47 #define OFFSET(x) offsetof(FSPPContext, x)
48 #define FLAGS AV_OPT_FLAG_FILTERING_PARAM|AV_OPT_FLAG_VIDEO_PARAM
60 { 0, 48, 12, 60, 3, 51, 15, 63, },
61 { 32, 16, 44, 28, 35, 19, 47, 31, },
62 { 8, 56, 4, 52, 11, 59, 7, 55, },
63 { 40, 24, 36, 20, 43, 27, 39, 23, },
64 { 2, 50, 14, 62, 1, 49, 13, 61, },
65 { 34, 18, 46, 30, 33, 17, 45, 29, },
66 { 10, 58, 6, 54, 9, 57, 5, 53, },
67 { 42, 26, 38, 22, 41, 25, 37, 21, },
74 71, 296, 295, 237, 71, 40, 38, 19,
75 245, 193, 185, 121, 102, 73, 53, 27,
76 158, 129, 141, 107, 97, 73, 50, 26,
77 102, 116, 109, 98, 82, 66, 45, 23,
78 71, 94, 95, 81, 70, 56, 38, 20,
79 56, 77, 74, 66, 56, 44, 30, 15,
80 38, 53, 50, 45, 38, 30, 21, 11,
81 20, 27, 26, 23, 20, 15, 11, 5
86 ptrdiff_t dst_stride, ptrdiff_t src_stride,
87 ptrdiff_t
width, ptrdiff_t
height, ptrdiff_t log2_scale)
91 temp = (src[x + pos] + (d[pos] >> log2_scale)) >> (6 - log2_scale); \
92 src[x + pos] = src[x + pos - 8 * src_stride] = 0; \
93 if (temp & 0x100) temp = ~(temp >> 31); \
96 for (y = 0; y <
height; y++) {
98 for (x = 0; x <
width; x += 8) {
116 ptrdiff_t dst_stride, ptrdiff_t src_stride,
117 ptrdiff_t
width, ptrdiff_t
height, ptrdiff_t log2_scale)
120 #define STORE2(pos) \
121 temp = (src[x + pos] + src[x + pos + 16 * src_stride] + (d[pos] >> log2_scale)) >> (6 - log2_scale); \
122 src[x + pos + 16 * src_stride] = 0; \
123 if (temp & 0x100) temp = ~(temp >> 31); \
126 for (y = 0; y <
height; y++) {
128 for (x = 0; x <
width; x += 8) {
144 static void mul_thrmat_c(int16_t *thr_adr_noq, int16_t *thr_adr,
int q)
147 for (
a = 0;
a < 64;
a++)
148 thr_adr[
a] = q * thr_adr_noq[
a];
152 int dst_stride,
int src_stride,
154 uint8_t *qp_store,
int qp_stride,
int is_luma)
156 int x, x0, y, es, qy, t;
160 const int qpsh = 4 - p->
hsub * !is_luma;
161 const int qpsv = 4 - p->
vsub * !is_luma;
164 int16_t *
block = (int16_t *)block_align;
165 int16_t *block3 = (int16_t *)(block_align + 4 * 8 *
BLOCKSZ);
167 memset(block3, 0, 4 * 8 *
BLOCKSZ);
169 if (!
src || !dst)
return;
171 for (y = 0; y <
height; y++) {
174 for (x = 0; x < 8; x++) {
180 for (y = 0; y < 8; y++) {
186 for (y = 8; y < 24; y++)
189 for (y = step; y <
height + 8; y += step) {
190 const int y1 = y - 8 + step;
196 qy = (qy >> qpsv) * qp_stride;
205 for (x = 0; x < 8 * (
BLOCKSZ - 1); x += 8) {
210 t = qp_store[qy + (t >> qpsh)];
218 memmove(block3, block3 + (
BLOCKSZ - 1) * 64, 6 * 8 *
sizeof(int16_t));
229 if (!(y1 & 7) && y1) {
251 int_simd16_t tmp0, tmp1, tmp2, tmp3, tmp4, tmp5, tmp6, tmp7;
264 for (; cnt > 0; cnt -= 2) {
265 threshold = (int16_t *)thr_adr;
266 for (ctr =
DCTSIZE; ctr > 0; ctr--) {
301 tmp10 = (tmp0 + tmp2) >> 2;
302 tmp11 = (tmp0 - tmp2) >> 2;
304 tmp13 = (tmp1 + tmp3) >>2;
307 tmp0 = tmp10 + tmp13;
308 tmp3 = tmp10 - tmp13;
309 tmp1 = tmp11 + tmp12;
310 tmp2 = tmp11 - tmp12;
340 z10 = (tmp6 - tmp5) << 1;
342 z12 = (tmp4 - tmp7) << 1;
344 tmp7 = (z11 + z13) >> 2;
354 wsptr[
DCTSIZE * 0] += (tmp0 + tmp7);
355 wsptr[
DCTSIZE * 1] += (tmp1 + tmp6);
356 wsptr[
DCTSIZE * 2] += (tmp2 + tmp5);
357 wsptr[
DCTSIZE * 3] += (tmp3 - tmp4);
358 wsptr[
DCTSIZE * 4] += (tmp3 + tmp4);
359 wsptr[
DCTSIZE * 5] += (tmp2 - tmp5);
360 wsptr[
DCTSIZE * 6] = (tmp1 - tmp6);
361 wsptr[
DCTSIZE * 7] = (tmp0 - tmp7);
372 static void row_idct_c(int16_t *workspace, int16_t *output_adr, ptrdiff_t output_stride,
int cnt)
374 int_simd16_t tmp0, tmp1, tmp2, tmp3, tmp4, tmp5, tmp6, tmp7;
383 for (; cnt > 0; cnt--) {
386 tmp10 = wsptr[2] + wsptr[3];
387 tmp11 = wsptr[2] - wsptr[3];
389 tmp13 = wsptr[0] + wsptr[1];
392 tmp0 = tmp10 + tmp13;
393 tmp3 = tmp10 - tmp13;
394 tmp1 = tmp11 + tmp12;
395 tmp2 = tmp11 - tmp12;
403 z13 = wsptr[4] + wsptr[5];
404 z10 = wsptr[4] - wsptr[5];
405 z11 = wsptr[6] + wsptr[7];
406 z12 = wsptr[6] - wsptr[7];
415 tmp6 = (tmp12 << 3) - tmp7;
416 tmp5 = (tmp11 << 3) - tmp6;
417 tmp4 = (tmp10 << 3) + tmp5;
420 outptr[0 * output_stride] +=
DESCALE(tmp0 + tmp7, 3);
421 outptr[1 * output_stride] +=
DESCALE(tmp1 + tmp6, 3);
422 outptr[2 * output_stride] +=
DESCALE(tmp2 + tmp5, 3);
423 outptr[3 * output_stride] +=
DESCALE(tmp3 - tmp4, 3);
424 outptr[4 * output_stride] +=
DESCALE(tmp3 + tmp4, 3);
425 outptr[5 * output_stride] +=
DESCALE(tmp2 - tmp5, 3);
426 outptr[6 * output_stride] +=
DESCALE(tmp1 - tmp6, 3);
427 outptr[7 * output_stride] +=
DESCALE(tmp0 - tmp7, 3);
436 int_simd16_t tmp0, tmp1, tmp2, tmp3, tmp4, tmp5, tmp6, tmp7;
445 for (; cnt > 0; cnt--) {
446 tmp0 = pixels[line_size * 0] + pixels[line_size * 7];
447 tmp7 = pixels[line_size * 0] - pixels[line_size * 7];
448 tmp1 = pixels[line_size * 1] + pixels[line_size * 6];
449 tmp6 = pixels[line_size * 1] - pixels[line_size * 6];
450 tmp2 = pixels[line_size * 2] + pixels[line_size * 5];
451 tmp5 = pixels[line_size * 2] - pixels[line_size * 5];
452 tmp3 = pixels[line_size * 3] + pixels[line_size * 4];
453 tmp4 = pixels[line_size * 3] - pixels[line_size * 4];
464 dataptr[2] = tmp10 + tmp11;
465 dataptr[3] = tmp10 - tmp11;
468 dataptr[0] = tmp13 + z1;
469 dataptr[1] = tmp13 - z1;
473 tmp10 = (tmp4 + tmp5) << 2;
474 tmp11 = (tmp5 + tmp6) << 2;
475 tmp12 = (tmp6 + tmp7) << 2;
485 dataptr[4] = z13 + z2;
486 dataptr[5] = z13 - z2;
487 dataptr[6] = z11 + z4;
488 dataptr[7] = z11 - z4;
517 const int h =
FFALIGN(inlink->
h + 16, 16);
551 int8_t *qp_table =
NULL;
554 int custom_threshold_m[64];
561 for (
i = 0;
i < 8;
i++) {
563 |(((uint64_t)custom_threshold_m[
i * 8 + 6]) << 16)
564 |(((uint64_t)custom_threshold_m[
i * 8 + 0]) << 32)
565 |(((uint64_t)custom_threshold_m[
i * 8 + 4]) << 48);
568 |(((uint64_t)custom_threshold_m[
i * 8 + 3]) << 16)
569 |(((uint64_t)custom_threshold_m[
i * 8 + 1]) << 32)
570 |(((uint64_t)custom_threshold_m[
i * 8 + 7]) << 48);
600 if (qp_table || fspp->
qp) {
607 const int aligned_w =
FFALIGN(inlink->
w, 8);
608 const int aligned_h =
FFALIGN(inlink->
h, 8);
617 out->width =
in->width;
618 out->height =
in->height;
622 inlink->
w, inlink->
h, qp_table, qp_stride, 1);
624 cw, ch, qp_table, qp_stride, 0);
626 cw, ch, qp_table, qp_stride, 0);
634 in ->data[3],
in ->linesize[3],
635 inlink->
w, inlink->
h);
679 .priv_class = &fspp_class,
static const AVFilterPad inputs[]
static const AVFilterPad outputs[]
uint8_t pi<< 24) CONV_FUNC_GROUP(AV_SAMPLE_FMT_FLT, float, AV_SAMPLE_FMT_U8, uint8_t,(*(const uint8_t *) pi - 0x80) *(1.0f/(1<< 7))) CONV_FUNC_GROUP(AV_SAMPLE_FMT_DBL, double, AV_SAMPLE_FMT_U8, uint8_t,(*(const uint8_t *) pi - 0x80) *(1.0/(1<< 7))) CONV_FUNC_GROUP(AV_SAMPLE_FMT_U8, uint8_t, AV_SAMPLE_FMT_S16, int16_t,(*(const int16_t *) pi >> 8)+0x80) CONV_FUNC_GROUP(AV_SAMPLE_FMT_FLT, float, AV_SAMPLE_FMT_S16, int16_t, *(const int16_t *) pi *(1.0f/(1<< 15))) CONV_FUNC_GROUP(AV_SAMPLE_FMT_DBL, double, AV_SAMPLE_FMT_S16, int16_t, *(const int16_t *) pi *(1.0/(1<< 15))) CONV_FUNC_GROUP(AV_SAMPLE_FMT_U8, uint8_t, AV_SAMPLE_FMT_S32, int32_t,(*(const int32_t *) pi >> 24)+0x80) CONV_FUNC_GROUP(AV_SAMPLE_FMT_FLT, float, AV_SAMPLE_FMT_S32, int32_t, *(const int32_t *) pi *(1.0f/(1U<< 31))) CONV_FUNC_GROUP(AV_SAMPLE_FMT_DBL, double, AV_SAMPLE_FMT_S32, int32_t, *(const int32_t *) pi *(1.0/(1U<< 31))) CONV_FUNC_GROUP(AV_SAMPLE_FMT_U8, uint8_t, AV_SAMPLE_FMT_FLT, float, av_clip_uint8(lrintf(*(const float *) pi *(1<< 7))+0x80)) CONV_FUNC_GROUP(AV_SAMPLE_FMT_S16, int16_t, AV_SAMPLE_FMT_FLT, float, av_clip_int16(lrintf(*(const float *) pi *(1<< 15)))) CONV_FUNC_GROUP(AV_SAMPLE_FMT_S32, int32_t, AV_SAMPLE_FMT_FLT, float, av_clipl_int32(llrintf(*(const float *) pi *(1U<< 31)))) CONV_FUNC_GROUP(AV_SAMPLE_FMT_U8, uint8_t, AV_SAMPLE_FMT_DBL, double, av_clip_uint8(lrint(*(const double *) pi *(1<< 7))+0x80)) CONV_FUNC_GROUP(AV_SAMPLE_FMT_S16, int16_t, AV_SAMPLE_FMT_DBL, double, av_clip_int16(lrint(*(const double *) pi *(1<< 15)))) CONV_FUNC_GROUP(AV_SAMPLE_FMT_S32, int32_t, AV_SAMPLE_FMT_DBL, double, av_clipl_int32(llrint(*(const double *) pi *(1U<< 31)))) #define SET_CONV_FUNC_GROUP(ofmt, ifmt) static void set_generic_function(AudioConvert *ac) { } void ff_audio_convert_free(AudioConvert **ac) { if(! *ac) return;ff_dither_free(&(*ac) ->dc);av_freep(ac);} AudioConvert *ff_audio_convert_alloc(AVAudioResampleContext *avr, enum AVSampleFormat out_fmt, enum AVSampleFormat in_fmt, int channels, int sample_rate, int apply_map) { AudioConvert *ac;int in_planar, out_planar;ac=av_mallocz(sizeof(*ac));if(!ac) return NULL;ac->avr=avr;ac->out_fmt=out_fmt;ac->in_fmt=in_fmt;ac->channels=channels;ac->apply_map=apply_map;if(avr->dither_method !=AV_RESAMPLE_DITHER_NONE &&av_get_packed_sample_fmt(out_fmt)==AV_SAMPLE_FMT_S16 &&av_get_bytes_per_sample(in_fmt) > 2) { ac->dc=ff_dither_alloc(avr, out_fmt, in_fmt, channels, sample_rate, apply_map);if(!ac->dc) { av_free(ac);return NULL;} return ac;} in_planar=ff_sample_fmt_is_planar(in_fmt, channels);out_planar=ff_sample_fmt_is_planar(out_fmt, channels);if(in_planar==out_planar) { ac->func_type=CONV_FUNC_TYPE_FLAT;ac->planes=in_planar ? ac->channels :1;} else if(in_planar) ac->func_type=CONV_FUNC_TYPE_INTERLEAVE;else ac->func_type=CONV_FUNC_TYPE_DEINTERLEAVE;set_generic_function(ac);if(ARCH_AARCH64) ff_audio_convert_init_aarch64(ac);if(ARCH_ARM) ff_audio_convert_init_arm(ac);if(ARCH_X86) ff_audio_convert_init_x86(ac);return ac;} int ff_audio_convert(AudioConvert *ac, AudioData *out, AudioData *in) { int use_generic=1;int len=in->nb_samples;int p;if(ac->dc) { av_log(ac->avr, AV_LOG_TRACE, "%d samples - audio_convert: %s to %s (dithered)\n", len, av_get_sample_fmt_name(ac->in_fmt), av_get_sample_fmt_name(ac->out_fmt));return ff_convert_dither(ac-> in
simple assert() macros that are a bit more flexible than ISO C assert().
int ff_filter_frame(AVFilterLink *link, AVFrame *frame)
Send a frame of data to the next filter.
#define flags(name, subs,...)
#define AV_CEIL_RSHIFT(a, b)
#define AVFILTER_FLAG_SUPPORT_TIMELINE_INTERNAL
Same as AVFILTER_FLAG_SUPPORT_TIMELINE_GENERIC, except that the filter will have its filter_frame() c...
int av_frame_is_writable(AVFrame *frame)
Check if the frame data is writable.
void av_frame_free(AVFrame **frame)
Free the frame and any dynamically allocated objects in it, e.g.
int av_frame_copy_props(AVFrame *dst, const AVFrame *src)
Copy only "metadata" fields from src to dst.
#define DECLARE_ALIGNED(n, t, v)
Declare a variable that is aligned in memory.
void av_image_copy_plane(uint8_t *dst, int dst_linesize, const uint8_t *src, int src_linesize, int bytewidth, int height)
Copy image plane from src to dst.
@ AV_PICTURE_TYPE_B
Bi-dir predicted.
static int ff_norm_qscale(int qscale, int type)
Normalize the qscale factor FIXME the H264 qscale is a log based scale, mpeg1/2 is not,...
common internal API header
#define NULL_IF_CONFIG_SMALL(x)
Return NULL if CONFIG_SMALL is true, otherwise the argument without modification.
static enum AVPixelFormat pix_fmts[]
const AVPixFmtDescriptor * av_pix_fmt_desc_get(enum AVPixelFormat pix_fmt)
AVPixelFormat
Pixel format.
@ AV_PIX_FMT_YUV420P
planar YUV 4:2:0, 12bpp, (1 Cr & Cb sample per 2x2 Y samples)
@ AV_PIX_FMT_YUV440P
planar YUV 4:4:0 (1 Cr & Cb sample per 1x2 Y samples)
@ AV_PIX_FMT_YUV422P
planar YUV 4:2:2, 16bpp, (1 Cr & Cb sample per 2x1 Y samples)
@ AV_PIX_FMT_GRAY8
Y , 8bpp.
@ AV_PIX_FMT_YUVJ440P
planar YUV 4:4:0 full scale (JPEG), deprecated in favor of AV_PIX_FMT_YUV440P and setting color_range
@ AV_PIX_FMT_YUV410P
planar YUV 4:1:0, 9bpp, (1 Cr & Cb sample per 4x4 Y samples)
@ AV_PIX_FMT_YUV411P
planar YUV 4:1:1, 12bpp, (1 Cr & Cb sample per 4x1 Y samples)
@ AV_PIX_FMT_YUV444P
planar YUV 4:4:4, 24bpp, (1 Cr & Cb sample per 1x1 Y samples)
@ AV_PIX_FMT_YUVJ422P
planar YUV 4:2:2, 16bpp, full scale (JPEG), deprecated in favor of AV_PIX_FMT_YUV422P and setting col...
@ AV_PIX_FMT_GBRP
planar GBR 4:4:4 24bpp
@ AV_PIX_FMT_YUVJ444P
planar YUV 4:4:4, 24bpp, full scale (JPEG), deprecated in favor of AV_PIX_FMT_YUV444P and setting col...
@ AV_PIX_FMT_YUVJ420P
planar YUV 4:2:0, 12bpp, full scale (JPEG), deprecated in favor of AV_PIX_FMT_YUV420P and setting col...
int ff_qp_table_extract(AVFrame *frame, int8_t **table, int *table_w, int *table_h, int *qscale_type)
Extract a libpostproc-compatible QP table - an 8-bit QP value per 16x16 macroblock,...
A link between two filters.
int w
agreed upon image width
int h
agreed upon image height
AVFilterContext * dst
dest filter
int format
agreed upon media format
A filter pad used for either input or output.
const char * name
Pad name.
const char * name
Filter name.
This structure describes decoded (raw) audio or video data.
Descriptor that unambiguously describes how the bits of a pixel are stored in the up to 4 data planes...
void(* column_fidct)(int16_t *thr_adr, int16_t *data, int16_t *output, int cnt)
void(* store_slice2)(uint8_t *dst, int16_t *src, ptrdiff_t dst_stride, ptrdiff_t src_stride, ptrdiff_t width, ptrdiff_t height, ptrdiff_t log2_scale)
void(* store_slice)(uint8_t *dst, int16_t *src, ptrdiff_t dst_stride, ptrdiff_t src_stride, ptrdiff_t width, ptrdiff_t height, ptrdiff_t log2_scale)
uint64_t threshold_mtx_noq[8 *2]
void(* row_idct)(int16_t *workspace, int16_t *output_adr, ptrdiff_t output_stride, int cnt)
void(* mul_thrmat)(int16_t *thr_adr_noq, int16_t *thr_adr, int q)
uint64_t threshold_mtx[8 *2]
void(* row_fdct)(int16_t *data, const uint8_t *pixels, ptrdiff_t line_size, int cnt)
#define av_malloc_array(a, b)
AVFILTER_DEFINE_CLASS(fspp)
static void row_fdct_c(int16_t *data, const uint8_t *pixels, ptrdiff_t line_size, int cnt)
static const AVOption fspp_options[]
static void column_fidct_c(int16_t *thr_adr, int16_t *data, int16_t *output, int cnt)
static void store_slice_c(uint8_t *dst, int16_t *src, ptrdiff_t dst_stride, ptrdiff_t src_stride, ptrdiff_t width, ptrdiff_t height, ptrdiff_t log2_scale)
static int query_formats(AVFilterContext *ctx)
static int config_input(AVFilterLink *inlink)
static void mul_thrmat_c(int16_t *thr_adr_noq, int16_t *thr_adr, int q)
static void row_idct_c(int16_t *workspace, int16_t *output_adr, ptrdiff_t output_stride, int cnt)
static int filter_frame(AVFilterLink *inlink, AVFrame *in)
static const AVFilterPad fspp_outputs[]
static av_cold void uninit(AVFilterContext *ctx)
static const short custom_threshold[64]
static const uint8_t dither[8][8]
static void store_slice2_c(uint8_t *dst, int16_t *src, ptrdiff_t dst_stride, ptrdiff_t src_stride, ptrdiff_t width, ptrdiff_t height, ptrdiff_t log2_scale)
static void filter(FSPPContext *p, uint8_t *dst, uint8_t *src, int dst_stride, int src_stride, int width, int height, uint8_t *qp_store, int qp_stride, int is_luma)
static const AVFilterPad fspp_inputs[]
#define THRESHOLD(r, x, t)
static const int16_t FIX_1_414213562_A
#define MULTIPLY16H(x, k)
void ff_fspp_init_x86(FSPPContext *fspp)
AVFrame * ff_get_video_buffer(AVFilterLink *link, int w, int h)
Request a picture buffer with a specific set of permissions.