Skip to content

V
vlc-gpu
Commit 91b3be07 authored by Simon Hailes's avatar Simon Hailes Committed by Jean-Baptiste Kempf
Browse files
Options

Modified the way DDS (display definition segment - HD res dvb subs) is handled... 

Modified the way DDS (display definition segment - HD res dvb subs) is handled and added comments/explanation.

Applied by jb
parent 4ffd915e
Hide whitespace changes
Inline Side-by-side
Showing with 64 additions and 14 deletions
modules/codec/dvbsub.c
View file @ 91b3be07
......@@ -38,7 +38,28 @@
* refers to the character index in the codepage belonging to the language specified
* in the subtitle descriptor. Potentially it's designed for widechar
* (but not for UTF-*) codepages.
*****************************************************************************/
*****************************************************************************
* Notes on DDS (Display Definition Segment)
* DDS (Display Definition Segment) tells the decoder how the subtitle image relates to
* the video image.
* For SD, the subtitle image is always considered to be for display at 720x576
* (although it's assumed that for NTSC, this is 720x480, this is not documented well)
* Also, for SD, the subtitle image is drawn 'on the glass' (i.e. after video scaling,
+ * letterbox, etc.)
* For 'HD' (subs marked type 0x14/0x24 in PSI), a DDS must be present, and the subs area
* is drawn onto the video area (scales if necessary). The DDS tells the decoder what
* resolution the subtitle images were intended for, and hence how to scale the subtitle
* images for a particular video size
* i.e. if HD video is presented as letterbox, the subs will
* be in the same place on the video as if the video was presented on an HD set
* indeed, if the HD video was pillarboxed by the decoder, the subs may be cut off as
* well as the video. The intent here is that the subs can be placed accurately on the video
* - somthing which was missed in the original spec.
*
* A DDS may also specify a window - this is where the subs images are moved so that the (0,0)
* origin of decode is offset.
********************************************************************************************/
#ifdef HAVE_CONFIG_H
# include "config.h"
#endif
......@@ -151,6 +172,7 @@ typedef struct dvbsub_display_s
int i_height;
bool b_windowed;
/* these values are only relevant if windowed */
int i_x;
int i_y;
int i_max_x;
......@@ -997,11 +1019,20 @@ static void decode_display_definition( decoder_t *p_dec, bs_t *s )
p_sys->display.i_max_y = bs_read( s, 16 );
i_processed_length += 64;
}
else
{
/* if not windowed, setup the window variables to good defaults */
/* not necessary, but to avoid future confusion.. */
p_sys->display.i_x = 0;
p_sys->display.i_max_x = p_sys->display.i_width-1;
p_sys->display.i_y = 0;
p_sys->display.i_max_y = p_sys->display.i_height-1;
}
if( i_processed_length != i_segment_length*8 )
{
msg_Err( p_dec, "processed length %d != segment length %d",
i_processed_length, i_segment_length );
msg_Err( p_dec, "processed length %d bytes != segment length %d bytes",
i_processed_length / 8 , i_segment_length );
}
#ifdef DEBUG_DVBSUB
......@@ -1474,9 +1505,14 @@ static subpicture_t *render( decoder_t *p_dec )
if( p_sys->display.b_windowed )
{
/* TODO: check that this actually works */
p_spu->i_original_picture_width = p_sys->display.i_max_x - p_sys->display.i_x;
p_spu->i_original_picture_height = p_sys->display.i_max_y - p_sys->display.i_y;
/* From en_300743v01 - */
/* the DDS is there to indicate intended size/position of text */
/* the intended video area is ->i_width/height */
/* the window is within this... SO... we should leave i_original_picture_width etc. as is */
/* and ONLY change i_base_x. effectively i_max_x/y are only there to limit memory requirements*/
/* we COULD use the upper limits to limit rendering to within these? */
/* see notes on DDS at the top of the file */
i_base_x += p_sys->display.i_x;
i_base_y += p_sys->display.i_y;
}
......@@ -1505,11 +1541,6 @@ static subpicture_t *render( decoder_t *p_dec )
p_regiondef = &p_sys->p_page->p_region_defs[i];
#ifdef DEBUG_DVBSUB
msg_Dbg( p_dec, "rendering region %i (%i,%i)", i,
p_regiondef->i_x, p_regiondef->i_y );
#endif
/* Find associated region */
for( p_region = p_sys->p_regions; p_region != NULL;
p_region = p_region->p_next )
......@@ -1517,6 +1548,22 @@ static subpicture_t *render( decoder_t *p_dec )
if( p_regiondef->i_id == p_region->i_id ) break;
}
#ifdef DEBUG_DVBSUB
/* if a region exists, then print it's size */
if (p_region)
{
msg_Dbg( p_dec, "rendering region %i (%i,%i) to (%i,%i)", i,
p_regiondef->i_x, p_regiondef->i_y,
p_regiondef->i_x + p_region->i_width,
p_regiondef->i_y + p_region->i_height );
}
else
{
msg_Dbg( p_dec, "rendering region %i (%i,%i) (no region matched to render)", i,
p_regiondef->i_x, p_regiondef->i_y );
}
#endif
if( !p_region )
{
msg_Dbg( p_dec, "region %i not found", p_regiondef->i_id );
......@@ -1716,7 +1763,8 @@ static subpicture_t *YuvaYuvp( subpicture_t *p_subpic )
+ p_region->p_picture->p[0].i_pitch * 1 / 3;
int i_tolerance = 0;
#ifdef DEBUG_DVBSUB
#ifdef DEBUG_DVBSUB1
/* p_enc not valid here */
msg_Dbg( p_enc, "YuvaYuvp: i_pixels=%d, i_iterator=%d", i_pixels, i_iterator );
#endif
p_fmt->i_chroma = VLC_CODEC_YUVP;
......@@ -1793,7 +1841,8 @@ static subpicture_t *YuvaYuvp( subpicture_t *p_subpic )
}
}
#ifdef DEBUG_DVBSUB
#ifdef DEBUG_DVBSUB1
/* p_enc not valid here */
msg_Dbg( p_enc, "best palette has %d colors", p_fmt->p_palette->i_entries );
#endif
......@@ -1887,7 +1936,8 @@ static subpicture_t *YuvaYuvp( subpicture_t *p_subpic )
p_fmt->p_palette->palette[i][3] = 0;
}
p_fmt->p_palette->i_entries = i_max_entries;
#ifdef DEBUG_DVBSUB
#ifdef DEBUG_DVBSUB1
/* p_enc not valid here */
msg_Dbg( p_enc, "best palette has %d colors", p_fmt->p_palette->i_entries );
#endif
}
......
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