sergiotarxz/libvips
1
0
Fork 0
Code Issues Pull Requests Releases Wiki Activity

split sRGB <-> XYZ up

Browse Source 
split sRGB <-> XYZ into sRGB <-> scRGB <-> XYZ so we can support
scRGB as a colourspace
This commit is contained in:
John Cupitt 2012-12-11 13:50:58 +00:00
parent 77ec00b495
commit 1bddb78f6a
1 changed files with 76 additions and 66 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

142
libvips/colour/LabQ2sRGB.c
View File

@ -18,6 +18,10 @@
* - faster and more accurate sRGB <-> XYZ conversion * - faster and more accurate sRGB <-> XYZ conversion
* 6/11/12 * 6/11/12
* - added a 16-bit path * - added a 16-bit path
* 11/12/12
* - spot NaN, Inf in XYZ2RGB, they break LUT indexing
* - split sRGB <-> XYZ into sRGB <-> scRGB <-> XYZ so we can support
* scRGB as a colourspace
*/ */
/* /*
@ -84,30 +88,6 @@ static int vips_Y2v_16[65536 + 1];
*/ */
static float vips_v2Y_16[65536]; static float vips_v2Y_16[65536];
/* We need Y in 0 - 100. We can save three muls later if we pre-scale the
* matrix.
*
* The matrix already includes the D65 channel weighting, so we just scale by
* Y.
*/
#define SCALE (VIPS_D65_Y0)
/* linear RGB -> XYZ matrix.
*/
static float vips_mat_RGB2XYZ[3][3] = {
{ SCALE * 0.4124, SCALE * 0.3576, SCALE * 0.18056 },
{ SCALE * 0.2126, SCALE * 0.7152, SCALE * 0.0722 },
{ SCALE * 0.0193, SCALE * 0.1192, SCALE * 0.9505 }
};
/* XYZ -> linear RGB matrix.
*/
static float vips_mat_XYZ2RGB[3][3] = {
{ 3.2406 / SCALE, -1.5372 / SCALE, -0.4986 / SCALE },
{ -0.9689 / SCALE, 1.8758 / SCALE, 0.0415 / SCALE },
{ 0.0557 / SCALE, -0.2040 / SCALE, 1.0570 / SCALE }
};
/* Do our own indexing of the arrays below to make sure we get efficient mults. /* Do our own indexing of the arrays below to make sure we get efficient mults.
*/ */
#define INDEX( L, A, B ) (L + (A << 6) + (B << 12)) #define INDEX( L, A, B ) (L + (A << 6) + (B << 12))
@ -176,33 +156,45 @@ vips_col_make_tables_RGB_8( void )
} }
} }
/* RGB to XYZ. /* sRGB to scRGB.
* *
* @range is eg. 256 for 8-bit data, * @range is eg. 256 for 8-bit data.
*/ */
static int static int
vips_col_sRGB2XYZ( int range, float *lut, vips_col_sRGB2scRGB( int range, float *lut,
int r, int g, int b, float *X, float *Y, float *Z ) int r, int g, int b, float *R, float *G, float *B )
{ {
int maxval = range - 1; int maxval = range - 1;
float Yr, Yg, Yb;
int i; int i;
i = VIPS_CLIP( 0, r, maxval ); i = VIPS_CLIP( 0, r, maxval );
Yr = lut[i]; *R = lut[i];
i = VIPS_CLIP( 0, g, maxval ); i = VIPS_CLIP( 0, g, maxval );
Yg = lut[i]; *G = lut[i];
i = VIPS_CLIP( 0, b, maxval ); i = VIPS_CLIP( 0, b, maxval );
Yb = lut[i]; *B = lut[i];
*X = vips_mat_RGB2XYZ[0][0] * Yr + return( 0 );
vips_mat_RGB2XYZ[0][1] * Yg + vips_mat_RGB2XYZ[0][2] * Yb; }
*Y = vips_mat_RGB2XYZ[1][0] * Yr +
vips_mat_RGB2XYZ[1][1] * Yg + vips_mat_RGB2XYZ[1][2] * Yb; /* We need Y in 0 - 100. We can save three muls later if we pre-scale the
*Z = vips_mat_RGB2XYZ[2][0] * Yr + * matrix.
vips_mat_RGB2XYZ[2][1] * Yg + vips_mat_RGB2XYZ[2][2] * Yb; *
* The matrix already includes the D65 channel weighting, so we just scale by
* Y.
*/
#define SCALE (VIPS_D65_Y0)
/* scRGB to XYZ.
*/
static int
vips_col_scRGB2XYZ( float R, float G, float B, float *X, float *Y, float *Z )
{
*X = SCALE * 0.4124 * R + SCALE * 0.3576 * G + SCALE * 0.18056 * B;
*Y = SCALE * 0.2126 * R + SCALE * 0.7152 * G + SCALE * 0.07220 * B;
*Z = SCALE * 0.0193 * R + SCALE * 0.1192 * G + SCALE * 0.9505 * B;
return( 0 ); return( 0 );
} }
@ -210,9 +202,12 @@ vips_col_sRGB2XYZ( int range, float *lut,
int int
vips_col_sRGB2XYZ_8( int r, int g, int b, float *X, float *Y, float *Z ) vips_col_sRGB2XYZ_8( int r, int g, int b, float *X, float *Y, float *Z )
{ {
float R, G, B;
vips_col_make_tables_RGB_8(); vips_col_make_tables_RGB_8();
return( vips_col_sRGB2XYZ( 256, vips_v2Y_8, r, g, b, X, Y, Z ) ); return( vips_col_sRGB2scRGB( 256, vips_v2Y_8, r, g, b, &R, &G, &B ) ||
vips_col_scRGB2XYZ( R, G, B, X, Y, Z ) );
} }
static void * static void *
@ -242,23 +237,40 @@ vips_col_make_tables_RGB_16( void )
int int
vips_col_sRGB2XYZ_16( int r, int g, int b, float *X, float *Y, float *Z ) vips_col_sRGB2XYZ_16( int r, int g, int b, float *X, float *Y, float *Z )
{ {
float R, G, B;
vips_col_make_tables_RGB_16(); vips_col_make_tables_RGB_16();
return( vips_col_sRGB2XYZ( 65536, vips_v2Y_16, r, g, b, X, Y, Z ) ); return( vips_col_sRGB2scRGB( 65536, vips_v2Y_16,
r, g, b, &R, &G, &B ) ||
vips_col_scRGB2XYZ( R, G, B, X, Y, Z ) );
} }
/* Turn XYZ into display colour. Return or=1 for out of gamut - rgb will /* Turn XYZ into scRGB.
* contain an approximation of the right colour.
*/ */
static int static int
vips_col_XYZ2sRGB( int range, int *lut, vips_col_XYZ2scRGB( float X, float Y, float Z, float *R, float *G, float *B )
float X, float Y, float Z, {
*R = 3.2406 / SCALE * X + -1.5372 / SCALE * Y + -0.4986 / SCALE * Z;
*G = -0.9689 / SCALE * X + 1.8758 / SCALE * Y + 0.0415 / SCALE * Z;
*B = 0.0557 / SCALE * X + -0.2040 / SCALE * Y + 1.0570 / SCALE * Z;
return( 0 );
}
/* Turn scRGB into sRGB. Return or=1 for out of gamut - rgb will contain an
* approximation of the right colour.
*
* Return -1 for NaN, Inf etc.
*/
static int
vips_col_scRGB2sRGB( int range, int *lut,
float R, float G, float B,
int *r, int *g, int *b, int *r, int *g, int *b,
int *or_ret ) int *or_ret )
{ {
int maxval = range - 1; int maxval = range - 1;
float Yr, Yg, Yb;
int or; int or;
float Yf; float Yf;
int Yi; int Yi;
@ -267,9 +279,9 @@ vips_col_XYZ2sRGB( int range, int *lut,
/* XYZ can be Nan, Inf etc. Throw those values out, they will break /* XYZ can be Nan, Inf etc. Throw those values out, they will break
* our clipping. * our clipping.
*/ */
if( !isnormal( X ) || if( !isnormal( R ) ||
!isnormal( Y ) || !isnormal( G ) ||
!isnormal( Z ) ) { !isnormal( B ) ) {
*r = 0; *r = 0;
*g = 0; *g = 0;
*b = 0; *b = 0;
@ -277,15 +289,6 @@ vips_col_XYZ2sRGB( int range, int *lut,
return( -1 ); return( -1 );
} }
/* Multiply through the matrix to get luminosity values.
*/
Yr = vips_mat_XYZ2RGB[0][0] * X +
vips_mat_XYZ2RGB[0][1] * Y + vips_mat_XYZ2RGB[0][2] * Z;
Yg = vips_mat_XYZ2RGB[1][0] * X +
vips_mat_XYZ2RGB[1][1] * Y + vips_mat_XYZ2RGB[1][2] * Z;
Yb = vips_mat_XYZ2RGB[2][0] * X +
vips_mat_XYZ2RGB[2][1] * Y + vips_mat_XYZ2RGB[2][2] * Z;
/* Clip range, set the out-of-range flag. /* Clip range, set the out-of-range flag.
*/ */
#define CLIP( L, V, H ) { \ #define CLIP( L, V, H ) { \
@ -307,20 +310,20 @@ vips_col_XYZ2sRGB( int range, int *lut,
or = 0; or = 0;
Yf = Yr * maxval; Yf = R * maxval;
CLIP( 0, Yf, maxval); CLIP( 0, Yf, maxval );
Yi = (int) Yf; Yi = (int) Yf;
v = lut[Yi] + (lut[Yi + 1] - lut[Yi]) * (Yf - Yi); v = lut[Yi] + (lut[Yi + 1] - lut[Yi]) * (Yf - Yi);
*r = VIPS_RINT( v ); *r = VIPS_RINT( v );
Yf = Yg * maxval; Yf = G * maxval;
CLIP( 0, Yf, maxval); CLIP( 0, Yf, maxval );
Yi = (int) Yf; Yi = (int) Yf;
v = lut[Yi] + (lut[Yi + 1] - lut[Yi]) * (Yf - Yi); v = lut[Yi] + (lut[Yi + 1] - lut[Yi]) * (Yf - Yi);
*g = VIPS_RINT( v ); *g = VIPS_RINT( v );
Yf = Yb * maxval; Yf = B * maxval;
CLIP( 0, Yf, maxval); CLIP( 0, Yf, maxval );
Yi = (int) Yf; Yi = (int) Yf;
v = lut[Yi] + (lut[Yi + 1] - lut[Yi]) * (Yf - Yi); v = lut[Yi] + (lut[Yi + 1] - lut[Yi]) * (Yf - Yi);
*b = VIPS_RINT( v ); *b = VIPS_RINT( v );
@ -336,9 +339,13 @@ vips_col_XYZ2sRGB_8( float X, float Y, float Z,
int *r, int *g, int *b, int *r, int *g, int *b,
int *or ) int *or )
{ {
float R, G, B;
vips_col_make_tables_RGB_8(); vips_col_make_tables_RGB_8();
return( vips_col_XYZ2sRGB( 256, vips_Y2v_8, X, Y, Z, r, g, b, or ) ); return( vips_col_XYZ2scRGB( X, Y, Z, &R, &G, &B ) ||
vips_col_scRGB2sRGB( 256, vips_Y2v_8,
R, G, B, r, g, b, or ) );
} }
int int
@ -346,10 +353,13 @@ vips_col_XYZ2sRGB_16( float X, float Y, float Z,
int *r, int *g, int *b, int *r, int *g, int *b,
int *or ) int *or )
{ {
float R, G, B;
vips_col_make_tables_RGB_16(); vips_col_make_tables_RGB_16();
return( vips_col_XYZ2sRGB( 65536, vips_Y2v_16, return( vips_col_XYZ2scRGB( X, Y, Z, &R, &G, &B ) ||
X, Y, Z, r, g, b, or ) ); vips_col_scRGB2sRGB( 65536, vips_Y2v_16,
R, G, B, r, g, b, or ) );
} }
/* Build Lab->disp dither tables. /* Build Lab->disp dither tables.