|
|
|
@ -51,6 +51,7 @@ typedef struct HYuvContext{ |
|
|
|
|
int width, height; |
|
|
|
|
int flags; |
|
|
|
|
int picture_number; |
|
|
|
|
int last_slice_end; |
|
|
|
|
int linesize[3]; |
|
|
|
|
uint8_t __align8 temp[3][2500]; |
|
|
|
|
uint64_t stats[3][256]; |
|
|
|
@ -157,15 +158,24 @@ static inline void add_left_prediction_bgr32(uint8_t *dst, uint8_t *src, int w, |
|
|
|
|
*blue= b; |
|
|
|
|
} |
|
|
|
|
|
|
|
|
|
//FIXME optimize
|
|
|
|
|
static inline int sub_left_prediction(uint8_t *dst, uint8_t *src, int w, int left){ |
|
|
|
|
static inline int sub_left_prediction(HYuvContext *s, uint8_t *dst, uint8_t *src, int w, int left){ |
|
|
|
|
int i; |
|
|
|
|
for(i=0; i<w; i++){ |
|
|
|
|
const int temp= src[i]; |
|
|
|
|
dst[i]= temp - left; |
|
|
|
|
left= temp; |
|
|
|
|
if(w<32){ |
|
|
|
|
for(i=0; i<w; i++){ |
|
|
|
|
const int temp= src[i]; |
|
|
|
|
dst[i]= temp - left; |
|
|
|
|
left= temp; |
|
|
|
|
} |
|
|
|
|
return left; |
|
|
|
|
}else{ |
|
|
|
|
for(i=0; i<16; i++){ |
|
|
|
|
const int temp= src[i]; |
|
|
|
|
dst[i]= temp - left; |
|
|
|
|
left= temp; |
|
|
|
|
} |
|
|
|
|
s->dsp.diff_bytes(dst+16, src+16, src+15, w-16); |
|
|
|
|
return src[w-1]; |
|
|
|
|
} |
|
|
|
|
return left; |
|
|
|
|
} |
|
|
|
|
|
|
|
|
|
static void read_len_table(uint8_t *dst, GetBitContext *gb){ |
|
|
|
@ -648,6 +658,32 @@ static void decode_bgr_bitstream(HYuvContext *s, int count){ |
|
|
|
|
} |
|
|
|
|
} |
|
|
|
|
|
|
|
|
|
static void draw_slice(HYuvContext *s, int y){ |
|
|
|
|
int h, cy; |
|
|
|
|
UINT8 *src_ptr[3]; |
|
|
|
|
|
|
|
|
|
if(s->avctx->draw_horiz_band==NULL)
|
|
|
|
|
return; |
|
|
|
|
|
|
|
|
|
h= y - s->last_slice_end; |
|
|
|
|
y -= h; |
|
|
|
|
|
|
|
|
|
if(s->bitstream_bpp==12){ |
|
|
|
|
cy= y>>1; |
|
|
|
|
}else{ |
|
|
|
|
cy= y; |
|
|
|
|
} |
|
|
|
|
|
|
|
|
|
src_ptr[0] = s->picture[0] + s->linesize[0]*y; |
|
|
|
|
src_ptr[1] = s->picture[1] + s->linesize[1]*cy; |
|
|
|
|
src_ptr[2] = s->picture[2] + s->linesize[2]*cy; |
|
|
|
|
emms_c(); |
|
|
|
|
|
|
|
|
|
s->avctx->draw_horiz_band(s->avctx, src_ptr, s->linesize[0], y, s->width, h); |
|
|
|
|
|
|
|
|
|
s->last_slice_end= y + h; |
|
|
|
|
} |
|
|
|
|
|
|
|
|
|
static int decode_frame(AVCodecContext *avctx, void *data, int *data_size, uint8_t *buf, int buf_size){ |
|
|
|
|
HYuvContext *s = avctx->priv_data; |
|
|
|
|
const int width= s->width; |
|
|
|
@ -684,6 +720,8 @@ static int decode_frame(AVCodecContext *avctx, void *data, int *data_size, uint8 |
|
|
|
|
fake_ystride= s->interlaced ? s->linesize[0]*2 : s->linesize[0]; |
|
|
|
|
fake_ustride= s->interlaced ? s->linesize[1]*2 : s->linesize[1]; |
|
|
|
|
fake_vstride= s->interlaced ? s->linesize[2]*2 : s->linesize[2]; |
|
|
|
|
|
|
|
|
|
s->last_slice_end= 0; |
|
|
|
|
|
|
|
|
|
if(s->bitstream_bpp<24){ |
|
|
|
|
int y, cy; |
|
|
|
@ -732,6 +770,8 @@ static int decode_frame(AVCodecContext *avctx, void *data, int *data_size, uint8 |
|
|
|
|
if(y>=s->height) break; |
|
|
|
|
} |
|
|
|
|
|
|
|
|
|
draw_slice(s, y); |
|
|
|
|
|
|
|
|
|
ydst= s->picture[0] + s->linesize[0]*y; |
|
|
|
|
udst= s->picture[1] + s->linesize[1]*cy; |
|
|
|
|
vdst= s->picture[2] + s->linesize[2]*cy; |
|
|
|
@ -753,6 +793,8 @@ static int decode_frame(AVCodecContext *avctx, void *data, int *data_size, uint8 |
|
|
|
|
} |
|
|
|
|
} |
|
|
|
|
} |
|
|
|
|
draw_slice(s, height); |
|
|
|
|
|
|
|
|
|
break; |
|
|
|
|
case MEDIAN: |
|
|
|
|
/* first line except first 2 pixels is left predicted */ |
|
|
|
@ -808,6 +850,7 @@ static int decode_frame(AVCodecContext *avctx, void *data, int *data_size, uint8 |
|
|
|
|
} |
|
|
|
|
if(y>=height) break; |
|
|
|
|
} |
|
|
|
|
draw_slice(s, y); |
|
|
|
|
|
|
|
|
|
decode_422_bitstream(s, width); |
|
|
|
|
|
|
|
|
@ -821,6 +864,8 @@ static int decode_frame(AVCodecContext *avctx, void *data, int *data_size, uint8 |
|
|
|
|
add_median_prediction(vdst, vdst - fake_vstride, s->temp[2], width2, &leftv, &lefttopv); |
|
|
|
|
} |
|
|
|
|
} |
|
|
|
|
|
|
|
|
|
draw_slice(s, height); |
|
|
|
|
break; |
|
|
|
|
} |
|
|
|
|
} |
|
|
|
@ -859,6 +904,7 @@ static int decode_frame(AVCodecContext *avctx, void *data, int *data_size, uint8 |
|
|
|
|
} |
|
|
|
|
} |
|
|
|
|
} |
|
|
|
|
draw_slice(s, height); // just 1 large slice as this isnt possible in reverse order
|
|
|
|
|
break; |
|
|
|
|
default: |
|
|
|
|
fprintf(stderr, "prediction type not supported!\n"); |
|
|
|
@ -922,9 +968,9 @@ static int encode_frame(AVCodecContext *avctx, unsigned char *buf, int buf_size, |
|
|
|
|
put_bits(&s->pb, 8, leftu= s->picture[1][0]); |
|
|
|
|
put_bits(&s->pb, 8, s->picture[0][0]); |
|
|
|
|
|
|
|
|
|
lefty= sub_left_prediction(s->temp[0], s->picture[0]+2, width-2 , lefty); |
|
|
|
|
leftu= sub_left_prediction(s->temp[1], s->picture[1]+1, width2-1, leftu); |
|
|
|
|
leftv= sub_left_prediction(s->temp[2], s->picture[2]+1, width2-1, leftv); |
|
|
|
|
lefty= sub_left_prediction(s, s->temp[0], s->picture[0]+2, width-2 , lefty); |
|
|
|
|
leftu= sub_left_prediction(s, s->temp[1], s->picture[1]+1, width2-1, leftu); |
|
|
|
|
leftv= sub_left_prediction(s, s->temp[2], s->picture[2]+1, width2-1, leftv); |
|
|
|
|
|
|
|
|
|
encode_422_bitstream(s, width-2); |
|
|
|
|
|
|
|
|
@ -932,17 +978,17 @@ static int encode_frame(AVCodecContext *avctx, unsigned char *buf, int buf_size, |
|
|
|
|
int lefttopy, lefttopu, lefttopv; |
|
|
|
|
cy=y=1; |
|
|
|
|
if(s->interlaced){ |
|
|
|
|
lefty= sub_left_prediction(s->temp[0], s->picture[0]+s->linesize[0], width , lefty); |
|
|
|
|
leftu= sub_left_prediction(s->temp[1], s->picture[1]+s->linesize[1], width2, leftu); |
|
|
|
|
leftv= sub_left_prediction(s->temp[2], s->picture[2]+s->linesize[2], width2, leftv); |
|
|
|
|
lefty= sub_left_prediction(s, s->temp[0], s->picture[0]+s->linesize[0], width , lefty); |
|
|
|
|
leftu= sub_left_prediction(s, s->temp[1], s->picture[1]+s->linesize[1], width2, leftu); |
|
|
|
|
leftv= sub_left_prediction(s, s->temp[2], s->picture[2]+s->linesize[2], width2, leftv); |
|
|
|
|
|
|
|
|
|
encode_422_bitstream(s, width); |
|
|
|
|
y++; cy++; |
|
|
|
|
} |
|
|
|
|
|
|
|
|
|
lefty= sub_left_prediction(s->temp[0], s->picture[0]+fake_ystride, 4, lefty); |
|
|
|
|
leftu= sub_left_prediction(s->temp[1], s->picture[1]+fake_ystride, 2, leftu); |
|
|
|
|
leftv= sub_left_prediction(s->temp[2], s->picture[2]+fake_ystride, 2, leftv); |
|
|
|
|
lefty= sub_left_prediction(s, s->temp[0], s->picture[0]+fake_ystride, 4, lefty); |
|
|
|
|
leftu= sub_left_prediction(s, s->temp[1], s->picture[1]+fake_ystride, 2, leftu); |
|
|
|
|
leftv= sub_left_prediction(s, s->temp[2], s->picture[2]+fake_ystride, 2, leftv); |
|
|
|
|
|
|
|
|
|
encode_422_bitstream(s, 4); |
|
|
|
|
|
|
|
|
@ -986,11 +1032,11 @@ static int encode_frame(AVCodecContext *avctx, unsigned char *buf, int buf_size, |
|
|
|
|
ydst= s->picture[0] + s->linesize[0]*y; |
|
|
|
|
|
|
|
|
|
if(s->predictor == PLANE && s->interlaced < y){ |
|
|
|
|
s->dsp.diff_bytes(s->temp[0], ydst, ydst - fake_ystride, width); |
|
|
|
|
s->dsp.diff_bytes(s->temp[1], ydst, ydst - fake_ystride, width); |
|
|
|
|
|
|
|
|
|
lefty= sub_left_prediction(s->temp[0], s->temp[0], width , lefty); |
|
|
|
|
lefty= sub_left_prediction(s, s->temp[0], s->temp[1], width , lefty); |
|
|
|
|
}else{ |
|
|
|
|
lefty= sub_left_prediction(s->temp[0], ydst, width , lefty); |
|
|
|
|
lefty= sub_left_prediction(s, s->temp[0], ydst, width , lefty); |
|
|
|
|
} |
|
|
|
|
encode_gray_bitstream(s, width); |
|
|
|
|
y++; |
|
|
|
@ -1002,17 +1048,17 @@ static int encode_frame(AVCodecContext *avctx, unsigned char *buf, int buf_size, |
|
|
|
|
vdst= s->picture[2] + s->linesize[2]*cy; |
|
|
|
|
|
|
|
|
|
if(s->predictor == PLANE && s->interlaced < cy){ |
|
|
|
|
s->dsp.diff_bytes(s->temp[0], ydst, ydst - fake_ystride, width); |
|
|
|
|
s->dsp.diff_bytes(s->temp[1], udst, udst - fake_ustride, width2); |
|
|
|
|
s->dsp.diff_bytes(s->temp[2], vdst, vdst - fake_vstride, width2); |
|
|
|
|
s->dsp.diff_bytes(s->temp[1], ydst, ydst - fake_ystride, width); |
|
|
|
|
s->dsp.diff_bytes(s->temp[2], udst, udst - fake_ustride, width2); |
|
|
|
|
s->dsp.diff_bytes(s->temp[3], vdst, vdst - fake_vstride, width2); |
|
|
|
|
|
|
|
|
|
lefty= sub_left_prediction(s->temp[0], s->temp[0], width , lefty); |
|
|
|
|
leftu= sub_left_prediction(s->temp[1], s->temp[1], width2, leftu); |
|
|
|
|
leftv= sub_left_prediction(s->temp[2], s->temp[2], width2, leftv); |
|
|
|
|
lefty= sub_left_prediction(s, s->temp[0], s->temp[1], width , lefty); |
|
|
|
|
leftu= sub_left_prediction(s, s->temp[1], s->temp[2], width2, leftu); |
|
|
|
|
leftv= sub_left_prediction(s, s->temp[2], s->temp[3], width2, leftv); |
|
|
|
|
}else{ |
|
|
|
|
lefty= sub_left_prediction(s->temp[0], ydst, width , lefty); |
|
|
|
|
leftu= sub_left_prediction(s->temp[1], udst, width2, leftu); |
|
|
|
|
leftv= sub_left_prediction(s->temp[2], vdst, width2, leftv); |
|
|
|
|
lefty= sub_left_prediction(s, s->temp[0], ydst, width , lefty); |
|
|
|
|
leftu= sub_left_prediction(s, s->temp[1], udst, width2, leftu); |
|
|
|
|
leftv= sub_left_prediction(s, s->temp[2], vdst, width2, leftv); |
|
|
|
|
} |
|
|
|
|
|
|
|
|
|
encode_422_bitstream(s, width); |
|
|
|
@ -1069,7 +1115,7 @@ AVCodec huffyuv_decoder = { |
|
|
|
|
NULL, |
|
|
|
|
decode_end, |
|
|
|
|
decode_frame, |
|
|
|
|
CODEC_CAP_DR1, |
|
|
|
|
CODEC_CAP_DR1 | CODEC_CAP_DRAW_HORIZ_BAND, |
|
|
|
|
NULL |
|
|
|
|
}; |
|
|
|
|
|
|
|
|
|