From c07d0c2d2f6f7b0eb6e92cc6204bf05037957e82 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?J=C3=B6rg=20Frings-F=C3=BCrst?= Date: Mon, 1 Sep 2014 15:43:52 +0200 Subject: Imported Upstream version 1.6.3 --- profile/profout.c | 156 ++++++++++++++++++++++++++++++++++++++++-------------- 1 file changed, 116 insertions(+), 40 deletions(-) (limited to 'profile/profout.c') diff --git a/profile/profout.c b/profile/profout.c index d89ae4e..8c6fa0d 100644 --- a/profile/profout.c +++ b/profile/profout.c @@ -65,6 +65,7 @@ #undef IMP_MONO /* [Undef] Turn on development code */ +#define EMPH_DISP_BLACKPOINT /* [Define] Increase weight near diplay black point */ #define NO_B2A_PCS_CURVES /* [Define] PCS curves seem to make B2A less accurate. Why ? */ #define USE_CAM_CLIP_OPT /* [Define] Clip out of gamut in CAM space rather than PCS */ #define USE_LEASTSQUARES_APROX /* [Define] Use least squares fitting approximation in B2A */ @@ -73,7 +74,7 @@ #undef DISABLE_GAMUT_TAG /* [Undef] To disable gamut tag */ #undef WARN_CLUT_CLIPPING /* [Undef] Print warning if setting clut clips */ #undef COMPARE_INV_CLUT /* [Undef] Compare result of inv_clut with clut to diag inv probs */ -#define FILTER_B2ACLIP /* [Define] Filter clip area of B2A */ +#undef FILTER_B2ACLIP /* [Undef] Filter clip area of B2A (Causes reversal problems) */ #define FILTER_THR_DE 3.0 /* [5.0] Filtering threshold DE */ #define FILTER_MAX_DE 5.0 /* [10.0] Filtering DE to gamut surface at whit MAX_RAD starts */ #define FILTER_MAX_RAD 0.1 /* [0.1] Filtering maximum radius in grid */ @@ -185,7 +186,7 @@ typedef struct { double swxyz[3]; /* Source white point in XYZ */ gamut *gam; /* Output gamut object for setting gamut Lut */ - int wantLab; /* 0 if want is XYZ PCS, 1 want is Lab PCS */ + int wantLab; /* 0 if is XYZ PCS, 1 if is Lab PCS */ } out_b2a_callback; /* Utility to handle abstract profile application to PCS. */ @@ -642,6 +643,7 @@ make_output_icc( int gamdiag, /* Make gamut mapping diagnostic wrl plots */ int verify, /* nz to print verification */ int clipprims, /* Clip white, black and primaries */ + double wpscale, /* >= 0.0 for media white point scale factor */ icxInk *oink, /* Ink limit/black generation setup (NULL if n/a) */ char *in_name, /* input .ti3 file name */ char *file_name, /* output icc name */ @@ -655,6 +657,7 @@ make_output_icc( int fwacomp, /* FWA compensation requested */ double smooth, /* RSPL smoothing factor, -ve if raw */ double avgdev, /* reading Average Deviation as a proportion of the input range */ + double demph, /* Emphasise dark region grid resolution in cLUT */ char *ipname, /* input icc profile - enables gamut map, NULL if none */ char *sgname, /* source image gamut - NULL if none */ char *absname[3], /* abstract profile name for each table */ @@ -737,7 +740,7 @@ make_output_icc( int ti; if ((ti = icg->find_kword(icg, 0, "LUMINANCE_XYZ_CDM2")) >= 0) { - if (sscanf(icg->t[0].kdata[ti], " %*lf %lf %*lf ",&dispLuminance) != 1) + if (sscanf(icg->t[0].kdata[ti], " %*f %lf %*f ",&dispLuminance) != 1) dispLuminance = 0.0; } @@ -1215,6 +1218,17 @@ make_output_icc( cc = 0.0; else if (cc > 1.0) cc = 1.0; + + if (cal->tvenc) { + cc = (cc * (235.0-16.0) + 16.0)/255.0; + + /* For video encoding the extra bits of precision are created by bit shifting */ + /* rather than scaling, so we need to scale the fp value to account for this. */ + /* We assume the precision is the vcgt table size = 16 */ + /* ~~99 ideally we should tag the fact that this is video encoded, so that */ + /* the vcgt loaded can adjust for a different bit precision ~~~~ */ + cc = (cc * 255 * (1 << (16 - 8)))/((1 << 16) - 1.0); + } ((unsigned short*)wo->u.table.data)[ncal * j + i] = (int)(cc * 65535.0 + 0.5); } } @@ -1389,12 +1403,12 @@ make_output_icc( wog->allocate((icmBase *)wog); wob->allocate((icmBase *)wob); - wor->data[0].X = 1.0; wor->data[0].Y = 0.0; wor->data[0].Z = 0.0; - wog->data[0].X = 0.0; wog->data[0].Y = 1.0; wog->data[0].Z = 0.0; - wob->data[0].X = 0.0; wob->data[0].Y = 0.0; wob->data[0].Z = 1.0; +// wor->data[0].X = 1.0; wor->data[0].Y = 0.0; wor->data[0].Z = 0.0; +// wog->data[0].X = 0.0; wog->data[0].Y = 1.0; wog->data[0].Z = 0.0; +// wob->data[0].X = 0.0; wob->data[0].Y = 0.0; wob->data[0].Z = 1.0; /* Setup deliberately wrong dummy values (channels rotated). */ - /* icxMatrix may override override these later */ + /* icxMatrix may override these later */ wor->data[0].X = 0.143066; wor->data[0].Y = 0.060608; wor->data[0].Z = 0.714096; wog->data[0].X = 0.436066; wog->data[0].Y = 0.222488; wog->data[0].Z = 0.013916; wob->data[0].X = 0.385147; wob->data[0].Y = 0.716873; wob->data[0].Z = 0.097076; @@ -1834,6 +1848,8 @@ make_output_icc( } + isLab = wantLab; /* We now have what we want */ + /* Normalize display values to Y = 1.0 if needed */ /* (re-norm spec derived, since observer may be different) */ if (isdisp && (isdnormed == 0 || spec != 0)) { @@ -1878,6 +1894,51 @@ make_output_icc( } } /* End of reading in CGATs file */ + +#ifdef EMPH_DISP_BLACKPOINT + /* Add extra weighting to sample points near black for additive display. */ + /* Not sure what the justification is, apart from making the black */ + /* point more accurately modelled. Does this reflect an underlying */ + /* problem with rspl in perceptual space ? Or is it a reflection */ + /* of the possible "scaled" viewing mode of additive display */ + /* usage ? What about print and scan ?? */ + if (isdisp && (imask == ICX_W || imask == ICX_RGB)) { + double minL = 1e6;; + + /* Locate the lowest L* value */ + for (i = 0; i < npat; i++) { + if (wantLab) { + if (tpat[i].v[0] < minL) + minL = tpat[i].v[0]; + } else { + double lab[2]; + icmXYZ2Lab(&icmD50, lab, tpat[i].v); + if (lab[0] < minL) + minL = lab[0]; + } + } + + /* Compute weighting factor */ + /* (Hard to guess what numbers to put in here.. */ + for (i = 0; i < npat; i++) { + double lab[3], L; + if (wantLab) { + L = tpat[i].v[0]; + } else { + icmXYZ2Lab(&icmD50, lab, tpat[i].v); + L = lab[0]; + } + L -= minL; + L /= 20.0; /* Just weight over 20 L* range */ + if (L > 1.0) + continue; + L = 1.0 + 19.0 * pow(1.0 - L, 3.0); + tpat[i].w *= L; +//printf("~1 pat %d %f %f %f weight %f\n", i,tpat[i].p[0], tpat[i].p[1], tpat[i].p[2], tpat[i].w); + } + } +#endif /* EMPH_DISP_BLACKPOINT */ + if (isLut) { xicc *wr_xicc; /* extention object */ icxLuBase *AtoB; /* AtoB ixcLu */ @@ -1920,7 +1981,7 @@ make_output_icc( ICX_2PASSSMTH | #endif flags, - npat, npat, tpat, NULL, dispLuminance, -1.0, smooth, avgdev, + npat, npat, tpat, NULL, dispLuminance, wpscale, smooth, avgdev, demph, NULL, oink, cal, iquality)) == NULL) error("%d, %s",wr_xicc->errc, wr_xicc->err); @@ -2016,18 +2077,20 @@ make_output_icc( vc->Lv = v->Lv; if (v->Yf >= 0.0) vc->Yf = v->Yf; - if (v->Fxyz[0] >= 0.0 && v->Fxyz[1] > 0.0 && v->Fxyz[2] >= 0.0) { + if (v->Yg >= 0.0) + vc->Yg = v->Yg; + if (v->Gxyz[0] >= 0.0 && v->Gxyz[1] > 0.0 && v->Gxyz[2] >= 0.0) { /* Normalise XYZ to current media white */ - vc->Fxyz[0] = v->Fxyz[0]/v->Fxyz[1] * vc->Fxyz[1]; - vc->Fxyz[2] = v->Fxyz[2]/v->Fxyz[1] * vc->Fxyz[1]; + vc->Gxyz[0] = v->Gxyz[0]/v->Gxyz[1] * vc->Gxyz[1]; + vc->Gxyz[2] = v->Gxyz[2]/v->Gxyz[1] * vc->Gxyz[1]; } - if (v->Fxyz[0] >= 0.0 && v->Fxyz[1] >= 0.0 && v->Fxyz[2] < 0.0) { + if (v->Gxyz[0] >= 0.0 && v->Gxyz[1] >= 0.0 && v->Gxyz[2] < 0.0) { /* Convert Yxy to XYZ */ - double x = v->Fxyz[0]; - double y = v->Fxyz[1]; /* If Y == 1.0, then X+Y+Z = 1/y */ + double x = v->Gxyz[0]; + double y = v->Gxyz[1]; /* If Y == 1.0, then X+Y+Z = 1/y */ double z = 1.0 - x - y; - vc->Fxyz[0] = x/y * vc->Fxyz[1]; - vc->Fxyz[2] = z/y * vc->Fxyz[1]; + vc->Gxyz[0] = x/y * vc->Gxyz[1]; + vc->Gxyz[2] = z/y * vc->Gxyz[1]; } } @@ -2182,31 +2245,40 @@ make_output_icc( /* cx.ix and cx.ox for each intent, but that would be expensive!. */ if (sepsat && (pgmi->usecas & 0xff) != (sgmi->usecas & 0xff)) error("Can't handle percept and sat table intents with different CAM spaces"); - /* Default perceptual input gamut mapping space is absolute perceptual */ - intentp = noptop ? icAbsoluteColorimetric : icmAbsolutePerceptual; - /* But override this for apperance space gamut mapping */ - if ((pgmi->usecas & 0xff) != 0x0) { + /* If apperance space perceptual table gamut mapping */ + if ((pgmi->usecas & 0xff) >= 0x2) { intentp = noptop ? icxAppearance : icxPerceptualAppearance; + } else { + if ((pgmi->usecas & 0xff) == 0x0) + intentp = noptop ? icRelativeColorimetric : icPerceptual; + else /* pgmi->usecas & 0xff == 0x1 */ + intentp = noptop ? icAbsoluteColorimetric : icmAbsolutePerceptual; } + if (sepsat) { - /* Default saturation gamut mapping space is absolute saturation */ - intents = nostos ? icAbsoluteColorimetric : icmAbsoluteSaturation; - /* But override this for apperance space gamut mapping */ - if ((sgmi->usecas & 0xff) != 0x0) { + /* If apperance space saturation table gamut mapping */ + if ((sgmi->usecas & 0xff) >= 0x2) { intents = nostos ? icxAppearance : icxSaturationAppearance; + } else { + if ((sgmi->usecas & 0xff) == 0x0) + intents = nostos ? icRelativeColorimetric : icSaturation; + else + intents = nostos ? icAbsoluteColorimetric : icmAbsoluteSaturation; } } - /* Default output gamut mapping space is absolute colorimetric */ - intento = icAbsoluteColorimetric; - /* But override this for apperance space gamut mapping */ - if ((pgmi->usecas & 0xff) != 0x0) { + if ((pgmi->usecas & 0xff) >= 0x2) { intento = icxAppearance; + } else { + if ((pgmi->usecas & 0xff) == 0x0) + intento = icRelativeColorimetric; + else /* pgmi->usecas & 0xff == 0x1 */ + intento = icAbsoluteColorimetric; } - if ((pgmi->usecas & 0xff) == 0x2) { + if ((pgmi->usecas & 0xff) == 0x3) { /* Abs. appearance */ double mxw; intentp = intents = intento = icxAbsAppearance; @@ -2274,7 +2346,7 @@ make_output_icc( if (sgname != NULL) { /* Optional source gamut - ie. from an images */ int isJab = 0; - if ((pgmi->usecas & 0xff) != 0) + if ((pgmi->usecas & 0xff) >= 0x2) isJab = 1; if (verb) @@ -2628,14 +2700,16 @@ make_output_icc( ICM_CLUT_SET_FILTER | #endif 0, - &cx, /* Context */ - cx.pcsspace, /* Input color space */ - devspace, /* Output color space */ - out_b2a_input, /* Input transform PCS->PCS' */ - NULL, NULL, /* Use default PCS range */ - out_b2a_clut, /* Lab' -> Device' transfer function */ - NULL, NULL, /* Use default Device' range */ - out_b2a_output) != 0) /* Output transfer function, Device'->Device */ + &cx, /* Context */ + cx.pcsspace, /* Input color space */ + devspace, /* Output color space */ + out_b2a_input, /* Input transform PCS->PCS' */ + NULL, NULL, /* Use default PCS range */ + out_b2a_clut, /* Lab' -> Device' transfer function */ + NULL, NULL, /* Use default Device' range */ + out_b2a_output, /* Output transfer function, Device'->Device */ + NULL, NULL /* Use default APXLS range */ + ) != 0) error("Setting 16 bit PCS->Device Lut failed: %d, %s",wr_icco->errc,wr_icco->err); if (cx.verb) { printf("\n"); @@ -2769,7 +2843,9 @@ make_output_icc( NULL, NULL, /* Use default Lab' range */ PCSp_bdist, /* Lab' -> Boundary distance */ NULL, NULL, /* Use default Device' range */ - gamut_output) != 0) /* Boundary distance to out of gamut value */ + gamut_output, /* Boundary distance to out of gamut value */ + NULL, NULL + ) != 0) error("Setting 16 bit PCS->Device Gamut Lut failed: %d, %s",wr_icco->errc,wr_icco->err); #endif /* !DEBUG_ONE */ if (cx.verb) { @@ -2823,7 +2899,7 @@ make_output_icc( wr_xicc, icmFwd, isdisp ? icmDefaultIntent : icRelativeColorimetric, icmLuOrdRev, flags, /* Compute white & black */ - npat, npat, tpat, NULL, dispLuminance, -1.0, smooth, avgdev, + npat, npat, tpat, NULL, dispLuminance, wpscale, smooth, avgdev, demph, NULL, oink, cal, iquality)) == NULL) error("%d, %s",wr_xicc->errc, wr_xicc->err); -- cgit v1.2.3