPS */
#define X264_CPU_MSA 0x0000001 /* MIPS MSA */
/* Analyse flags */
#define X264_ANALYSE_I4x4 0x0001 /* Analyse i4x4 */
#define X264_ANALYSE_I8x8 0x0002 /* Analyse i8x8 (requires 8x8 transform) */
#define X264_ANALYSE_PSUB16x16 0x0010 /* Analyse p16x8, p8x16 and p8x8 */
#define X264_ANALYSE_PSUB8x8 0x0020 /* Analyse p8x4, p4x8, p4x4 */
#define X264_ANALYSE_BSUB16x16 0x0100 /* Analyse b16x8, b8x16 and b8x8 */
#define X264_DIRECT_PRED_NONE 0 // 直接预测方法 direct predict
#define X264_DIRECT_PRED_SPATIAL 1 // 空间
#define X264_DIRECT_PRED_TEMPORAL 2 // 时间
#define X264_DIRECT_PRED_AUTO 3
#define X264_ME_DIA 0 //运动估计方法 motion estimation
#define X264_ME_HEX 1
#define X264_ME_UMH 2
#define X264_ME_ESA 3
#define X264_ME_TESA 4
#define X264_CQM_FLAT 0 //自定义量化矩阵方法 custom quant matrices
#define X264_CQM_JVT 1
#define X264_CQM_CUSTOM 2
#define X264_RC_CQP 0 // 码率控制方法 rate control
#define X264_RC_CRF 1
#define X264_RC_ABR 2
#define X264_QP_AUTO 0
#define X264_AQ_NONE 0 // adaptive quality
#define X264_AQ_VARIANCE 1
#define X264_AQ_AUTOVARIANCE 2
#define X264_AQ_AUTOVARIANCE_BIASED 3
#define X264_B_ADAPT_NONE 0 // 自适应B帧
#define X264_B_ADAPT_FAST 1
#define X264_B_ADAPT_TRELLIS 2
#define X264_WEIGHTP_NONE 0 // P帧权重
#define X264_WEIGHTP_SIMPLE 1
#define X264_WEIGHTP_SMART 2
#define X264_B_PYRAMID_NONE 0 // 参见param->i_bframe_pyramid ,B帧可作为参考帧
#define X264_B_PYRAMID_STRICT 1
#define X264_B_PYRAMID_NORMAL 2
#define X264_KEYINT_MIN_AUTO 0 // IDR帧间距param->i_keyint_max / i_keyint_min
#define X264_KEYINT_MAX_INFINITE (1<<30)
// 一些字符串式控制参数 或上面宏对应的字符串
static const char * const x264_direct_pred_names[] = { "none", "spatial", "temporal", "auto", 0 };
static const char * const x264_motion_est_names[] = { "dia", "hex", "umh", "esa", "tesa", 0 };
static const char * const x264_b_pyramid_names[] = { "none", "strict", "normal", 0 };
static const char * const x264_overscan_names[] = { "undef", "show", "crop", 0 };
static const char * const x264_vidformat_names[] = { "component", "pal", "ntsc", "secam", "mac", "undef", 0 };
static const char * const x264_fullrange_names[] = { "off", "on", 0 };
static const char * const x264_colorprim_names[] = { "", "bt709", "undef", "", "bt470m", "bt470bg", "smpte170m", "smpte240m", "film", "bt2020", "smpte428",
"smpte431", "smpte432", 0 };
static const char * const x264_transfer_names[] = { "", "bt709", "undef", "", "bt470m", "bt470bg", "smpte170m", "smpte240m", "linear", "log100", "log316",
"iec61966-2-4", "bt1361e", "iec61966-2-1", "bt2020-10", "bt2020-12", "smpte2084", "smpte428", 0 };
static const char * const x264_colmatrix_names[] = { "GBR", "bt709", "undef", "", "fcc", "bt470bg", "smpte170m", "smpte240m", "YCgCo", "bt2020nc", "bt2020c",
"smpte2085", 0 };
static const char * const x264_nal_hrd_names[] = { "none", "vbr", "cbr", 0 };
/* Colorspace type，色彩空间类型 */
#define X264_CSP_MASK 0x00ff /* */
#define X264_CSP_NONE 0x0000 /* Invalid mode */
#define X264_CSP_I420 0x0001 /* yuv 4:2:0 planar */
#define X264_CSP_YV12 0x0002 /* yvu 4:2:0 planar */
#define X264_CSP_NV12 0x0003 /* yuv 4:2:0, with one y plane and one packed u+v */
#define X264_CSP_NV21 0x0004 /* yuv 4:2:0, with one y plane and one packed v+u */
#define X264_CSP_I422 0x0005 /* yuv 4:2:2 planar */
#define X264_CSP_YV16 0x0006 /* yvu 4:2:2 planar */
#define X264_CSP_NV16 0x0007 /* yuv 4:2:2, with one y plane and one packed u+v */
#define X264_CSP_V210 0x0008 /* 10-bit yuv 4:2:2 packed in 32 */
#define X264_CSP_I444 0x0009 /* yuv 4:4:4 planar */
#define X264_CSP_YV24 0x000a /* yvu 4:4:4 planar */
#define X264_CSP_BGR 0x000b /* packed bgr 24bits */
#define X264_CSP_BGRA 0x000c /* packed bgr 32bits */
#define X264_CSP_RGB 0x000d /* packed rgb 24bits */
#define X264_CSP_MAX 0x000e /* end of list */
#define X264_CSP_VFLIP 0x1000 /* the csp is vertically flipped */
#define X264_CSP_HIGH_DEPTH 0x2000 /* the csp has a depth of 16 bits per pixel component */
/* Slice type，片类型，对应了上面nal_unit_type_e */
#define X264_TYPE_AUTO 0x0000 /* Let x264 choose the right type */
#define X264_TYPE_IDR 0x0001
#define X264_TYPE_I 0x0002
#define X264_TYPE_P 0x0003
#define X264_TYPE_BREF 0x0004 /* Non-disposable B-frame */
#define X264_TYPE_B 0x0005
#define X264_TYPE_KEYFRAME 0x0006 /* IDR or I depending on b_open_gop option */
#define IS_X264_TYPE_I(x) ((x)==X264_TYPE_I || (x)==X264_TYPE_IDR || (x)==X264_TYPE_KEYFRAME)
#define IS_X264_TYPE_B(x) ((x)==X264_TYPE_B || (x)==X264_TYPE_BREF)
/* Log level */
#define X264_LOG_NONE (-1)
#define X264_LOG_ERROR 0
#define X264_LOG_WARNING 1
#define X264_LOG_INFO 2
#define X264_LOG_DEBUG 3
/* Threading */
#define X264_THREADS_AUTO 0 /* Automatically select optimal number of threads */
#define X264_SYNC_LOOKAHEAD_AUTO (-1) /* Automatically select optimal lookahead thread buffer size */
/* HRD，hypothetical reference decoder (假定参考解码器),参见param->i_nal_hrd */
#define X264_NAL_HRD_NONE 0
#define X264_NAL_HRD_VBR 1 // variable bitrate
#define X264_NAL_HRD_CBR 2 // constant bitrate
/* Zones: override ratecontrol or other options for specific sections of the video.
* See x264_encoder_reconfig() for which options can be changed.
* If zones overlap, whichever comes later in the list takes precedence. */
typedef struct x264_zone_t
{
int i_start, i_end; /* range of frame numbers */
int b_force_qp; /* whether to use qp vs bitrate factor */
int i_qp;
float f_bitrate_factor;
struct x264_param_t *param;
} x264_zone_t;
/*
* 结构体详细介绍相关网址：
* http://www.docin.com/p-800561346.html //x264和xvid
* http://blog.csdn.net/zhubosa/article/details/51321780 //x264_param_t参数详细说明
* http://www.cnblogs.com/TaigaCon/p/5500110.html //关于去块滤波与PQ
*/
typedef struct x264_param_t
{
/* CPU flags */
unsigned int cpu;
int i_threads; /* encode multiple frames in parallel */
int i_lookahead_threads; /* multiple threads for lookahead analysis */
int b_sliced_threads; /* Whether to use slice-based threading. */
int b_deterministic; /* whether to allow non-deterministic optimizations when threaded */
int b_cpu_independent; /* force canonical behavior rather than cpu-dependent optimal algorithms */
int i_sync_lookahead; /* threaded lookahead buffer */
/* Video Properties */
int i_width;
int i_height;
int i_csp; /* CSP of encoded bitstream */
int i_level_idc; //decoder compatibility
/*取值范围10-51。设置比特流的Level。默认40，即4.0。用来告诉解码器需要支持的什么级别的
兼容性。只有在你知道自己在做什么的时候才设置该参数。*/
int i_frame_total; /* number of frames to encode if known, else 0 */
/* NAL HRD
* Uses Buffering and Picture Timing SEIs to signal HRD
* The HRD in H.264 was not designed with VFR in mind.
* It is therefore not recommendeded to use NAL HRD with VFR.
* Furthermore, reconfiguring the VBV (via x264_encoder_reconfig)
* will currently generate invalid HRD. */
/* hrd : hypothetical reference decoder (假定参考解码器) , 检验编码器产生的符合
该标准的NAL单元流或字节流的偏差值。蓝光视频、电视广播及其它特殊领域有此要求 */
int i_nal_hrd;
/* 视频可用性、视频标准化选项 */
struct
{
/* they will be reduced to be 0 < x <= 65535 and prime */
int i_sar_height;// 样本宽高比的高度
int i_sar_width;// 样本宽高比的宽度
int i_overscan; /* 0=undef, 1=no overscan, 2=overscan */
/* see h264 annex E for the values of the following */
int i_vidformat;
// 视频在编码/数字化之前是什么类型，默认"undef".
// 取值有：Component, PAL, NTSC, SECAM, MAC 等
int b_fullrange;
int i_colorprim;
int i_transfer;
int i_colmatrix;
int i_chroma_loc; /* both top & bottom */
} vui; //video useability information，视频可用性信息
/* Bitstream parameters,比特流参数 */
int i_frame_reference; /* Maximum number of reference frames,帧参考其他帧的最大数目 */
/*
B和P帧向前预测参考的帧数。取值范围1-16。该值不影响解码的速度，但是越大解码
所需的内存越大。这个值在一般情况下越大效果越好，但是超过6以后效果就 不明显了。
*/
int i_dpb_size; /* Force a DPB size larger than that implied by B-frames and reference frames.
* Useful in combination with interactive error resilience.
* dpb -> decoded picture buffer解码图像的缓冲区大小
*/
int i_keyint_max; /* Force an IDR keyframe at this interval, 设定IDR帧之间的最间隔， */
int i_keyint_min; /* Scenecuts closer together than this are coded as I, not IDR.
设定IDR帧之间的最小间隔, 场景切换小于此值编码位I帧, 而不是 IDR帧. */
int i_scenecut_threshold; /* how aggressively to insert extra I frames,场景切换阈值，插入I帧 */
/*
x264 为每个帧计算一个度量来评估它与它前面的帧有多大不同。假如值小于 scenecut 设定值，
意味着发生场景切换。如果此帧与前面最后一个 IDR 帧间隔小于 --min-keyint 就放置一个 I 帧，
否则就放置一个 IDR 帧。过大的 scenecut 值会导致产生大量的场景切换。
默认40，scencecut 设置为0时相当于--no-scenecut。
*/
int b_intra_refresh; /* Whether or not to use periodic intra refresh instead of IDR frames.
是否使用周期帧内刷新替代IDR帧，默认为false */
int i_bframe; /* how many b-frame between 2 references pictures，b帧最大连续数目 */
int i_bframe_adaptive;// 自适应B帧判定, 可选取值：X264_B_ADAPT_FAST等,默认开启
//允许编码器根据情况自动调整是否用B帧，
//如果为true，则最大连续数目可达到i_bframe，但可能小于此值
//如果为false，那么参考帧之间B帧数目一定是i_bframe（测试发现，当i_bframe设置为20，实际上得到的文件i_bframe是16，可能也跟文件有关）
int i_bframe_bias; // 控制B帧替代P帧的概率，范围-100 ~ +100，
// 该值越高越容易插入B帧，默认0.
//负值倾向少用B帧，正值倾向使用B帧，值大小作为衡量
int i_bframe_pyramid; /* 允许B帧作为参考帧Keep some B-frames as references: 0=off, 1=strict hierarchical, 2=normal */
//当连续出现2个或以上B帧时，能够增加压缩率,视频质量略微提高，但编码速度不受影响
// 当此值为非off时，编码才能出现B帧，同时Baseline模式不支持B帧
int b_open_gop; // Close GOP是指帧间的预测都是在GOP中进行的。
// 使用Open GOP，后一个GOP会参考前一个GOP的信息
// GOP被定义为IDR帧之间的距离
int b_bluray_compat; // 支持蓝光碟
int i_avcintra_class;
/* 去块滤波参数，概念http://www.cnblogs.com/TaigaCon/p/5500110.html 关于去块滤波与QP */
// 由于在DCT变换中产生了锯齿
/* 去块滤波器需要的参数, alpha和beta是去块滤波器参数 */
//在编码器中，滤波后的图像会作为后续编码运动补偿的参考图像；
//在解码器中，滤波后的图像会被输出显示并且作为后续图像解码重建的参考图像。
int b_deblocking_filter; // 去块滤波开关
int i_deblocking_filter_alphac0; /* [-6, 6] -6 light filter, 6 strong */
int i_deblocking_filter_beta; /* [-6, 6] idem */
int b_cabac; // 自适应算术编码cabac开关
//使用CABAC熵编码技术,为引起轻微的编码和解码的速度损失，但是可以提高10%-15%的编码质量。
/*
* b_cabac 打开后，才能将Adaptive DCT设置为 8*8 DCT,
* I4*4/P4*4/I8*8/P8*8/B8*8 ,avc标准允许的DCT块划分方式，前面字母代表帧类型，数字代表块大小
* 貌似 8*8 对画面质量和亚索效率都由好处
*/
int i_cabac_init_idc; // 给出算术编码初始化时表格的选择
/*
* CABAC 是H.264/AVC标准中两种熵编码方法中的一种，是将自适应的二进制算术编码
* 与一个设计精良的上下文模型结合起来得到的方法
*/
int b_interlaced; // 隔行扫描
int b_constrained_intra; //强行分区
/* 量化 */
int i_cqm_preset; // constant quant macro （恒定量化宏块）
char *psz_cqm_file; /* filename (in UTF-8) of CQM file, JM format */
uint8_t cqm_4iy[16]; /* used only if i_cqm_preset == X264_CQM_CUSTOM */
uint8_t cqm_4py[16];
uint8_t cqm_4ic[16];
uint8_t cqm_4pc[16];
uint8_t cqm_8iy[64];
uint8_t cqm_8py[64];
uint8_t cqm_8ic[64];
uint8_t cqm_8pc[64];
/* Log，日志 */
void (*pf_log)( void *, int i_level, const char *psz, va_list );
void *p_log_private;
int i_log_level;
int b_full_recon; /* fully reconstruct frames, even when not necessary for encoding. Implied by psz_dump_yuv */
char *psz_dump_yuv; /* filename (in UTF-8) for reconstructed frames */
/* Encoder analyser parameters，编码分析参数 */
struct
{
unsigned int intra; /* intra partitions,帧内分区 */
unsigned int inter; /* inter partitions，帧间分区 */
/*
X264_ANALYSE_I4x4
X264_ANALYSE_I8x8
X264_ANALYSE_PSUB16x16
X264_ANALYSE_PSUB8x8
X264_ANALYSE_BSUB16x16
*/
int b_transform_8x8; //是否使用8*8 DCT
int i_weighted_pred; /* weighting for P-frames */
int b_weighted_bipred; /* implicit weighting for B-frames,当B帧设置大于1使用 */
int i_direct_mv_pred; /* spatial vs temporal mv prediction */
// 时间空间运动向量预测模式
/*
B帧采用的运动侦测的方式,设置为auto质量会好一些，但是速度会下降一些，设置为0
，质量和速度都会下降.可以选择none, auto, temporal, spatial.
*/
int i_chroma_qp_offset; // 色度量化步长偏移量
// me -> motion estimation(运动估计)
int i_me_method; /* motion estimation algorithm to use (X264_ME_*) */
// 运动估计算法
int i_me_range; /* integer pixel motion estimation search range (from predicted mv) */
// 整像素运动估计搜索范围 (from predicted mv)
// mv -> motion vector (运动矢量)
int i_mv_range; /* maximum length of a mv (in pixels). -1 = auto, based on level */
int i_mv_range_thread; /* minimum space between threads. -1 = auto, based on number of threads. */
// 线程之间的最小运动向量缓冲
int i_subpel_refine; /* subpixel motion estimation quality */
// 亚像素运动估计质量,这个参数控制在运动估算过程中质量和速度的权衡。
int b_chroma_me; /* chroma ME for subpel and mode decision in P-frames */
// 是否开启 亚像素运动估计 和 P帧模式选择
int b_mixed_references; /* allow each mb partition to have its own reference number */
// 是否允许每个宏块的分区有它自己的参考号
//也就是允许8*8,16*16运动块独立地选择参考帧，如果disable,则所有的宏块必须参考同一帧。
int i_trellis; /* trellis RD quantization */
// Trellis量化提高效率，对每个8x8的块寻找合适的量化值，需要CABAC，
// 0 ：即关闭 1：只在最后编码时使用 2：在所有模式决策上启用
int b_fast_pskip; /* early SKIP detection on P-frames */
// 是否P帧提早跳过检测,这个经常在没有任何损失的前提下提高了速度。
int b_dct_decimate; /* transform coefficient thresholding on P-frames */
//P帧变换系数阈值
int i_noise_reduction; /* adaptive pseudo-deadzone */
//自适应位盲区，减少噪声0意味着关闭，对于噪声很大的内容你需要打开。范围：0-100000
float f_psy_rd; /* Psy(Psychovisual精神性视觉?) RD(rate dropped?码率丢失) strength */
// Psy RD强度
float f_psy_trellis; /* Psy trellis(格子，框架) strength */
// Psy Trellis强度
int b_psy; /* Toggle all psy optimizations */
//；Psy优化开关，可能会增强细节
int b_mb_info; /* Use input mb_info data in x264_picture_t */
int b_mb_info_update; /* Update the values in mb_info according to the results of encoding. */
/* the deadzone size that will be used in luma quantization */
int i_luma_deadzone[2]; /* {inter, intra} */
// 允许计算和打印峰值信噪比
// 它是原图像与被处理图像之间的均方误差相对于(2^n-1)^2的对数值(信号最大值的平方，n是每个采样值的比特数)，它的单位是dB
// PSNR值越大，就代表失真越少。
int b_psnr; /* compute and print PSNR(Peak signal-to-noise ratio 峰值信噪比) stats */
// 允许计算和打印结构相似法,SSIM是一种衡量两幅图像相似度的指标,结构相似性的范围为[-1,1]。当两张图像一模一样时，SSIM的值等于1。
int b_ssim; /* compute and print SSIM（Structural Similarity结构相似法） stats */
} analyse;
/* Rate control parameters，码率控制参数 */
struct
{ // DCT变换 和 量化 均是为了压缩，降低码率
int i_rc_method; /* X264_RC_* */
// 码率控制方式 ： X264_RC_CQP恒定质量(恒定量化？),
// X264_RC_CRF恒定码率(constant rate factor), X264_RC_ABR平均码率（average bitrate）
//优先级是ABR > CQP > CRF. 缺省是CRF
//量化来消除图像中的相关性以减少图像编码的动态范围
// quant（量化） p frame？
int i_qp_constant; /* 指定量化值(QP value)，0 to (51 + 6*(x264_bit_depth-8)). 0=lossless（无损） */
//固定量化因子。取值范围0到51。经常取值在20-40之间，越小质量越好,要求的码率越高。0表示无损压缩
int i_qp_min; /* 允许的最小量化值，默认10，min allowed QP value */
int i_qp_max; /* 允许的最大量化值，默认51，max allowed QP value */
int i_qp_step; /* 量化步长，即相邻两帧之间量化值之差的最大值，max QP step between frames */
int i_bitrate; // 平均码率大小,单位？
float f_rf_constant; /* (rate factor=码率因子)实际质量，值越大图像越花,越小越清晰，1pass VBR, nominal QP */
float f_rf_constant_max; /* In CRF(constant rate factor=恒定码率因子) mode, maximum CRF as caused by VBV */
// 最大码率因子，该选项仅在使用CRF并开启VBV时有效，
// 图像质量的最大值，可能会导致VBV下溢。
float f_rate_tolerance; // 码率允许的误差
/*
* VBV (Video Buffering Verifier,视频缓存检验区)它作用是处理各帧编码后大小不一和恒定输出
* 码率的矛盾。可以将VBV想象成为一个水池，水池的入口连接着encoder的输出，出口为恒定码率的网
* 络输出。为了使得输出恒定，encoder必须保证水池既不上溢也不下溢。下溢会导致无数据输出，上溢
* 会导致数据丢失。所以encoder在编码一帧时会参考当前vbv的充盈情况，并由此计算出，当前帧应当编
* 码出多少比特，从而保证既不上溢（增加QP）也不下溢（减少QP）。为了防止上溢，VBV会要求降低编
* 码比特，encoder会采用更大一些的QP，这样必然降低了视频序列质量，因此只在真正需要的时候才设定
* 它。一般在硬件解码端的时候，一定要做这个设置。
*/
int i_vbv_max_bitrate; //平均码率模式下，最大瞬时码率,默认0单位是kbps(ffmpeg里的rc_max_rate单位是bps)b->bit
int i_vbv_buffer_size; //设置VBV可用的最大缓冲区，单位是kbits，默认0
float f_vbv_buffer_init; /* <=1: fraction of buffer_size. >1: kbit */
// 设置码率控制缓冲区（VBV）缓冲达到多满(百分比)，才开始回放，
// 范围0~1.0，默认0.9
float f_ip_factor;// I帧和P帧之间的量化因子（QP）比值(ratio)，默认1.4
float f_pb_factor; // P帧和B帧之间的量化因子（QP）比值(ratio)，默认1.3
/* VBV filler: force CBR VBV and use filler bytes to ensure hard-CBR.
* Implied by NAL-HRD CBR. */
int b_filler;
// aq -> adaptive quant = 自适应量化
int i_aq_mode; /* psy adaptive QP. (X264_AQ_*) */
// 自适应量化（AQ）模式。 0：关闭AQ
// 1：允许AQ在整个视频中和帧内重新分配码
// 2：自方差AQ(实验阶段)，尝试逐帧调整强度
float f_aq_strength; // AQ强度，减少平趟和纹理区域的块效应和模糊度
int b_mb_tree; /* Macroblock-tree ratecontrol. */
// 是否开启基于macroblock的qp控制方法，开启之后质量有所提升
int i_lookahead; // 决定mbtree向前预测的帧数
/* 2pass */
int b_stat_write; /* Enable stat writing in psz_stat_out */
// 是否将统计数据写入到文件psz_stat_out中
char *psz_stat_out; /* output filename (in UTF-8) of the 2pass stats file */
// 输出文件用于保存第一次编码统计数据
int b_stat_read; /* Read stat from psz_stat_in and use it */
// 是否从文件psz_stat_in中读入统计数据
char *psz_stat_in; /* input filename (in UTF-8) of the 2pass stats file */
// 输入文件存有第一次编码的统计数据
/* 2pass params (same as ffmpeg ones) */
float f_qcompress; /* 量化曲线(quantizer curve)压缩因子。范围0-1
0.0 => cbr(constant bitrate), 1.0 => constant qp(constatnt quantity(p frame))
越小-》码率越恒定，越大-》量化越恒定
*/
float f_qblur; /* temporally blur quants */
// 时间上模糊量化，减少QP的波动(after curve compression)
float f_complexity_blur; /* temporally blur complexity */
// 时间上模糊复杂性，减少QP的波动(before curve compression)
x264_zone_t *zones; /* ratecontrol overrides */
// 码率控制覆盖
int i_zones; /* number of zone_t's */
char *psz_zones; /* alternate method of specifying zones */
} rc;
/* Cropping Rectangle parameters: added to those implicitly defined by
non-mod16 video resolutions. */
struct
{
unsigned int i_left;
unsigned int i_top;
unsigned int i_right;
unsigned int i_bottom;
} crop_rect; //剪裁矩形
/* frame packing arrangement flag */
int i_frame_packing;
/* Muxing parameters */
int b_aud; /* generate access unit delimiters */
// 生成访问单元分隔符
int b_repeat_headers; /* put SPS/PPS before each keyframe */
// 是否复制sps和pps放在每个关键帧的前面
int b_annexb; /* if set, place start codes (4 bytes) before NAL units,
* otherwise place size (4 bytes) before NAL units. */
// 值为true，则NALU之前是4字节前缀码0x00000001；
// 值为false，则NALU之前的4个字节为NALU长度
int i_sps_id; /* SPS and PPS id number */
int b_vfr_input; /* VFR input. If 1, use timebase and timestamps for ratecontrol purposes.
* If 0, use fps only. */
// VFR输入。1 ：时间基和时间戳用于码率控制 0 ：仅帧率用于码率控制
// 是否对恒定帧率（CFR）设置明确的时间基
int b_pulldown; /* use explicity set timebase for CFR */
/* 以某个预设模式将输入流(隔行，恒定帧率)标记为软交错(soft telecine)。默认none. 可用预设有：
none, 22, 32, 64, double, triple, euro. 使用除none以外任一预设，都会连带开启--pic-struct */
uint32_t i_fps_num; // 帧率分子
uint32_t i_fps_den; // 帧率分母
uint32_t i_timebase_num; /* Timebase numerator */ //fps两两互换，但为了提高时间基精度，会乘以1000
uint32_t i_timebase_den; /* Timebase denominator */
int b_tff;
/* Pulldown:
* The correct pic_struct must be passed with each input frame.
* The input timebase should be the timebase corresponding to the output framerate. This should be constant.
* e.g. for 3:2 pulldown timebase should be 1001/30000
* The PTS passed with each frame must be the PTS of the frame after pulldown is applied.
* Frame doubling and tripling require b_vfr_input set to zero (see H.264 Table D-1)
*
* Pulldown changes are not clearly defined in H.264. Therefore, it is the calling app's responsibility to manage this.
*/
int b_pic_struct;
/* Fake Interlaced.
*
* Used only when b_interlaced=0. Setting this flag makes it possible to flag the stream as PAFF interlaced yet
* encode all frames progessively. It is useful for encoding 25p and 30p Blu-Ray streams.
*/
/* 将视频流标记为交错(隔行)，哪怕并非为交错式编码。可用于编码蓝光兼容的25p和30p视频。默认是未开启 */
int b_fake_interlaced;
/* Don't optimize header parameters based on video content, e.g. ensure that splitting an input video, compressing
* each part, and stitching them back together will result in identical SPS/PPS. This is necessary for stitching
* with container formats that don't allow multiple SPS/PPS. */
int b_stitchable;
// 开放运算语言
int b_opencl; /* use OpenCL when available */
int i_opencl_device; /* specify count of GPU devices to skip, for CLI users */
void *opencl_device_id; /* pass explicit cl_device_id as void*, for API users */
char *psz_clbin_file; /* filename (in UTF-8) of the compiled OpenCL kernel cache file */
/* Slicing parameters */
int i_slice_max_size; /* Max size per slice in bytes; includes estimated NAL overhead. */
int i_slice_max_mbs; /* Max number of MBs per slice; overrides i_slice_count. */
int i_slice_min_mbs; /* Min number of MBs per slice */
int i_slice_count; /* Number of slices per frame: forces rectangular slices. */
int i_slice_count_max; /* Absolute cap on slices per frame; stops applying slice-max-size
* and slice-max-mbs if this is reached. */
/* Optional callback for freeing this x264_param_t when it is done being used.
* Only used when the x264_param_t sits in memory for an indefinite period of time,
* i.e. when an x264_param_t is passed to x264_t in an x264_picture_t or in zones.
* Not used when x264_encoder_reconfig is called directly. */
void (*param_free)( void* );
/* Optional low-level callback for low-latency encoding. Called for each output NAL unit
* immediately after the NAL unit is finished encoding. This allows the calling application
* to begin processing video data (e.g. by sending packets over a network) before the frame
* is done encoding.
*
* This callback MUST do the following in order to work correctly:
* 1) Have available an output buffer of at least size nal->i_payload*3/2 + 5 + 64.
* 2) Call x264_nal_encode( h, dst, nal ), where dst is the output buffer.
* After these steps, the content of nal is valid and can be used in the same way as if
* the NAL unit were output by x264_encoder_encode.
*
* This does not need to be synchronous with the encoding process: the data pointed to
* by nal (both before and after x264_nal_encode) will remain valid until the next
* x264_encoder_encode call. The callback must be re-entrant.
*
* This callback does not work with frame-based threads; threads must be disabled
* or sliced-threads enabled. This callback also does not work as one would expect
* with HRD -- since the buffering period SEI cannot be calculated until the frame
* is finished encoding, it will not be sent via this callback.
*
* Note also that the NALs are not necessarily returned in order when sliced threads is
* enabled. Accordingly, the variable i_first_mb and i_last_mb are available in
* x264_nal_t to help the calling application reorder the slices if necessary.
*
* When this callback is enabled, x264_encoder_encode does not return valid NALs;
* the calling application is expected to acquire all output NALs through the callback.
*
* It is generally sensible to combine this callback with a use of slice-max-mbs or
* slice-max-size.
*
* The opaque pointer is the opaque pointer from the input frame associated with this
* NAL unit. This helps distinguish between nalu_process calls from different sources,
* e.g. if doing multiple encodes in one process.
*/
void (*nalu_process) ( x264_t *h, x264_nal_t *nal, void *opaque );
} x264_param_t;
void x264_nal_encode( x264_t *h, uint8_t *dst, x264_nal_t *nal );
/****************************************************************************
* H.264 level restriction information
****************************************************************************/
// http://blog.csdn.net/sphone89/article/details/17492433 profile&level 介绍
//在给定的profile下，level通常与解码器的处理能力和内存容量相对应。每一个档次设置不同的参数
//（如取样速率、图像尺寸、编码比特率等），得到对应的编解码器性能的不同level。
typedef struct x264_level_t
{
int level_idc;
int mbps; /* max macroblock processing rate (macroblocks/sec) */
int frame_size; /* max frame size (macroblocks) */
int dpb; /* max decoded picture buffer (mbs) */
int bitrate; /* max bitrate (kbit/sec) */
int cpb; /* max vbv buffer (kbit) */
int mv_range; /* max vertical mv component range (pixels) */
int mvs_per_2mb; /* max mvs per 2 consecutive mbs. */
int slice_rate; /* ?? */
int mincr; /* min compression ratio */
int bipred8x8; /* limit bipred to >=8x8 */
int direct8x8; /* limit b_direct to >=8x8 */
int frame_only; /* forbid interlacing */
} x264_level_t;
/* all of the levels defined in the standard, terminated by .level_idc=0 */
X264_API extern const x264_level_t x264_levels[];
/****************************************************************************
* Basic parameter handling functions
****************************************************************************/
/* x264_param_default:
* fill x264_param_t with default values and do CPU detection */
void x264_param_default( x264_param_t * );
/* x264_param_parse:
* set one parameter by name.
* returns 0 on success, or returns one of the following errors.
* note: BAD_VALUE occurs only if it can't even parse the value,
* numerical range is not checked until x264_encoder_open() or
* x264_encoder_reconfig().
* value=NULL means "true" for boolean options, but is a BAD_VALUE for non-booleans. */
#define X264_PARAM_BAD_NAME (-1)
#define X264_PARAM_BAD_VALUE (-2)
int x264_param_parse( x264_param_t *, const char *name, const char *value );
/****************************************************************************
* Advanced parameter handling functions
****************************************************************************/
/* These functions expose the full power of x264's preset-tune-profile system for
* easy adjustment of large numbers of internal parameters.
*
* In order to replicate x264CLI's option handling, these functions MUST be called
* in the following order:
* 1) x264_param_default_preset
* 2) Custom user options (via param_parse or directly assigned variables)
* 3) x264_param_apply_fastfirstpass
* 4) x264_param_apply_profile
*
* Additionally, x264CLI does not apply step 3 if the preset chosen is "placebo"
* or --slow-firstpass is set. */
/* x264_param_default_preset:
* The same as x264_param_default, but also use the passed preset and tune
* to modify the default settings.
* (either can be NULL, which implies no preset or no tune, respectively)
* Currently available presets are, ordered from fastest to slowest:
* preset 的参数主要调节编码速度和视频质量的平衡，ultrafast 速度是 placebo 的100倍。 */
static const char * const x264_preset_names[] = { "ultrafast", "superfast", "veryfast", "faster", "fast", "medium", "slow", "slower", "veryslow", "placebo", 0 };
/* The presets can also be indexed numerically, as in:
* x264_param_default_preset( ¶m, "3", ... )
* with ultrafast mapping to "0" and placebo mapping to "9". This mapping may
* of course change if new presets are added in between, but will always be
* ordered from fastest to slowest.
*
* Warning: the speed of these presets scales dramatically. Ultrafast is a full
* 100 times faster than placebo!
* Currently available tunings are:
* --tune的参数主要配合视频类型和视觉优化的参数，或特别的情况。
film： 电影、真人类型；
animation： 动画；
grain： 需要保留大量的grain时用；
stillimage： 静态图像编码时使用；
psnr： 为提高psnr做了优化的参数；
ssim： 为提高ssim做了优化的参数；
fastdecode： 可以快速解码的参数；
zerolatency：零延迟，用在需要非常低的延迟的情况下，比如电视电话会议的编码。 */
static const char * const x264_tune_names[] = { "film", "animation", "grain", "stillimage", "psnr", "ssim", "fastdecode", "zerolatency", 0 };
/* Multiple tunings can be used if separated by a delimiter in ",./-+",
* however multiple psy tunings cannot be used.
* film, animation, grain, stillimage, psnr, and ssim are psy tunings.
*
* returns 0 on success, negative on failure (e.g. invalid preset/tune name). */
int x264_param_default_preset( x264_param_t *, const char *preset, const char *tune );
/* x264_param_apply_fastfirstpass:
* If first-pass mode is set (rc.b_stat_read == 0, rc.b_stat_write == 1),
* modify the encoder settings to disable options generally not useful on
* the first pass. */
void x264_param_apply_fastfirstpass( x264_param_t * );
/* x264_param_apply_profile:
* Applies the restrictions of the given profile.
* Currently available profiles are, from most to least restrictive: */
// http://blog.csdn.net/sphone89/article/details/17492433 profile&level 介绍
// 每个profile支持一组特定的编码功能，并支持一类特定的应用,分别是BP、EP、MP、HP：
static const char * const x264_profile_names[] = { "baseline", "main", "high", "high10", "high422", "high444", 0 };
/* (can be NULL, in which case the function will do nothing)
*
* Does NOT guarantee that the given profile will be used: if the restrictions
* of "High" are applied to settings that are already Baseline-compatible, the
* stream will remain baseline. In short, it does not increase settings, only
* decrease them.
*
* returns 0 on success, negative on failure (e.g. invalid profile name). */
int x264_param_apply_profile( x264_param_t *, const char *profile );
/****************************************************************************
* Picture structures and functions
****************************************************************************/
/* x264_bit_depth:位深
* Specifies the number of bits per pixel that x264 uses. This is also the
* bit depth that x264 encodes in. If this value is > 8, x264 will read
* two bytes of input data for each pixel sample, and expect the upper
* (16-x264_bit_depth) bits to be zero.
* Note: The flag X264_CSP_HIGH_DEPTH must be used to specify the
* colorspace depth as well. */
X264_API extern const int x264_bit_depth;
/* x264_chroma_format:色彩格式
* Specifies the chroma formats that x264 supports encoding. When this
* value is non-zero, then it represents a X264_CSP_* that is the only
* chroma format that x264 supports encoding. If the value is 0 then
* there are no restrictions. */
X264_API extern const int x264_chroma_format;
enum pic_struct_e
{
PIC_STRUCT_AUTO = 0, // automatically decide (default)
PIC_STRUCT_PROGRESSIVE = 1, // progressive frame
// "TOP" and "BOTTOM" are not supported in x264 (PAFF only)
PIC_STRUCT_TOP_BOTTOM = 4, // top field followed by bottom
PIC_STRUCT_BOTTOM_TOP = 5, // bottom field followed by top
PIC_STRUCT_TOP_BOTTOM_TOP = 6, // top field, bottom field, top field repeated
PIC_STRUCT_BOTTOM_TOP_BOTTOM = 7, // bottom field, top field, bottom field repeated
PIC_STRUCT_DOUBLE = 8, // double frame
PIC_STRUCT_TRIPLE = 9, // triple frame
};
typedef struct x264_hrd_t
{
double cpb_initial_arrival_time;
double cpb_final_arrival_time;
double cpb_removal_time;
double dpb_output_time;
} x264_hrd_t;
/* Arbitrary user SEI:
* Payload size is in bytes and the payload pointer must be valid.
* Payload types and syntax can be found in Annex D of the H.264 Specification.
* SEI payload alignment bits as described in Annex D must be included at the
* end of the payload if needed.
* The payload should not be NAL-encapsulated.
* Payloads are written first in order of input, apart from in the case when HRD
* is enabled where payloads are written after the Buffering Period SEI. */
typedef struct x264_sei_payload_t
{
int payload_size;
int payload_type;
uint8_t *payload;
} x264_sei_payload_t;
typedef struct x264_sei_t
{
int num_payloads;
x264_sei_payload_t *payloads;
/* In: optional callback to free each payload AND x264_sei_payload_t when used. */
void (*sei_free)( void* );
} x264_sei_t;
typedef struct x264_image_t
{
int i_csp; /* Colorspace */
int i_plane; /* Number of image planes */
int i_stride[4]; /* Strides for each plane */
uint8_t *plane[4]; /* Pointers to each plane */
} x264_image_t;
typedef struct x264_image_properties_t
{
/* All arrays of data here are ordered as follows:
* each array contains one offset per macroblock, in raster scan order. In interlaced
* mode, top-field MBs and bottom-field MBs are interleaved at the row level.
* Macroblocks are 16x16 blocks of pixels (with respect to the luma plane). For the
* purposes of calculating the number of macroblocks, width and height are rounded up to
* the nearest 16. If in interlaced mode, height is rounded up to the nearest 32 instead. */
/* In: an array of quantizer offsets to be applied to this image during encoding.
* These are added on top of the decisions made by x264.
* Offsets can be fractional; they are added before QPs are rounded to integer.
* Adaptive quantization must be enabled to use this feature. Behavior if quant
* offsets differ between encoding passes is undefined. */
float *quant_offsets;
/* In: optional callback to free quant_offsets when used.
* Useful if one wants to use a different quant_offset array for each frame. */
void (*quant_offsets_free)( void* );
/* In: optional array of flags for each macroblock.
* Allows specifying additional information for the encoder such as which macroblocks
* remain unchanged. Usable flags are listed below.
* x264_param_t.analyse.b_mb_info must be set to use this, since x264 needs to track
* extra data internally to make full use of this information.
*
* Out: if b_mb_info_update is set, x264 will update this array as a result of encoding.
*
* For "MBINFO_CONSTANT", it will remove this flag on any macroblock whose decoded
* pixels have changed. This can be useful for e.g. noting which areas of the
* frame need to actually be blitted. Note: this intentionally ignores the effects
* of deblocking for the current frame, which should be fine unless one needs exact
* pixel-perfect accuracy.
*
* Results for MBINFO_CONSTANT are currently only set for P-frames, and are not
* guaranteed to enumerate all blocks which haven't changed. (There may be false
* negatives, but no false positives.)
*/
uint8_t *mb_info;
/* In: optional callback to free mb_info when used. */
void (*mb_info_free)( void* );
/* The macroblock is constant and remains unchanged from the previous frame. */
#define X264_MBINFO_CONSTANT (1<<0)
/* More flags may be added in the future. */
/* Out: SSIM of the the frame luma (if x264_param_t.b_ssim is set) */
double f_ssim;
/* Out: Average PSNR of the frame (if x264_param_t.b_psnr is set) */
double f_psnr_avg;
/* Out: PSNR of Y, U, and V (if x264_param_t.b_psnr is set) */
double f_psnr[3];
/* Out: Average effective CRF of the encoded frame */
double f_crf_avg;
} x264_image_properties_t;
typedef struct x264_picture_t
{
/* In: force picture type (if not auto)
* If x264 encoding parameters are violated in the forcing of picture types,
* x264 will correct the input picture type and log a warning.
* Out: type of the picture encoded */
int i_type;
/* In: force quantizer for != X264_QP_AUTO */
int i_qpplus1;
/* In: pic_struct, for pulldown/doubling/etc...used only if b_pic_struct=1.
* use pic_struct_e for pic_struct inputs
* Out: pic_struct element associated with frame */
int i_pic_struct;
/* Out: whether this frame is a keyframe. Important when using modes that result in
* SEI recovery points being used instead of IDR frames. */
int b_keyframe;
/* In: user pts, Out: pts of encoded picture (user)*/
int64_t i_pts;
/* Out: frame dts. When the pts of the first frame is close to zero,
* initial frames may have a negative dts which must be dealt with by any muxer */
int64_t i_dts;
/* In: custom encoding parameters to be set from this frame forwards
(in coded order, not display order). If NULL, continue using
parameters from the previous frame. Some parameters, such as
aspect ratio, can only be changed per-GOP due to the limitations
of H.264 itself; in this case, the caller must force an IDR frame
if it needs the changed parameter to apply immediately. */
x264_param_t *param;
/* In: raw image data */
/* Out: reconstructed image data. x264 may skip part of the reconstruction process,
e.g. deblocking, in frames where it isn't necessary. To force complete
reconstruction, at a small speed cost, set b_full_recon. */
x264_image_t img;
/* In: optional information to modify encoder decisions for this frame
* Out: information about the encoded frame */
x264_image_properties_t prop;
/* Out: HRD timing information. Output only when i_nal_hrd is set. */
x264_hrd_t hrd_timing;
/* In: arbitrary user SEI (e.g subtitles, AFDs) */
x264_sei_t extra_sei;
/* private user data. copied from input to output frames. */
void *opaque;
} x264_picture_t;
/* x264_picture_init:
* initialize an x264_picture_t. Needs to be done if the calling application
* allocates its own x264_picture_t as opposed to using x264_picture_alloc. */
void x264_picture_init( x264_picture_t *pic );
/* x264_picture_alloc:
* alloc data for a picture. You must call x264_picture_clean on it.
* returns 0 on success, or -1 on malloc failure or invalid colorspace. */
int x264_picture_alloc( x264_picture_t *pic, int i_csp, int i_width, int i_height );
/* x264_picture_clean:
* free associated resource for a x264_picture_t allocated with
* x264_picture_alloc ONLY */
void x264_picture_clean( x264_picture_t *pic );
/****************************************************************************
* Encoder functions
****************************************************************************/
/* Force a link error in the case of linking against an incompatible API version.
* Glue #defines exist to force correct macro expansion; the final output of the macro
* is x264_encoder_open_##X264_BUILD (for purposes of dlopen). */
#define x264_encoder_glue1(x,y) x##y
#define x264_encoder_glue2(x,y) x264_encoder_glue1(x,y)
#define x264_encoder_open x264_encoder_glue2(x264_encoder_open_,X264_BUILD)
/* x264_encoder_open:
* create a new encoder handler, all parameters from x264_param_t are copied */
x264_t *x264_encoder_open( x264_param_t * );
/* x264_encoder_reconfig:
* various parameters from x264_param_t are copied.
* this takes effect immediately, on whichever frame is encoded next;
* due to delay, this may not be the next frame passed to encoder_encode.
* if the change should apply to some particular frame, use x264_picture_t->param instead.
* returns 0 on success, negative on parameter validation error.
* not all parameters can be changed; see the actual function for a detailed breakdown.
*
* since not all parameters can be changed, moving from preset to preset may not always
* fully copy all relevant parameters, but should still work usably in practice. however,
* more so than for other presets, many of the speed shortcuts used in ultrafast cannot be
* switched out of; using reconfig to switch between ultrafast and other presets is not
* recommended without a more fine-grained breakdown of parameters to take this into account. */
int x264_encoder_reconfig( x264_t *, x264_param_t * );
/* x264_encoder_parameters:
* copies the current internal set of parameters to the pointer provided
* by the caller. useful when the calling application needs to know
* how x264_encoder_open has changed the parameters, or the current state
* of the encoder after multiple x264_encoder_reconfig calls.
* note that the data accessible through pointers in the returned param struct
* (e.g. filenames) should not be modified by the calling application. */
void x264_encoder_parameters( x264_t *, x264_param_t * );
/* x264_encoder_headers:
* return the SPS and PPS that will be used for the whole stream.
* *pi_nal is the number of NAL units outputted in pp_nal.
* returns the number of bytes in the returned NALs.
* returns negative on error.
* the payloads of all output NALs are guaranteed to be sequential in memory. */
int x264_encoder_headers( x264_t *, x264_nal_t **pp_nal, int *pi_nal );
/* x264_encoder_encode:
* encode one picture.
* *pi_nal is the number of NAL units outputted in pp_nal.
* returns the number of bytes in the returned NALs.
* returns negative on error and zero if no NAL units returned.
* the payloads of all output NALs are guaranteed to be sequential in memory. */
int x264_encoder_encode( x264_t *, x264_nal_t **pp_nal, int *pi_nal, x264_picture_t *pic_in, x264_picture_t *pic_out );
/* x264_encoder_close:
* close an encoder handler */
void x264_encoder_close ( x264_t * );
/* x264_encoder_delayed_frames:
* return the number of currently delayed (buffered) frames
* this should be used at the end of the stream, to know when you have all the encoded frames. */
int x264_encoder_delayed_frames( x264_t * );
/* x264_encoder_maximum_delayed_frames( x264_t *h ):
* return the maximum number of delayed (buffered) frames that can occur with the current
* parameters. */
int x264_encoder_maximum_delayed_frames( x264_t *h );
/* x264_encoder_intra_refresh:
* If an intra refresh is not in progress, begin one with the next P-frame.
* If an intra refresh is in progress, begin one as soon as the current one finishes.
* Requires that b_intra_refresh be set.
*
* Useful for interactive streaming where the client can tell the server that packet loss has
* occurred. In this case, keyint can be set to an extremely high value so that intra refreshes
* only occur when calling x264_encoder_intra_refresh.
*
* In multi-pass encoding, if x264_encoder_intra_refresh is called differently in each pass,
* behavior is undefined.
*
* Should not be called during an x264_encoder_encode. */
void x264_encoder_intra_refresh( x264_t * );
/* x264_encoder_invalidate_reference:
* An interactive error resilience tool, designed for use in a low-latency one-encoder-few-clients
* system. When the client has packet loss or otherwise incorrectly decodes a frame, the encoder
* can be told with this command to "forget" the frame and all frames that depend on it, referencing
* only frames that occurred before the loss. This will force a keyframe if no frames are left to
* reference after the aforementioned "forgetting".
*
* It is strongly recommended to use a large i_dpb_size in this case, which allows the encoder to
* keep around extra, older frames to fall back on in case more recent frames are all invalidated.
* Unlike increasing i_frame_reference, this does not increase the number of frames used for motion
* estimation and thus has no speed impact. It is also recommended to set a very large keyframe
* interval, so that keyframes are not used except as necessary for error recovery.
*
* x264_encoder_invalidate_reference is not currently compatible with the use of B-frames or intra
* refresh.
*
* In multi-pass encoding, if x264_encoder_invalidate_reference is called differently in each pass,
* behavior is undefined.
*
* Should not be called during an x264_encoder_encode, but multiple calls can be made simultaneously.
*
* Returns 0 on success, negative on failure. */
int x264_encoder_invalidate_reference( x264_t *, int64_t pts );
#ifdef __cplusplus
}
#endif
#endif
类似资料：