/* * transition_affine.c -- affine transformations * Copyright (C) 2003-2004 Ushodaya Enterprises Limited * Author: Charles Yates * * This library is free software; you can redistribute it and/or * modify it under the terms of the GNU Lesser General Public * License as published by the Free Software Foundation; either * version 2.1 of the License, or (at your option) any later version. * * This library 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 * Lesser General Public License for more details. * * You should have received a copy of the GNU Lesser General Public * License along with this library; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA */ #include "transition_affine.h" #include #include #include #include #include #include /** Calculate real geometry. */ static void geometry_calculate( mlt_transition this, char *store, struct mlt_geometry_item_s *output, float position ) { mlt_properties properties = MLT_TRANSITION_PROPERTIES( this ); mlt_geometry geometry = mlt_properties_get_data( properties, store, NULL ); int mirror_off = mlt_properties_get_int( properties, "mirror_off" ); int repeat_off = mlt_properties_get_int( properties, "repeat_off" ); int length = mlt_geometry_get_length( geometry ); // Allow wrapping if ( !repeat_off && position >= length && length != 0 ) { int section = position / length; position -= section * length; if ( !mirror_off && section % 2 == 1 ) position = length - position; } // Fetch the key for the position mlt_geometry_fetch( geometry, output, position ); } static mlt_geometry transition_parse_keys( mlt_transition this, char *name, char *store, int normalised_width, int normalised_height ) { // Get the properties of the transition mlt_properties properties = MLT_TRANSITION_PROPERTIES( this ); // Try to fetch it first mlt_geometry geometry = mlt_properties_get_data( properties, store, NULL ); // Get the in and out position mlt_position in = mlt_transition_get_in( this ); mlt_position out = mlt_transition_get_out( this ); // Determine length and obtain cycle int length = out - in + 1; double cycle = mlt_properties_get_double( properties, "cycle" ); // Allow a geometry repeat cycle if ( cycle >= 1 ) length = cycle; else if ( cycle > 0 ) length *= cycle; if ( geometry == NULL ) { // Get the new style geometry string char *property = mlt_properties_get( properties, name ); // Create an empty geometries object geometry = mlt_geometry_init( ); // Parse the geometry if we have one mlt_geometry_parse( geometry, property, length, normalised_width, normalised_height ); // Store it mlt_properties_set_data( properties, store, geometry, 0, ( mlt_destructor )mlt_geometry_close, NULL ); } else { // Check for updates and refresh if necessary mlt_geometry_refresh( geometry, mlt_properties_get( properties, name ), length, normalised_width, normalised_height ); } return geometry; } static mlt_geometry composite_calculate( mlt_transition this, struct mlt_geometry_item_s *result, int nw, int nh, float position ) { // Structures for geometry mlt_geometry start = transition_parse_keys( this, "geometry", "geometries", nw, nh ); // Do the calculation geometry_calculate( this, "geometries", result, position ); return start; } static inline float composite_calculate_key( mlt_transition this, char *name, char *store, int norm, float position ) { // Struct for the result struct mlt_geometry_item_s result; // Structures for geometry transition_parse_keys( this, name, store, norm, 0 ); // Do the calculation geometry_calculate( this, store, &result, position ); return result.x; } typedef struct { float matrix[3][3]; } affine_t; static void affine_init( float this[3][3] ) { this[0][0] = 1; this[0][1] = 0; this[0][2] = 0; this[1][0] = 0; this[1][1] = 1; this[1][2] = 0; this[2][0] = 0; this[2][1] = 0; this[2][2] = 1; } // Multiply two this affine transform with that static void affine_multiply( float this[3][3], float that[3][3] ) { float output[3][3]; int i; int j; for ( i = 0; i < 3; i ++ ) for ( j = 0; j < 3; j ++ ) output[i][j] = this[i][0] * that[j][0] + this[i][1] * that[j][1] + this[i][2] * that[j][2]; this[0][0] = output[0][0]; this[0][1] = output[0][1]; this[0][2] = output[0][2]; this[1][0] = output[1][0]; this[1][1] = output[1][1]; this[1][2] = output[1][2]; this[2][0] = output[2][0]; this[2][1] = output[2][1]; this[2][2] = output[2][2]; } // Rotate by a given angle static void affine_rotate_x( float this[3][3], float angle ) { float affine[3][3]; affine[0][0] = cos( angle * M_PI / 180 ); affine[0][1] = 0 - sin( angle * M_PI / 180 ); affine[0][2] = 0; affine[1][0] = sin( angle * M_PI / 180 ); affine[1][1] = cos( angle * M_PI / 180 ); affine[1][2] = 0; affine[2][0] = 0; affine[2][1] = 0; affine[2][2] = 1; affine_multiply( this, affine ); } static void affine_rotate_y( float this[3][3], float angle ) { float affine[3][3]; affine[0][0] = cos( angle * M_PI / 180 ); affine[0][1] = 0; affine[0][2] = 0 - sin( angle * M_PI / 180 ); affine[1][0] = 0; affine[1][1] = 1; affine[1][2] = 0; affine[2][0] = sin( angle * M_PI / 180 ); affine[2][1] = 0; affine[2][2] = cos( angle * M_PI / 180 ); affine_multiply( this, affine ); } static void affine_rotate_z( float this[3][3], float angle ) { float affine[3][3]; affine[0][0] = 1; affine[0][1] = 0; affine[0][2] = 0; affine[1][0] = 0; affine[1][1] = cos( angle * M_PI / 180 ); affine[1][2] = sin( angle * M_PI / 180 ); affine[2][0] = 0; affine[2][1] = - sin( angle * M_PI / 180 ); affine[2][2] = cos( angle * M_PI / 180 ); affine_multiply( this, affine ); } static void affine_scale( float this[3][3], float sx, float sy ) { float affine[3][3]; affine[0][0] = sx; affine[0][1] = 0; affine[0][2] = 0; affine[1][0] = 0; affine[1][1] = sy; affine[1][2] = 0; affine[2][0] = 0; affine[2][1] = 0; affine[2][2] = 1; affine_multiply( this, affine ); } // Shear by a given value static void affine_shear( float this[3][3], float shear_x, float shear_y, float shear_z ) { float affine[3][3]; affine[0][0] = 1; affine[0][1] = tan( shear_x * M_PI / 180 ); affine[0][2] = 0; affine[1][0] = tan( shear_y * M_PI / 180 ); affine[1][1] = 1; affine[1][2] = tan( shear_z * M_PI / 180 ); affine[2][0] = 0; affine[2][1] = 0; affine[2][2] = 1; affine_multiply( this, affine ); } static void affine_offset( float this[3][3], int x, int y ) { this[0][2] += x; this[1][2] += y; } // Obtain the mapped x coordinate of the input static inline double MapX( float this[3][3], int x, int y ) { return this[0][0] * x + this[0][1] * y + this[0][2]; } // Obtain the mapped y coordinate of the input static inline double MapY( float this[3][3], int x, int y ) { return this[1][0] * x + this[1][1] * y + this[1][2]; } static inline double MapZ( float this[3][3], int x, int y ) { return this[2][0] * x + this[2][1] * y + this[2][2]; } #define MAX( x, y ) x > y ? x : y #define MIN( x, y ) x < y ? x : y static void affine_max_output( float this[3][3], float *w, float *h, float dz ) { int tlx = MapX( this, -720, 576 ) / dz; int tly = MapY( this, -720, 576 ) / dz; int trx = MapX( this, 720, 576 ) / dz; int try = MapY( this, 720, 576 ) / dz; int blx = MapX( this, -720, -576 ) / dz; int bly = MapY( this, -720, -576 ) / dz; int brx = MapX( this, 720, -576 ) / dz; int bry = MapY( this, 720, -576 ) / dz; int max_x; int max_y; int min_x; int min_y; max_x = MAX( tlx, trx ); max_x = MAX( max_x, blx ); max_x = MAX( max_x, brx ); min_x = MIN( tlx, trx ); min_x = MIN( min_x, blx ); min_x = MIN( min_x, brx ); max_y = MAX( tly, try ); max_y = MAX( max_y, bly ); max_y = MAX( max_y, bry ); min_y = MIN( tly, try ); min_y = MIN( min_y, bly ); min_y = MIN( min_y, bry ); *w = ( float )( max_x - min_x + 1 ) / 1440.0; *h = ( float )( max_y - min_y + 1 ) / 1152.0; } #define IN_RANGE( v, r ) ( v >= - r / 2 && v < r / 2 ) static inline void get_affine( affine_t *affine, mlt_transition this, float position ) { mlt_properties properties = MLT_TRANSITION_PROPERTIES( this ); int keyed = mlt_properties_get_int( properties, "keyed" ); affine_init( affine->matrix ); if ( keyed == 0 ) { float fix_rotate_x = mlt_properties_get_double( properties, "fix_rotate_x" ); float fix_rotate_y = mlt_properties_get_double( properties, "fix_rotate_y" ); float fix_rotate_z = mlt_properties_get_double( properties, "fix_rotate_z" ); float rotate_x = mlt_properties_get_double( properties, "rotate_x" ); float rotate_y = mlt_properties_get_double( properties, "rotate_y" ); float rotate_z = mlt_properties_get_double( properties, "rotate_z" ); float fix_shear_x = mlt_properties_get_double( properties, "fix_shear_x" ); float fix_shear_y = mlt_properties_get_double( properties, "fix_shear_y" ); float fix_shear_z = mlt_properties_get_double( properties, "fix_shear_z" ); float shear_x = mlt_properties_get_double( properties, "shear_x" ); float shear_y = mlt_properties_get_double( properties, "shear_y" ); float shear_z = mlt_properties_get_double( properties, "shear_z" ); float ox = mlt_properties_get_double( properties, "ox" ); float oy = mlt_properties_get_double( properties, "oy" ); affine_rotate_x( affine->matrix, fix_rotate_x + rotate_x * position ); affine_rotate_y( affine->matrix, fix_rotate_y + rotate_y * position ); affine_rotate_z( affine->matrix, fix_rotate_z + rotate_z * position ); affine_shear( affine->matrix, fix_shear_x + shear_x * position, fix_shear_y + shear_y * position, fix_shear_z + shear_z * position ); affine_offset( affine->matrix, ox, oy ); } else { float rotate_x = composite_calculate_key( this, "rotate_x", "rotate_x_info", 360, position ); float rotate_y = composite_calculate_key( this, "rotate_y", "rotate_y_info", 360, position ); float rotate_z = composite_calculate_key( this, "rotate_z", "rotate_z_info", 360, position ); float shear_x = composite_calculate_key( this, "shear_x", "shear_x_info", 360, position ); float shear_y = composite_calculate_key( this, "shear_y", "shear_y_info", 360, position ); float shear_z = composite_calculate_key( this, "shear_z", "shear_z_info", 360, position ); affine_rotate_x( affine->matrix, rotate_x ); affine_rotate_y( affine->matrix, rotate_y ); affine_rotate_z( affine->matrix, rotate_z ); affine_shear( affine->matrix, shear_x, shear_y, shear_z ); } } /** Get the image. */ static int transition_get_image( mlt_frame a_frame, uint8_t **image, mlt_image_format *format, int *width, int *height, int writable ) { // Get the b frame from the stack mlt_frame b_frame = mlt_frame_pop_frame( a_frame ); // Get the transition object mlt_transition this = mlt_frame_pop_service( a_frame ); // Get the properties of the transition mlt_properties properties = MLT_TRANSITION_PROPERTIES( this ); // Get the properties of the a frame mlt_properties a_props = MLT_FRAME_PROPERTIES( a_frame ); // Get the properties of the b frame mlt_properties b_props = MLT_FRAME_PROPERTIES( b_frame ); // Image, format, width, height and image for the b frame uint8_t *b_image = NULL; mlt_image_format b_format = mlt_image_yuv422; int b_width; int b_height; // Get the unique name to retrieve the frame position char *name = mlt_properties_get( properties, "_unique_id" ); // Assign the current position to the name mlt_position position = mlt_properties_get_position( a_props, name ); mlt_position in = mlt_properties_get_position( properties, "in" ); mlt_position out = mlt_properties_get_position( properties, "out" ); int mirror = mlt_properties_get_position( properties, "mirror" ); int length = out - in + 1; // Obtain the normalised width and height from the a_frame int normalised_width = mlt_properties_get_int( a_props, "normalised_width" ); int normalised_height = mlt_properties_get_int( a_props, "normalised_height" ); double consumer_ar = mlt_properties_get_double( a_props, "consumer_aspect_ratio" ) ; // Structures for geometry struct mlt_geometry_item_s result; if ( mirror && position > length / 2 ) position = abs( position - length ); // Fetch the a frame image mlt_frame_get_image( a_frame, image, format, width, height, 1 ); // Calculate the region now composite_calculate( this, &result, normalised_width, normalised_height, ( float )position ); // Fetch the b frame image result.w = ( int )( result.w * *width / normalised_width ); result.h = ( int )( result.h * *height / normalised_height ); result.x = ( int )( result.x * *width / normalised_width ); result.y = ( int )( result.y * *height / normalised_height ); //result.w -= ( int )abs( result.w ) % 2; //result.x -= ( int )abs( result.x ) % 2; b_width = result.w; b_height = result.h; if ( mlt_properties_get_double( b_props, "aspect_ratio" ) == 0.0 ) mlt_properties_set_double( b_props, "aspect_ratio", consumer_ar ); if ( !strcmp( mlt_properties_get( a_props, "rescale.interp" ), "none" ) ) { mlt_properties_set( b_props, "rescale.interp", "nearest" ); mlt_properties_set_double( b_props, "consumer_aspect_ratio", consumer_ar ); } else { mlt_properties_set( b_props, "rescale.interp", mlt_properties_get( a_props, "rescale.interp" ) ); mlt_properties_set_double( b_props, "consumer_aspect_ratio", consumer_ar ); } mlt_properties_set_int( b_props, "distort", mlt_properties_get_int( properties, "distort" ) ); mlt_frame_get_image( b_frame, &b_image, &b_format, &b_width, &b_height, 0 ); result.w = b_width; result.h = b_height; // Check that both images are of the correct format and process if ( *format == mlt_image_yuv422 && b_format == mlt_image_yuv422 ) { register int x, y; register int dx, dy; double dz; float sw, sh; // Get values from the transition float scale_x = mlt_properties_get_double( properties, "scale_x" ); float scale_y = mlt_properties_get_double( properties, "scale_y" ); int scale = mlt_properties_get_int( properties, "scale" ); uint8_t *p = *image; uint8_t *q = *image; int cx = result.x + ( b_width >> 1 ); int cy = result.y + ( b_height >> 1 ); int lower_x = 0 - cx; int upper_x = *width - cx; int lower_y = 0 - cy; int upper_y = *height - cy; int b_stride = b_width << 1; int a_stride = *width << 1; int x_offset = ( int )result.w >> 1; int y_offset = ( int )result.h >> 1; uint8_t *alpha = mlt_frame_get_alpha_mask( b_frame ); uint8_t *mask = mlt_frame_get_alpha_mask( a_frame ); uint8_t *pmask = mask; float mix; affine_t affine; get_affine( &affine, this, ( float )position ); q = *image; dz = MapZ( affine.matrix, 0, 0 ); if ( mask == NULL ) { mask = mlt_pool_alloc( *width * *height ); pmask = mask; memset( mask, 255, *width * *height ); } if ( ( int )abs( dz * 1000 ) < 25 ) goto getout; if ( scale ) { affine_max_output( affine.matrix, &sw, &sh, dz ); affine_scale( affine.matrix, sw, sh ); } else if ( scale_x != 0 && scale_y != 0 ) { affine_scale( affine.matrix, scale_x, scale_y ); } if ( alpha == NULL ) { for ( y = lower_y; y < upper_y; y ++ ) { p = q; for ( x = lower_x; x < upper_x; x ++ ) { dx = MapX( affine.matrix, x, y ) / dz + x_offset; dy = MapY( affine.matrix, x, y ) / dz + y_offset; if ( dx >= 0 && dx < b_width && dy >=0 && dy < b_height ) { pmask ++; dx -= dx & 1; *p ++ = *( b_image + dy * b_stride + ( dx << 1 ) ); *p ++ = *( b_image + dy * b_stride + ( dx << 1 ) + ( ( x & 1 ) << 1 ) + 1 ); } else { p += 2; pmask ++; } } q += a_stride; } } else { for ( y = lower_y; y < upper_y; y ++ ) { p = q; for ( x = lower_x; x < upper_x; x ++ ) { dx = MapX( affine.matrix, x, y ) / dz + x_offset; dy = MapY( affine.matrix, x, y ) / dz + y_offset; if ( dx >= 0 && dx < b_width && dy >=0 && dy < b_height ) { *pmask ++ = *( alpha + dy * b_width + dx ); mix = ( float )*( alpha + dy * b_width + dx ) / 255.0; dx -= dx & 1; *p = *p * ( 1 - mix ) + mix * *( b_image + dy * b_stride + ( dx << 1 ) ); p ++; *p = *p * ( 1 - mix ) + mix * *( b_image + dy * b_stride + ( dx << 1 ) + ( ( x & 1 ) << 1 ) + 1 ); p ++; } else { p += 2; pmask ++; } } q += a_stride; } } getout: a_frame->get_alpha_mask = NULL; mlt_properties_set_data( a_props, "alpha", mask, 0, mlt_pool_release, NULL ); } return 0; } /** Affine transition processing. */ static mlt_frame transition_process( mlt_transition transition, mlt_frame a_frame, mlt_frame b_frame ) { // Get a unique name to store the frame position char *name = mlt_properties_get( MLT_TRANSITION_PROPERTIES( transition ), "_unique_id" ); // Assign the current position to the name mlt_properties a_props = MLT_FRAME_PROPERTIES( a_frame ); mlt_properties_set_position( a_props, name, mlt_frame_get_position( a_frame ) ); // Push the transition on to the frame mlt_frame_push_service( a_frame, transition ); // Push the b_frame on to the stack mlt_frame_push_frame( a_frame, b_frame ); // Push the transition method mlt_frame_push_get_image( a_frame, transition_get_image ); return a_frame; } /** Constructor for the filter. */ mlt_transition transition_affine_init( char *arg ) { mlt_transition transition = mlt_transition_new( ); if ( transition != NULL ) { mlt_properties_set_int( MLT_TRANSITION_PROPERTIES( transition ), "sx", 1 ); mlt_properties_set_int( MLT_TRANSITION_PROPERTIES( transition ), "sy", 1 ); mlt_properties_set_int( MLT_TRANSITION_PROPERTIES( transition ), "distort", 0 ); mlt_properties_set( MLT_TRANSITION_PROPERTIES( transition ), "geometry", "0,0:100%x100%" ); // Inform apps and framework that this is a video only transition mlt_properties_set_int( MLT_TRANSITION_PROPERTIES( transition ), "_transition_type", 1 ); transition->process = transition_process; } return transition; }