Commit 2edcd5be authored by Gildas Bazin's avatar Gildas Bazin

* modules/video_filter/blend.c: chroma fixes for the YUY2 blending.

parent e8e8633f
...@@ -512,6 +512,7 @@ static void BlendYUY2( filter_t *p_filter, picture_t *p_dst_pic, ...@@ -512,6 +512,7 @@ static void BlendYUY2( filter_t *p_filter, picture_t *p_dst_pic,
uint8_t *p_src2_u, *p_src2_v; uint8_t *p_src2_u, *p_src2_v;
uint8_t *p_trans; uint8_t *p_trans;
int i_x, i_y, i_pix_pitch, i_trans; int i_x, i_y, i_pix_pitch, i_trans;
vlc_bool_t b_even = !((i_x_offset + p_filter->fmt_out.video.i_x_offset)%2);
i_pix_pitch = 2; i_pix_pitch = 2;
i_dst_pitch = p_dst_pic->p->i_pitch; i_dst_pitch = p_dst_pic->p->i_pitch;
...@@ -541,6 +542,8 @@ static void BlendYUY2( filter_t *p_filter, picture_t *p_dst_pic, ...@@ -541,6 +542,8 @@ static void BlendYUY2( filter_t *p_filter, picture_t *p_dst_pic,
p_filter->fmt_in.video.i_x_offset + p_filter->fmt_in.video.i_x_offset +
p_src->p[A_PLANE].i_pitch * p_filter->fmt_in.video.i_y_offset; p_src->p[A_PLANE].i_pitch * p_filter->fmt_in.video.i_y_offset;
i_width = (i_width >> 1) << 1; /* Needs to be a multiple of 2 */
#define MAX_TRANS 255 #define MAX_TRANS 255
#define TRANS_BITS 8 #define TRANS_BITS 8
...@@ -551,7 +554,7 @@ static void BlendYUY2( filter_t *p_filter, picture_t *p_dst_pic, ...@@ -551,7 +554,7 @@ static void BlendYUY2( filter_t *p_filter, picture_t *p_dst_pic,
p_src2_v += i_src2_pitch ) p_src2_v += i_src2_pitch )
{ {
/* Draw until we reach the end of the line */ /* Draw until we reach the end of the line */
for( i_x = 0; i_x < i_width; i_x += 2 ) for( i_x = 0; i_x < i_width; i_x++, b_even = !b_even )
{ {
i_trans = ( p_trans[i_x] * i_alpha ) / 255; i_trans = ( p_trans[i_x] * i_alpha ) / 255;
if( !i_trans ) if( !i_trans )
...@@ -562,8 +565,12 @@ static void BlendYUY2( filter_t *p_filter, picture_t *p_dst_pic, ...@@ -562,8 +565,12 @@ static void BlendYUY2( filter_t *p_filter, picture_t *p_dst_pic,
{ {
/* Completely opaque. Completely overwrite underlying pixel */ /* Completely opaque. Completely overwrite underlying pixel */
p_dst[i_x * 2] = p_src2_y[i_x]; p_dst[i_x * 2] = p_src2_y[i_x];
p_dst[i_x * 2 + 1] = p_src2_u[i_x];
p_dst[i_x * 2 + 3] = p_src2_v[i_x]; if( b_even )
{
p_dst[i_x * 2 + 1] = p_src2_u[i_x];
p_dst[i_x * 2 + 3] = p_src2_v[i_x];
}
} }
else else
{ {
...@@ -571,30 +578,16 @@ static void BlendYUY2( filter_t *p_filter, picture_t *p_dst_pic, ...@@ -571,30 +578,16 @@ static void BlendYUY2( filter_t *p_filter, picture_t *p_dst_pic,
p_dst[i_x * 2] = ( (uint16_t)p_src2_y[i_x] * i_trans + p_dst[i_x * 2] = ( (uint16_t)p_src2_y[i_x] * i_trans +
(uint16_t)p_src1[i_x * 2] * (MAX_TRANS - i_trans) ) (uint16_t)p_src1[i_x * 2] * (MAX_TRANS - i_trans) )
>> TRANS_BITS; >> TRANS_BITS;
p_dst[i_x * 2 + 1] = ( (uint16_t)p_src2_u[i_x] * i_trans +
(uint16_t)p_src1[i_x * 2 + 1] * (MAX_TRANS - i_trans) )
>> TRANS_BITS;
p_dst[i_x * 2 + 3] = ( (uint16_t)p_src2_v[i_x] * i_trans +
(uint16_t)p_src1[i_x * 2 + 3] * (MAX_TRANS - i_trans) )
>> TRANS_BITS;
}
i_trans = ( p_trans[i_x+1] * i_alpha ) / 255; if( b_even )
if( !i_trans ) {
{ p_dst[i_x * 2 + 1] = ( (uint16_t)p_src2_u[i_x] * i_trans +
/* Completely transparent. Don't change pixel */ (uint16_t)p_src1[i_x * 2 + 1] * (MAX_TRANS - i_trans) )
} >> TRANS_BITS;
else if( i_trans == MAX_TRANS ) p_dst[i_x * 2 + 3] = ( (uint16_t)p_src2_v[i_x] * i_trans +
{ (uint16_t)p_src1[i_x * 2 + 3] * (MAX_TRANS - i_trans) )
/* Completely opaque. Completely overwrite underlying pixel */ >> TRANS_BITS;
p_dst[i_x * 2 + 2] = p_src2_y[i_x + 1]; }
}
else
{
/* Blending */
p_dst[i_x * 2 + 2] = ( (uint16_t)p_src2_y[i_x+1] * i_trans +
(uint16_t)p_src1[i_x * 2 + 2] * (MAX_TRANS - i_trans) )
>> TRANS_BITS;
} }
} }
} }
...@@ -720,6 +713,7 @@ static void BlendPalYUY2( filter_t *p_filter, picture_t *p_dst_pic, ...@@ -720,6 +713,7 @@ static void BlendPalYUY2( filter_t *p_filter, picture_t *p_dst_pic,
int i_src1_pitch, i_src2_pitch, i_dst_pitch; int i_src1_pitch, i_src2_pitch, i_dst_pitch;
uint8_t *p_src1, *p_src2, *p_dst; uint8_t *p_src1, *p_src2, *p_dst;
int i_x, i_y, i_pix_pitch, i_trans; int i_x, i_y, i_pix_pitch, i_trans;
vlc_bool_t b_even = !((i_x_offset + p_filter->fmt_out.video.i_x_offset)%2);
i_pix_pitch = 2; i_pix_pitch = 2;
i_dst_pitch = p_dst_pic->p->i_pitch; i_dst_pitch = p_dst_pic->p->i_pitch;
...@@ -736,6 +730,8 @@ static void BlendPalYUY2( filter_t *p_filter, picture_t *p_dst_pic, ...@@ -736,6 +730,8 @@ static void BlendPalYUY2( filter_t *p_filter, picture_t *p_dst_pic,
p_src2 = p_src->p->p_pixels + p_filter->fmt_in.video.i_x_offset + p_src2 = p_src->p->p_pixels + p_filter->fmt_in.video.i_x_offset +
i_src2_pitch * p_filter->fmt_in.video.i_y_offset; i_src2_pitch * p_filter->fmt_in.video.i_y_offset;
i_width = (i_width >> 1) << 1; /* Needs to be a multiple of 2 */
#define MAX_TRANS 255 #define MAX_TRANS 255
#define TRANS_BITS 8 #define TRANS_BITS 8
#define p_trans p_src2 #define p_trans p_src2
...@@ -746,7 +742,7 @@ static void BlendPalYUY2( filter_t *p_filter, picture_t *p_dst_pic, ...@@ -746,7 +742,7 @@ static void BlendPalYUY2( filter_t *p_filter, picture_t *p_dst_pic,
p_dst += i_dst_pitch, p_src1 += i_src1_pitch, p_src2 += i_src2_pitch ) p_dst += i_dst_pitch, p_src1 += i_src1_pitch, p_src2 += i_src2_pitch )
{ {
/* Draw until we reach the end of the line */ /* Draw until we reach the end of the line */
for( i_x = 0; i_x < i_width; i_x += 2 ) for( i_x = 0; i_x < i_width; i_x++, b_even = !b_even )
{ {
i_trans = ( p_pal[p_trans[i_x]][3] * i_alpha ) / 255; i_trans = ( p_pal[p_trans[i_x]][3] * i_alpha ) / 255;
if( !i_trans ) if( !i_trans )
...@@ -756,42 +752,30 @@ static void BlendPalYUY2( filter_t *p_filter, picture_t *p_dst_pic, ...@@ -756,42 +752,30 @@ static void BlendPalYUY2( filter_t *p_filter, picture_t *p_dst_pic,
else if( i_trans == MAX_TRANS ) else if( i_trans == MAX_TRANS )
{ {
/* Completely opaque. Completely overwrite underlying pixel */ /* Completely opaque. Completely overwrite underlying pixel */
/* NOTE: YUVP is actually YVUP */
p_dst[i_x * 2] = p_pal[p_src2[i_x]][0]; p_dst[i_x * 2] = p_pal[p_src2[i_x]][0];
p_dst[i_x * 2 + 1] = p_pal[p_src2[i_x]][2];
p_dst[i_x * 2 + 3] = p_pal[p_src2[i_x]][1]; if( b_even )
{
p_dst[i_x * 2 + 1] = p_pal[p_src2[i_x]][1];
p_dst[i_x * 2 + 3] = p_pal[p_src2[i_x]][2];
}
} }
else else
{ {
/* Blending */ /* Blending */
/* NOTE: YUVP is actually YVUP */
p_dst[i_x * 2] = ( (uint16_t)p_pal[p_src2[i_x]][0] * p_dst[i_x * 2] = ( (uint16_t)p_pal[p_src2[i_x]][0] *
i_trans + (uint16_t)p_src1[i_x * 2] * i_trans + (uint16_t)p_src1[i_x * 2] *
(MAX_TRANS - i_trans) ) >> TRANS_BITS; (MAX_TRANS - i_trans) ) >> TRANS_BITS;
p_dst[i_x * 2 + 1] = ( (uint16_t)p_pal[p_src2[i_x]][2] *
i_trans + (uint16_t)p_src1[i_x * 2 + 1] *
(MAX_TRANS - i_trans) ) >> TRANS_BITS;
p_dst[i_x * 2 + 3] = ( (uint16_t)p_pal[p_src2[i_x]][1] *
i_trans + (uint16_t)p_src1[i_x * 2 + 3] *
(MAX_TRANS - i_trans) ) >> TRANS_BITS;
}
i_trans = ( p_pal[p_trans[i_x+1]][3] * i_alpha ) / 255; if( b_even )
if( !i_trans ) {
{ p_dst[i_x * 2 + 1] = ( (uint16_t)p_pal[p_src2[i_x]][1] *
/* Completely transparent. Don't change pixel */ i_trans + (uint16_t)p_src1[i_x * 2 + 1] *
} (MAX_TRANS - i_trans) ) >> TRANS_BITS;
else if( i_trans == MAX_TRANS ) p_dst[i_x * 2 + 3] = ( (uint16_t)p_pal[p_src2[i_x]][2] *
{ i_trans + (uint16_t)p_src1[i_x * 2 + 3] *
/* Completely opaque. Completely overwrite underlying pixel */ (MAX_TRANS - i_trans) ) >> TRANS_BITS;
p_dst[i_x * 2 + 2] = p_pal[p_src2[i_x + 1]][0]; }
}
else
{
/* Blending */
p_dst[i_x * 2 + 2] = ( (uint16_t)p_pal[p_src2[i_x+1]][0] *
i_trans + (uint16_t)p_src1[i_x * 2 + 2] *
(MAX_TRANS - i_trans) ) >> TRANS_BITS;
} }
} }
} }
......
Markdown is supported
0%
or
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment