1//---------------------------------------------------------------------------------
2//
3// Little Color Management System
4// Copyright (c) 1998-2013 Marti Maria Saguer
5//
6// Permission is hereby granted, free of charge, to any person obtaining
7// a copy of this software and associated documentation files (the "Software"),
8// to deal in the Software without restriction, including without limitation
9// the rights to use, copy, modify, merge, publish, distribute, sublicense,
10// and/or sell copies of the Software, and to permit persons to whom the Software
11// is furnished to do so, subject to the following conditions:
12//
13// The above copyright notice and this permission notice shall be included in
14// all copies or substantial portions of the Software.
15//
16// THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
17// EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO
18// THE WARRANTIES OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND
19// NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE
20// LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION
21// OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION
22// WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.
23//
24//---------------------------------------------------------------------------------
25//
26#include "lcms2_internal.h"
27
28// Tone curves are powerful constructs that can contain curves specified in diverse ways.
29// The curve is stored in segments, where each segment can be sampled or specified by parameters.
30// a 16.bit simplification of the *whole* curve is kept for optimization purposes. For float operation,
31// each segment is evaluated separately. Plug-ins may be used to define new parametric schemes,
32// each plug-in may define up to MAX_TYPES_IN_LCMS_PLUGIN functions types. For defining a function,
33// the plug-in should provide the type id, how many parameters each type has, and a pointer to
34// a procedure that evaluates the function. In the case of reverse evaluation, the evaluator will
35// be called with the type id as a negative value, and a sampled version of the reversed curve
36// will be built.
37
38// ----------------------------------------------------------------- Implementation
39// Maxim number of nodes
40#define MAX_NODES_IN_CURVE 4097
41#define MINUS_INF (-1E22F)
42#define PLUS_INF (+1E22F)
43
44// The list of supported parametric curves
45typedef struct _cmsParametricCurvesCollection_st {
46
47 cmsUInt32Number nFunctions; // Number of supported functions in this chunk
48 cmsInt32Number FunctionTypes[MAX_TYPES_IN_LCMS_PLUGIN]; // The identification types
49 cmsUInt32Number ParameterCount[MAX_TYPES_IN_LCMS_PLUGIN]; // Number of parameters for each function
50
51 cmsParametricCurveEvaluator Evaluator; // The evaluator
52
53 struct _cmsParametricCurvesCollection_st* Next; // Next in list
54
55} _cmsParametricCurvesCollection;
56
57// This is the default (built-in) evaluator
58static cmsFloat64Number DefaultEvalParametricFn(cmsContext ContextID, cmsInt32Number Type, const cmsFloat64Number Params[], cmsFloat64Number R);
59
60// The built-in list
61static _cmsParametricCurvesCollection DefaultCurves = {
62 9, // # of curve types
63 { 1, 2, 3, 4, 5, 6, 7, 8, 108 }, // Parametric curve ID
64 { 1, 3, 4, 5, 7, 4, 5, 5, 1 }, // Parameters by type
65 DefaultEvalParametricFn, // Evaluator
66 NULL // Next in chain
67};
68
69// Duplicates the zone of memory used by the plug-in in the new context
70static
71void DupPluginCurvesList(struct _cmsContext_struct* ctx,
72 const struct _cmsContext_struct* src)
73{
74 _cmsCurvesPluginChunkType newHead = { NULL };
75 _cmsParametricCurvesCollection* entry;
76 _cmsParametricCurvesCollection* Anterior = NULL;
77 _cmsCurvesPluginChunkType* head = (_cmsCurvesPluginChunkType*) src->chunks[CurvesPlugin];
78
79 _cmsAssert(head != NULL);
80
81 // Walk the list copying all nodes
82 for (entry = head->ParametricCurves;
83 entry != NULL;
84 entry = entry ->Next) {
85
86 _cmsParametricCurvesCollection *newEntry = ( _cmsParametricCurvesCollection *) _cmsSubAllocDup(ctx ->MemPool, entry, sizeof(_cmsParametricCurvesCollection));
87
88 if (newEntry == NULL)
89 return;
90
91 // We want to keep the linked list order, so this is a little bit tricky
92 newEntry -> Next = NULL;
93 if (Anterior)
94 Anterior -> Next = newEntry;
95
96 Anterior = newEntry;
97
98 if (newHead.ParametricCurves == NULL)
99 newHead.ParametricCurves = newEntry;
100 }
101
102 ctx ->chunks[CurvesPlugin] = _cmsSubAllocDup(ctx->MemPool, &newHead, sizeof(_cmsCurvesPluginChunkType));
103}
104
105// The allocator have to follow the chain
106void _cmsAllocCurvesPluginChunk(struct _cmsContext_struct* ctx,
107 const struct _cmsContext_struct* src)
108{
109 _cmsAssert(ctx != NULL);
110
111 if (src != NULL) {
112
113 // Copy all linked list
114 DupPluginCurvesList(ctx, src);
115 }
116 else {
117 static _cmsCurvesPluginChunkType CurvesPluginChunk = { NULL };
118 ctx ->chunks[CurvesPlugin] = _cmsSubAllocDup(ctx ->MemPool, &CurvesPluginChunk, sizeof(_cmsCurvesPluginChunkType));
119 }
120}
121
122
123// The linked list head
124_cmsCurvesPluginChunkType _cmsCurvesPluginChunk = { NULL };
125
126// As a way to install new parametric curves
127cmsBool _cmsRegisterParametricCurvesPlugin(cmsContext ContextID, cmsPluginBase* Data)
128{
129 _cmsCurvesPluginChunkType* ctx = ( _cmsCurvesPluginChunkType*) _cmsContextGetClientChunk(ContextID, CurvesPlugin);
130 cmsPluginParametricCurves* Plugin = (cmsPluginParametricCurves*) Data;
131 _cmsParametricCurvesCollection* fl;
132
133 if (Data == NULL) {
134
135 ctx -> ParametricCurves = NULL;
136 return TRUE;
137 }
138
139 fl = (_cmsParametricCurvesCollection*) _cmsPluginMalloc(ContextID, sizeof(_cmsParametricCurvesCollection));
140 if (fl == NULL) return FALSE;
141
142 // Copy the parameters
143 fl ->Evaluator = Plugin ->Evaluator;
144 fl ->nFunctions = Plugin ->nFunctions;
145
146 // Make sure no mem overwrites
147 if (fl ->nFunctions > MAX_TYPES_IN_LCMS_PLUGIN)
148 fl ->nFunctions = MAX_TYPES_IN_LCMS_PLUGIN;
149
150 // Copy the data
151 memmove(fl->FunctionTypes, Plugin ->FunctionTypes, fl->nFunctions * sizeof(cmsUInt32Number));
152 memmove(fl->ParameterCount, Plugin ->ParameterCount, fl->nFunctions * sizeof(cmsUInt32Number));
153
154 // Keep linked list
155 fl ->Next = ctx->ParametricCurves;
156 ctx->ParametricCurves = fl;
157
158 // All is ok
159 return TRUE;
160}
161
162
163// Search in type list, return position or -1 if not found
164static
165int IsInSet(int Type, _cmsParametricCurvesCollection* c)
166{
167 int i;
168
169 for (i=0; i < (int) c ->nFunctions; i++)
170 if (abs(Type) == c ->FunctionTypes[i]) return i;
171
172 return -1;
173}
174
175
176// Search for the collection which contains a specific type
177static
178_cmsParametricCurvesCollection *GetParametricCurveByType(cmsContext ContextID, int Type, int* index)
179{
180 _cmsParametricCurvesCollection* c;
181 int Position;
182 _cmsCurvesPluginChunkType* ctx = ( _cmsCurvesPluginChunkType*) _cmsContextGetClientChunk(ContextID, CurvesPlugin);
183
184 for (c = ctx->ParametricCurves; c != NULL; c = c ->Next) {
185
186 Position = IsInSet(Type, c);
187
188 if (Position != -1) {
189 if (index != NULL)
190 *index = Position;
191 return c;
192 }
193 }
194 // If none found, revert for defaults
195 for (c = &DefaultCurves; c != NULL; c = c ->Next) {
196
197 Position = IsInSet(Type, c);
198
199 if (Position != -1) {
200 if (index != NULL)
201 *index = Position;
202 return c;
203 }
204 }
205
206 return NULL;
207}
208
209// Low level allocate, which takes care of memory details. nEntries may be zero, and in this case
210// no optimation curve is computed. nSegments may also be zero in the inverse case, where only the
211// optimization curve is given. Both features simultaneously is an error
212static
213cmsToneCurve* AllocateToneCurveStruct(cmsContext ContextID, cmsUInt32Number nEntries,
214 cmsUInt32Number nSegments, const cmsCurveSegment* Segments,
215 const cmsUInt16Number* Values)
216{
217 cmsToneCurve* p;
218 cmsUInt32Number i;
219
220 // We allow huge tables, which are then restricted for smoothing operations
221 if (nEntries > 65530) {
222 cmsSignalError(ContextID, cmsERROR_RANGE, "Couldn't create tone curve of more than 65530 entries");
223 return NULL;
224 }
225
226 if (nEntries == 0 && nSegments == 0) {
227 cmsSignalError(ContextID, cmsERROR_RANGE, "Couldn't create tone curve with zero segments and no table");
228 return NULL;
229 }
230
231 // Allocate all required pointers, etc.
232 p = (cmsToneCurve*) _cmsMallocZero(ContextID, sizeof(cmsToneCurve));
233 if (!p) return NULL;
234
235 // In this case, there are no segments
236 if (nSegments == 0) {
237 p ->Segments = NULL;
238 p ->Evals = NULL;
239 }
240 else {
241 p ->Segments = (cmsCurveSegment*) _cmsCalloc(ContextID, nSegments, sizeof(cmsCurveSegment));
242 if (p ->Segments == NULL) goto Error;
243
244 p ->Evals = (cmsParametricCurveEvaluator*) _cmsCalloc(ContextID, nSegments, sizeof(cmsParametricCurveEvaluator));
245 if (p ->Evals == NULL) goto Error;
246 }
247
248 p -> nSegments = nSegments;
249
250 // This 16-bit table contains a limited precision representation of the whole curve and is kept for
251 // increasing xput on certain operations.
252 if (nEntries == 0) {
253 p ->Table16 = NULL;
254 }
255 else {
256 p ->Table16 = (cmsUInt16Number*) _cmsCalloc(ContextID, nEntries, sizeof(cmsUInt16Number));
257 if (p ->Table16 == NULL) goto Error;
258 }
259
260 p -> nEntries = nEntries;
261
262 // Initialize members if requested
263 if (Values != NULL && (nEntries > 0)) {
264
265 for (i=0; i < nEntries; i++)
266 p ->Table16[i] = Values[i];
267 }
268
269 // Initialize the segments stuff. The evaluator for each segment is located and a pointer to it
270 // is placed in advance to maximize performance.
271 if (Segments != NULL && (nSegments > 0)) {
272
273 _cmsParametricCurvesCollection *c;
274
275 p ->SegInterp = (cmsInterpParams**) _cmsCalloc(ContextID, nSegments, sizeof(cmsInterpParams*));
276 if (p ->SegInterp == NULL) goto Error;
277
278 for (i=0; i < nSegments; i++) {
279
280 // Type 0 is a special marker for table-based curves
281 if (Segments[i].Type == 0)
282 p ->SegInterp[i] = _cmsComputeInterpParams(ContextID, Segments[i].nGridPoints, 1, 1, NULL, CMS_LERP_FLAGS_FLOAT);
283
284 memmove(&p ->Segments[i], &Segments[i], sizeof(cmsCurveSegment));
285
286 if (Segments[i].Type == 0 && Segments[i].SampledPoints != NULL)
287 p ->Segments[i].SampledPoints = (cmsFloat32Number*) _cmsDupMem(ContextID, Segments[i].SampledPoints, sizeof(cmsFloat32Number) * Segments[i].nGridPoints);
288 else
289 p ->Segments[i].SampledPoints = NULL;
290
291
292 c = GetParametricCurveByType(ContextID, Segments[i].Type, NULL);
293 if (c != NULL)
294 p ->Evals[i] = c ->Evaluator;
295 }
296 }
297
298 p ->InterpParams = _cmsComputeInterpParams(ContextID, p ->nEntries, 1, 1, p->Table16, CMS_LERP_FLAGS_16BITS);
299 if (p->InterpParams != NULL)
300 return p;
301
302Error:
303 if (p->SegInterp) _cmsFree(ContextID, p->SegInterp);
304 if (p -> Segments) _cmsFree(ContextID, p ->Segments);
305 if (p -> Evals) _cmsFree(ContextID, p -> Evals);
306 if (p ->Table16) _cmsFree(ContextID, p ->Table16);
307 _cmsFree(ContextID, p);
308 return NULL;
309}
310
311
312// Parametric Fn using floating point
313static
314cmsFloat64Number DefaultEvalParametricFn(cmsContext ContextID, cmsInt32Number Type, const cmsFloat64Number Params[], cmsFloat64Number R)
315{
316 cmsFloat64Number e, Val, disc;
317 cmsUNUSED_PARAMETER(ContextID);
318
319 switch (Type) {
320
321 // X = Y ^ Gamma
322 case 1:
323 if (R < 0) {
324
325 if (fabs(Params[0] - 1.0) < MATRIX_DET_TOLERANCE)
326 Val = R;
327 else
328 Val = 0;
329 }
330 else
331 Val = pow(R, Params[0]);
332 break;
333
334 // Type 1 Reversed: X = Y ^1/gamma
335 case -1:
336 if (R < 0) {
337
338 if (fabs(Params[0] - 1.0) < MATRIX_DET_TOLERANCE)
339 Val = R;
340 else
341 Val = 0;
342 }
343 else
344 {
345 if (fabs(Params[0]) < MATRIX_DET_TOLERANCE)
346 Val = PLUS_INF;
347 else
348 Val = pow(R, 1 / Params[0]);
349 }
350 break;
351
352 // CIE 122-1966
353 // Y = (aX + b)^Gamma | X >= -b/a
354 // Y = 0 | else
355 case 2:
356 {
357
358 if (fabs(Params[1]) < MATRIX_DET_TOLERANCE)
359 {
360 Val = 0;
361 }
362 else
363 {
364 disc = -Params[2] / Params[1];
365
366 if (R >= disc) {
367
368 e = Params[1] * R + Params[2];
369
370 if (e > 0)
371 Val = pow(e, Params[0]);
372 else
373 Val = 0;
374 }
375 else
376 Val = 0;
377 }
378 }
379 break;
380
381 // Type 2 Reversed
382 // X = (Y ^1/g - b) / a
383 case -2:
384 {
385 if (fabs(Params[0]) < MATRIX_DET_TOLERANCE ||
386 fabs(Params[1]) < MATRIX_DET_TOLERANCE)
387 {
388 Val = 0;
389 }
390 else
391 {
392 if (R < 0)
393 Val = 0;
394 else
395 Val = (pow(R, 1.0 / Params[0]) - Params[2]) / Params[1];
396
397 if (Val < 0)
398 Val = 0;
399 }
400 }
401 break;
402
403
404 // IEC 61966-3
405 // Y = (aX + b)^Gamma | X <= -b/a
406 // Y = c | else
407 case 3:
408 {
409 if (fabs(Params[1]) < MATRIX_DET_TOLERANCE)
410 {
411 Val = 0;
412 }
413 else
414 {
415 disc = -Params[2] / Params[1];
416 if (disc < 0)
417 disc = 0;
418
419 if (R >= disc) {
420
421 e = Params[1] * R + Params[2];
422
423 if (e > 0)
424 Val = pow(e, Params[0]) + Params[3];
425 else
426 Val = 0;
427 }
428 else
429 Val = Params[3];
430 }
431 }
432 break;
433
434
435 // Type 3 reversed
436 // X=((Y-c)^1/g - b)/a | (Y>=c)
437 // X=-b/a | (Y<c)
438 case -3:
439 {
440 if (fabs(Params[1]) < MATRIX_DET_TOLERANCE)
441 {
442 Val = 0;
443 }
444 else
445 {
446 if (R >= Params[3]) {
447
448 e = R - Params[3];
449
450 if (e > 0)
451 Val = (pow(e, 1 / Params[0]) - Params[2]) / Params[1];
452 else
453 Val = 0;
454 }
455 else {
456 Val = -Params[2] / Params[1];
457 }
458 }
459 }
460 break;
461
462
463 // IEC 61966-2.1 (sRGB)
464 // Y = (aX + b)^Gamma | X >= d
465 // Y = cX | X < d
466 case 4:
467 if (R >= Params[4]) {
468
469 e = Params[1]*R + Params[2];
470
471 if (e > 0)
472 Val = pow(e, Params[0]);
473 else
474 Val = 0;
475 }
476 else
477 Val = R * Params[3];
478 break;
479
480 // Type 4 reversed
481 // X=((Y^1/g-b)/a) | Y >= (ad+b)^g
482 // X=Y/c | Y< (ad+b)^g
483 case -4:
484 {
485 if (fabs(Params[0]) < MATRIX_DET_TOLERANCE ||
486 fabs(Params[1]) < MATRIX_DET_TOLERANCE ||
487 fabs(Params[3]) < MATRIX_DET_TOLERANCE)
488 {
489 Val = 0;
490 }
491 else
492 {
493 e = Params[1] * Params[4] + Params[2];
494 if (e < 0)
495 disc = 0;
496 else
497 disc = pow(e, Params[0]);
498
499 if (R >= disc) {
500
501 Val = (pow(R, 1.0 / Params[0]) - Params[2]) / Params[1];
502 }
503 else {
504 Val = R / Params[3];
505 }
506 }
507 }
508 break;
509
510
511 // Y = (aX + b)^Gamma + e | X >= d
512 // Y = cX + f | X < d
513 case 5:
514 if (R >= Params[4]) {
515
516 e = Params[1]*R + Params[2];
517
518 if (e > 0)
519 Val = pow(e, Params[0]) + Params[5];
520 else
521 Val = Params[5];
522 }
523 else
524 Val = R*Params[3] + Params[6];
525 break;
526
527
528 // Reversed type 5
529 // X=((Y-e)1/g-b)/a | Y >=(ad+b)^g+e), cd+f
530 // X=(Y-f)/c | else
531 case -5:
532 {
533 if (fabs(Params[1]) < MATRIX_DET_TOLERANCE ||
534 fabs(Params[3]) < MATRIX_DET_TOLERANCE)
535 {
536 Val = 0;
537 }
538 else
539 {
540 disc = Params[3] * Params[4] + Params[6];
541 if (R >= disc) {
542
543 e = R - Params[5];
544 if (e < 0)
545 Val = 0;
546 else
547 Val = (pow(e, 1.0 / Params[0]) - Params[2]) / Params[1];
548 }
549 else {
550 Val = (R - Params[6]) / Params[3];
551 }
552 }
553 }
554 break;
555
556
557 // Types 6,7,8 comes from segmented curves as described in ICCSpecRevision_02_11_06_Float.pdf
558 // Type 6 is basically identical to type 5 without d
559
560 // Y = (a * X + b) ^ Gamma + c
561 case 6:
562 e = Params[1]*R + Params[2];
563
564 if (e < 0)
565 Val = Params[3];
566 else
567 Val = pow(e, Params[0]) + Params[3];
568 break;
569
570 // ((Y - c) ^1/Gamma - b) / a
571 case -6:
572 {
573 if (fabs(Params[1]) < MATRIX_DET_TOLERANCE)
574 {
575 Val = 0;
576 }
577 else
578 {
579 e = R - Params[3];
580 if (e < 0)
581 Val = 0;
582 else
583 Val = (pow(e, 1.0 / Params[0]) - Params[2]) / Params[1];
584 }
585 }
586 break;
587
588
589 // Y = a * log (b * X^Gamma + c) + d
590 case 7:
591
592 e = Params[2] * pow(R, Params[0]) + Params[3];
593 if (e <= 0)
594 Val = Params[4];
595 else
596 Val = Params[1]*log10(e) + Params[4];
597 break;
598
599 // (Y - d) / a = log(b * X ^Gamma + c)
600 // pow(10, (Y-d) / a) = b * X ^Gamma + c
601 // pow((pow(10, (Y-d) / a) - c) / b, 1/g) = X
602 case -7:
603 {
604 if (fabs(Params[0]) < MATRIX_DET_TOLERANCE ||
605 fabs(Params[1]) < MATRIX_DET_TOLERANCE ||
606 fabs(Params[2]) < MATRIX_DET_TOLERANCE)
607 {
608 Val = 0;
609 }
610 else
611 {
612 Val = pow((pow(10.0, (R - Params[4]) / Params[1]) - Params[3]) / Params[2], 1.0 / Params[0]);
613 }
614 }
615 break;
616
617
618 //Y = a * b^(c*X+d) + e
619 case 8:
620 Val = (Params[0] * pow(Params[1], Params[2] * R + Params[3]) + Params[4]);
621 break;
622
623
624 // Y = (log((y-e) / a) / log(b) - d ) / c
625 // a=0, b=1, c=2, d=3, e=4,
626 case -8:
627
628 disc = R - Params[4];
629 if (disc < 0) Val = 0;
630 else
631 {
632 if (fabs(Params[0]) < MATRIX_DET_TOLERANCE ||
633 fabs(Params[2]) < MATRIX_DET_TOLERANCE)
634 {
635 Val = 0;
636 }
637 else
638 {
639 Val = (log(disc / Params[0]) / log(Params[1]) - Params[3]) / Params[2];
640 }
641 }
642 break;
643
644 // S-Shaped: (1 - (1-x)^1/g)^1/g
645 case 108:
646 if (fabs(Params[0]) < MATRIX_DET_TOLERANCE)
647 Val = 0;
648 else
649 Val = pow(1.0 - pow(1 - R, 1/Params[0]), 1/Params[0]);
650 break;
651
652 // y = (1 - (1-x)^1/g)^1/g
653 // y^g = (1 - (1-x)^1/g)
654 // 1 - y^g = (1-x)^1/g
655 // (1 - y^g)^g = 1 - x
656 // 1 - (1 - y^g)^g
657 case -108:
658 Val = 1 - pow(1 - pow(R, Params[0]), Params[0]);
659 break;
660
661 default:
662 // Unsupported parametric curve. Should never reach here
663 return 0;
664 }
665
666 return Val;
667}
668
669// Evaluate a segmented function for a single value. Return -Inf if no valid segment found .
670// If fn type is 0, perform an interpolation on the table
671static
672cmsFloat64Number EvalSegmentedFn(cmsContext ContextID, const cmsToneCurve *g, cmsFloat64Number R)
673{
674 int i;
675 cmsFloat32Number Out32;
676 cmsFloat64Number Out;
677
678 for (i = (int) g->nSegments - 1; i >= 0; --i) {
679
680 // Check for domain
681 if ((R > g->Segments[i].x0) && (R <= g->Segments[i].x1)) {
682
683 // Type == 0 means segment is sampled
684 if (g->Segments[i].Type == 0) {
685
686 cmsFloat32Number R1 = (cmsFloat32Number)(R - g->Segments[i].x0) / (g->Segments[i].x1 - g->Segments[i].x0);
687
688 // Setup the table (TODO: clean that)
689 g->SegInterp[i]->Table = g->Segments[i].SampledPoints;
690
691 g->SegInterp[i]->Interpolation.LerpFloat(ContextID, &R1, &Out32, g->SegInterp[i]);
692 Out = (cmsFloat64Number) Out32;
693
694 }
695 else {
696 Out = g->Evals[i](ContextID, g->Segments[i].Type, g->Segments[i].Params, R);
697 }
698
699 if (isinf(Out))
700 return PLUS_INF;
701 else
702 {
703 if (isinf(-Out))
704 return MINUS_INF;
705 }
706
707 return Out;
708 }
709 }
710
711 return MINUS_INF;
712}
713
714// Access to estimated low-res table
715cmsUInt32Number CMSEXPORT cmsGetToneCurveEstimatedTableEntries(cmsContext ContextID, const cmsToneCurve* t)
716{
717 cmsUNUSED_PARAMETER(ContextID);
718 _cmsAssert(t != NULL);
719 return t ->nEntries;
720}
721
722const cmsUInt16Number* CMSEXPORT cmsGetToneCurveEstimatedTable(cmsContext ContextID, const cmsToneCurve* t)
723{
724 cmsUNUSED_PARAMETER(ContextID);
725 _cmsAssert(t != NULL);
726 return t ->Table16;
727}
728
729
730// Create an empty gamma curve, by using tables. This specifies only the limited-precision part, and leaves the
731// floating point description empty.
732cmsToneCurve* CMSEXPORT cmsBuildTabulatedToneCurve16(cmsContext ContextID, cmsUInt32Number nEntries, const cmsUInt16Number Values[])
733{
734 return AllocateToneCurveStruct(ContextID, nEntries, 0, NULL, Values);
735}
736
737static
738cmsUInt32Number EntriesByGamma(cmsFloat64Number Gamma)
739{
740 if (fabs(Gamma - 1.0) < 0.001) return 2;
741 return 4096;
742}
743
744
745// Create a segmented gamma, fill the table
746cmsToneCurve* CMSEXPORT cmsBuildSegmentedToneCurve(cmsContext ContextID,
747 cmsUInt32Number nSegments, const cmsCurveSegment Segments[])
748{
749 cmsUInt32Number i;
750 cmsFloat64Number R, Val;
751 cmsToneCurve* g;
752 cmsUInt32Number nGridPoints = 4096;
753
754 _cmsAssert(Segments != NULL);
755
756 // Optimizatin for identity curves.
757 if (nSegments == 1 && Segments[0].Type == 1) {
758
759 nGridPoints = EntriesByGamma(Segments[0].Params[0]);
760 }
761
762 g = AllocateToneCurveStruct(ContextID, nGridPoints, nSegments, Segments, NULL);
763 if (g == NULL) return NULL;
764
765 // Once we have the floating point version, we can approximate a 16 bit table of 4096 entries
766 // for performance reasons. This table would normally not be used except on 8/16 bits transforms.
767 for (i = 0; i < nGridPoints; i++) {
768
769 R = (cmsFloat64Number) i / (nGridPoints-1);
770
771 Val = EvalSegmentedFn(ContextID, g, R);
772
773 // Round and saturate
774 g ->Table16[i] = _cmsQuickSaturateWord(Val * 65535.0);
775 }
776
777 return g;
778}
779
780// Use a segmented curve to store the floating point table
781cmsToneCurve* CMSEXPORT cmsBuildTabulatedToneCurveFloat(cmsContext ContextID, cmsUInt32Number nEntries, const cmsFloat32Number values[])
782{
783 cmsCurveSegment Seg[3];
784
785 // A segmented tone curve should have function segments in the first and last positions
786 // Initialize segmented curve part up to 0 to constant value = samples[0]
787 Seg[0].x0 = MINUS_INF;
788 Seg[0].x1 = 0;
789 Seg[0].Type = 6;
790
791 Seg[0].Params[0] = 1;
792 Seg[0].Params[1] = 0;
793 Seg[0].Params[2] = 0;
794 Seg[0].Params[3] = values[0];
795 Seg[0].Params[4] = 0;
796
797 // From zero to 1
798 Seg[1].x0 = 0;
799 Seg[1].x1 = 1.0;
800 Seg[1].Type = 0;
801
802 Seg[1].nGridPoints = nEntries;
803 Seg[1].SampledPoints = (cmsFloat32Number*) values;
804
805 // Final segment is constant = lastsample
806 Seg[2].x0 = 1.0;
807 Seg[2].x1 = PLUS_INF;
808 Seg[2].Type = 6;
809
810 Seg[2].Params[0] = 1;
811 Seg[2].Params[1] = 0;
812 Seg[2].Params[2] = 0;
813 Seg[2].Params[3] = values[nEntries-1];
814 Seg[2].Params[4] = 0;
815
816
817 return cmsBuildSegmentedToneCurve(ContextID, 3, Seg);
818}
819
820// Parametric curves
821//
822// Parameters goes as: Curve, a, b, c, d, e, f
823// Type is the ICC type +1
824// if type is negative, then the curve is analytically inverted
825cmsToneCurve* CMSEXPORT cmsBuildParametricToneCurve(cmsContext ContextID, cmsInt32Number Type, const cmsFloat64Number Params[])
826{
827 cmsCurveSegment Seg0;
828 int Pos = 0;
829 cmsUInt32Number size;
830 _cmsParametricCurvesCollection* c = GetParametricCurveByType(ContextID, Type, &Pos);
831
832 _cmsAssert(Params != NULL);
833
834 if (c == NULL) {
835 cmsSignalError(ContextID, cmsERROR_UNKNOWN_EXTENSION, "Invalid parametric curve type %d", Type);
836 return NULL;
837 }
838
839 memset(&Seg0, 0, sizeof(Seg0));
840
841 Seg0.x0 = MINUS_INF;
842 Seg0.x1 = PLUS_INF;
843 Seg0.Type = Type;
844
845 size = c->ParameterCount[Pos] * sizeof(cmsFloat64Number);
846 memmove(Seg0.Params, Params, size);
847
848 return cmsBuildSegmentedToneCurve(ContextID, 1, &Seg0);
849}
850
851
852
853// Build a gamma table based on gamma constant
854cmsToneCurve* CMSEXPORT cmsBuildGamma(cmsContext ContextID, cmsFloat64Number Gamma)
855{
856 return cmsBuildParametricToneCurve(ContextID, 1, &Gamma);
857}
858
859
860// Free all memory taken by the gamma curve
861void CMSEXPORT cmsFreeToneCurve(cmsContext ContextID, cmsToneCurve* Curve)
862{
863 if (Curve == NULL) return;
864
865 _cmsFreeInterpParams(ContextID, Curve ->InterpParams);
866
867 if (Curve -> Table16)
868 _cmsFree(ContextID, Curve ->Table16);
869
870 if (Curve ->Segments) {
871
872 cmsUInt32Number i;
873
874 for (i=0; i < Curve ->nSegments; i++) {
875
876 if (Curve ->Segments[i].SampledPoints) {
877 _cmsFree(ContextID, Curve ->Segments[i].SampledPoints);
878 }
879
880 if (Curve ->SegInterp[i] != 0)
881 _cmsFreeInterpParams(ContextID, Curve->SegInterp[i]);
882 }
883
884 _cmsFree(ContextID, Curve ->Segments);
885 _cmsFree(ContextID, Curve ->SegInterp);
886 }
887
888 if (Curve -> Evals)
889 _cmsFree(ContextID, Curve -> Evals);
890
891 if (Curve) _cmsFree(ContextID, Curve);
892}
893
894// Utility function, free 3 gamma tables
895void CMSEXPORT cmsFreeToneCurveTriple(cmsContext ContextID, cmsToneCurve* Curve[3])
896{
897
898 _cmsAssert(Curve != NULL);
899
900 if (Curve[0] != NULL) cmsFreeToneCurve(ContextID, Curve[0]);
901 if (Curve[1] != NULL) cmsFreeToneCurve(ContextID, Curve[1]);
902 if (Curve[2] != NULL) cmsFreeToneCurve(ContextID, Curve[2]);
903
904 Curve[0] = Curve[1] = Curve[2] = NULL;
905}
906
907
908// Duplicate a gamma table
909cmsToneCurve* CMSEXPORT cmsDupToneCurve(cmsContext ContextID, const cmsToneCurve* In)
910{
911 if (In == NULL) return NULL;
912
913 return AllocateToneCurveStruct(ContextID, In ->nEntries, In ->nSegments, In ->Segments, In ->Table16);
914}
915
916// Joins two curves for X and Y. Curves should be monotonic.
917// We want to get
918//
919// y = Y^-1(X(t))
920//
921cmsToneCurve* CMSEXPORT cmsJoinToneCurve(cmsContext ContextID,
922 const cmsToneCurve* X,
923 const cmsToneCurve* Y, cmsUInt32Number nResultingPoints)
924{
925 cmsToneCurve* out = NULL;
926 cmsToneCurve* Yreversed = NULL;
927 cmsFloat32Number t, x;
928 cmsFloat32Number* Res = NULL;
929 cmsUInt32Number i;
930
931
932 _cmsAssert(X != NULL);
933 _cmsAssert(Y != NULL);
934
935 Yreversed = cmsReverseToneCurveEx(ContextID, nResultingPoints, Y);
936 if (Yreversed == NULL) goto Error;
937
938 Res = (cmsFloat32Number*) _cmsCalloc(ContextID, nResultingPoints, sizeof(cmsFloat32Number));
939 if (Res == NULL) goto Error;
940
941 //Iterate
942 for (i=0; i < nResultingPoints; i++) {
943
944 t = (cmsFloat32Number) i / (nResultingPoints-1);
945 x = cmsEvalToneCurveFloat(ContextID, X, t);
946 Res[i] = cmsEvalToneCurveFloat(ContextID, Yreversed, x);
947 }
948
949 // Allocate space for output
950 out = cmsBuildTabulatedToneCurveFloat(ContextID, nResultingPoints, Res);
951
952Error:
953
954 if (Res != NULL) _cmsFree(ContextID, Res);
955 if (Yreversed != NULL) cmsFreeToneCurve(ContextID, Yreversed);
956
957 return out;
958}
959
960
961
962// Get the surrounding nodes. This is tricky on non-monotonic tables
963static
964int GetInterval(cmsFloat64Number In, const cmsUInt16Number LutTable[], const struct _cms_interp_struc* p)
965{
966 int i;
967 int y0, y1;
968
969 // A 1 point table is not allowed
970 if (p -> Domain[0] < 1) return -1;
971
972 // Let's see if ascending or descending.
973 if (LutTable[0] < LutTable[p ->Domain[0]]) {
974
975 // Table is overall ascending
976 for (i = (int) p->Domain[0] - 1; i >= 0; --i) {
977
978 y0 = LutTable[i];
979 y1 = LutTable[i+1];
980
981 if (y0 <= y1) { // Increasing
982 if (In >= y0 && In <= y1) return i;
983 }
984 else
985 if (y1 < y0) { // Decreasing
986 if (In >= y1 && In <= y0) return i;
987 }
988 }
989 }
990 else {
991 // Table is overall descending
992 for (i=0; i < (int) p -> Domain[0]; i++) {
993
994 y0 = LutTable[i];
995 y1 = LutTable[i+1];
996
997 if (y0 <= y1) { // Increasing
998 if (In >= y0 && In <= y1) return i;
999 }
1000 else
1001 if (y1 < y0) { // Decreasing
1002 if (In >= y1 && In <= y0) return i;
1003 }
1004 }
1005 }
1006
1007 return -1;
1008}
1009
1010// Reverse a gamma table
1011cmsToneCurve* CMSEXPORT cmsReverseToneCurveEx(cmsContext ContextID, cmsUInt32Number nResultSamples, const cmsToneCurve* InCurve)
1012{
1013 cmsToneCurve *out;
1014 cmsFloat64Number a = 0, b = 0, y, x1, y1, x2, y2;
1015 int i, j;
1016 int Ascending;
1017
1018 _cmsAssert(InCurve != NULL);
1019
1020 // Try to reverse it analytically whatever possible
1021
1022 if (InCurve ->nSegments == 1 && InCurve ->Segments[0].Type > 0 &&
1023 /* InCurve -> Segments[0].Type <= 5 */
1024 GetParametricCurveByType(ContextID, InCurve ->Segments[0].Type, NULL) != NULL) {
1025
1026 return cmsBuildParametricToneCurve(ContextID,
1027 -(InCurve -> Segments[0].Type),
1028 InCurve -> Segments[0].Params);
1029 }
1030
1031 // Nope, reverse the table.
1032 out = cmsBuildTabulatedToneCurve16(ContextID, nResultSamples, NULL);
1033 if (out == NULL)
1034 return NULL;
1035
1036 // We want to know if this is an ascending or descending table
1037 Ascending = !cmsIsToneCurveDescending(ContextID, InCurve);
1038
1039 // Iterate across Y axis
1040 for (i=0; i < (int) nResultSamples; i++) {
1041
1042 y = (cmsFloat64Number) i * 65535.0 / (nResultSamples - 1);
1043
1044 // Find interval in which y is within.
1045 j = GetInterval(y, InCurve->Table16, InCurve->InterpParams);
1046 if (j >= 0) {
1047
1048
1049 // Get limits of interval
1050 x1 = InCurve ->Table16[j];
1051 x2 = InCurve ->Table16[j+1];
1052
1053 y1 = (cmsFloat64Number) (j * 65535.0) / (InCurve ->nEntries - 1);
1054 y2 = (cmsFloat64Number) ((j+1) * 65535.0 ) / (InCurve ->nEntries - 1);
1055
1056 // If collapsed, then use any
1057 if (x1 == x2) {
1058
1059 out ->Table16[i] = _cmsQuickSaturateWord(Ascending ? y2 : y1);
1060 continue;
1061
1062 } else {
1063
1064 // Interpolate
1065 a = (y2 - y1) / (x2 - x1);
1066 b = y2 - a * x2;
1067 }
1068 }
1069
1070 out ->Table16[i] = _cmsQuickSaturateWord(a* y + b);
1071 }
1072
1073
1074 return out;
1075}
1076
1077// Reverse a gamma table
1078cmsToneCurve* CMSEXPORT cmsReverseToneCurve(cmsContext ContextID, const cmsToneCurve* InGamma)
1079{
1080 _cmsAssert(InGamma != NULL);
1081
1082 return cmsReverseToneCurveEx(ContextID, 4096, InGamma);
1083}
1084
1085// From: Eilers, P.H.C. (1994) Smoothing and interpolation with finite
1086// differences. in: Graphic Gems IV, Heckbert, P.S. (ed.), Academic press.
1087//
1088// Smoothing and interpolation with second differences.
1089//
1090// Input: weights (w), data (y): vector from 1 to m.
1091// Input: smoothing parameter (lambda), length (m).
1092// Output: smoothed vector (z): vector from 1 to m.
1093
1094static
1095cmsBool smooth2(cmsContext ContextID, cmsFloat32Number w[], cmsFloat32Number y[],
1096 cmsFloat32Number z[], cmsFloat32Number lambda, int m)
1097{
1098 int i, i1, i2;
1099 cmsFloat32Number *c, *d, *e;
1100 cmsBool st;
1101
1102
1103 c = (cmsFloat32Number*) _cmsCalloc(ContextID, MAX_NODES_IN_CURVE, sizeof(cmsFloat32Number));
1104 d = (cmsFloat32Number*) _cmsCalloc(ContextID, MAX_NODES_IN_CURVE, sizeof(cmsFloat32Number));
1105 e = (cmsFloat32Number*) _cmsCalloc(ContextID, MAX_NODES_IN_CURVE, sizeof(cmsFloat32Number));
1106
1107 if (c != NULL && d != NULL && e != NULL) {
1108
1109
1110 d[1] = w[1] + lambda;
1111 c[1] = -2 * lambda / d[1];
1112 e[1] = lambda /d[1];
1113 z[1] = w[1] * y[1];
1114 d[2] = w[2] + 5 * lambda - d[1] * c[1] * c[1];
1115 c[2] = (-4 * lambda - d[1] * c[1] * e[1]) / d[2];
1116 e[2] = lambda / d[2];
1117 z[2] = w[2] * y[2] - c[1] * z[1];
1118
1119 for (i = 3; i < m - 1; i++) {
1120 i1 = i - 1; i2 = i - 2;
1121 d[i]= w[i] + 6 * lambda - c[i1] * c[i1] * d[i1] - e[i2] * e[i2] * d[i2];
1122 c[i] = (-4 * lambda -d[i1] * c[i1] * e[i1])/ d[i];
1123 e[i] = lambda / d[i];
1124 z[i] = w[i] * y[i] - c[i1] * z[i1] - e[i2] * z[i2];
1125 }
1126
1127 i1 = m - 2; i2 = m - 3;
1128
1129 d[m - 1] = w[m - 1] + 5 * lambda -c[i1] * c[i1] * d[i1] - e[i2] * e[i2] * d[i2];
1130 c[m - 1] = (-2 * lambda - d[i1] * c[i1] * e[i1]) / d[m - 1];
1131 z[m - 1] = w[m - 1] * y[m - 1] - c[i1] * z[i1] - e[i2] * z[i2];
1132 i1 = m - 1; i2 = m - 2;
1133
1134 d[m] = w[m] + lambda - c[i1] * c[i1] * d[i1] - e[i2] * e[i2] * d[i2];
1135 z[m] = (w[m] * y[m] - c[i1] * z[i1] - e[i2] * z[i2]) / d[m];
1136 z[m - 1] = z[m - 1] / d[m - 1] - c[m - 1] * z[m];
1137
1138 for (i = m - 2; 1<= i; i--)
1139 z[i] = z[i] / d[i] - c[i] * z[i + 1] - e[i] * z[i + 2];
1140
1141 st = TRUE;
1142 }
1143 else st = FALSE;
1144
1145 if (c != NULL) _cmsFree(ContextID, c);
1146 if (d != NULL) _cmsFree(ContextID, d);
1147 if (e != NULL) _cmsFree(ContextID, e);
1148
1149 return st;
1150}
1151
1152// Smooths a curve sampled at regular intervals.
1153cmsBool CMSEXPORT cmsSmoothToneCurve(cmsContext ContextID, cmsToneCurve* Tab, cmsFloat64Number lambda)
1154{
1155 cmsBool SuccessStatus = TRUE;
1156 cmsFloat32Number *w, *y, *z;
1157 cmsUInt32Number i, nItems, Zeros, Poles;
1158
1159 if (Tab != NULL && Tab->InterpParams != NULL)
1160 {
1161 if (!cmsIsToneCurveLinear(ContextID, Tab)) // Only non-linear curves need smoothing
1162 {
1163 nItems = Tab->nEntries;
1164 if (nItems < MAX_NODES_IN_CURVE)
1165 {
1166 // Allocate one more item than needed
1167 w = (cmsFloat32Number *)_cmsCalloc(ContextID, nItems + 1, sizeof(cmsFloat32Number));
1168 y = (cmsFloat32Number *)_cmsCalloc(ContextID, nItems + 1, sizeof(cmsFloat32Number));
1169 z = (cmsFloat32Number *)_cmsCalloc(ContextID, nItems + 1, sizeof(cmsFloat32Number));
1170
1171 if (w != NULL && y != NULL && z != NULL) // Ensure no memory allocation failure
1172 {
1173 memset(w, 0, (nItems + 1) * sizeof(cmsFloat32Number));
1174 memset(y, 0, (nItems + 1) * sizeof(cmsFloat32Number));
1175 memset(z, 0, (nItems + 1) * sizeof(cmsFloat32Number));
1176
1177 for (i = 0; i < nItems; i++)
1178 {
1179 y[i + 1] = (cmsFloat32Number)Tab->Table16[i];
1180 w[i + 1] = 1.0;
1181 }
1182
1183 if (smooth2(ContextID, w, y, z, (cmsFloat32Number)lambda, (int)nItems))
1184 {
1185 // Do some reality - checking...
1186
1187 Zeros = Poles = 0;
1188 for (i = nItems; i > 1; --i)
1189 {
1190 if (z[i] == 0.) Zeros++;
1191 if (z[i] >= 65535.) Poles++;
1192 if (z[i] < z[i - 1])
1193 {
1194 cmsSignalError(ContextID, cmsERROR_RANGE, "cmsSmoothToneCurve: Non-Monotonic.");
1195 SuccessStatus = FALSE;
1196 break;
1197 }
1198 }
1199
1200 if (SuccessStatus && Zeros > (nItems / 3))
1201 {
1202 cmsSignalError(ContextID, cmsERROR_RANGE, "cmsSmoothToneCurve: Degenerated, mostly zeros.");
1203 SuccessStatus = FALSE;
1204 }
1205
1206 if (SuccessStatus && Poles > (nItems / 3))
1207 {
1208 cmsSignalError(ContextID, cmsERROR_RANGE, "cmsSmoothToneCurve: Degenerated, mostly poles.");
1209 SuccessStatus = FALSE;
1210 }
1211
1212 if (SuccessStatus) // Seems ok
1213 {
1214 for (i = 0; i < nItems; i++)
1215 {
1216 // Clamp to cmsUInt16Number
1217 Tab->Table16[i] = _cmsQuickSaturateWord(z[i + 1]);
1218 }
1219 }
1220 }
1221 else // Could not smooth
1222 {
1223 cmsSignalError(ContextID, cmsERROR_RANGE, "cmsSmoothToneCurve: Function smooth2 failed.");
1224 SuccessStatus = FALSE;
1225 }
1226 }
1227 else // One or more buffers could not be allocated
1228 {
1229 cmsSignalError(ContextID, cmsERROR_RANGE, "cmsSmoothToneCurve: Could not allocate memory.");
1230 SuccessStatus = FALSE;
1231 }
1232
1233 if (z != NULL)
1234 _cmsFree(ContextID, z);
1235
1236 if (y != NULL)
1237 _cmsFree(ContextID, y);
1238
1239 if (w != NULL)
1240 _cmsFree(ContextID, w);
1241 }
1242 else // too many items in the table
1243 {
1244 cmsSignalError(ContextID, cmsERROR_RANGE, "cmsSmoothToneCurve: Too many points.");
1245 SuccessStatus = FALSE;
1246 }
1247 }
1248 }
1249 else // Tab parameter or Tab->InterpParams is NULL
1250 {
1251 // Can't signal an error here since the ContextID is not known at this point
1252 SuccessStatus = FALSE;
1253 }
1254
1255 return SuccessStatus;
1256}
1257
1258// Is a table linear? Do not use parametric since we cannot guarantee some weird parameters resulting
1259// in a linear table. This way assures it is linear in 12 bits, which should be enough in most cases.
1260cmsBool CMSEXPORT cmsIsToneCurveLinear(cmsContext ContextID, const cmsToneCurve* Curve)
1261{
1262 int i;
1263 int diff;
1264 cmsUNUSED_PARAMETER(ContextID);
1265
1266 _cmsAssert(Curve != NULL);
1267
1268 for (i=0; i < (int) Curve ->nEntries; i++) {
1269
1270 diff = abs((int) Curve->Table16[i] - (int) _cmsQuantizeVal(i, Curve ->nEntries));
1271 if (diff > 0x0f)
1272 return FALSE;
1273 }
1274
1275 return TRUE;
1276}
1277
1278// Same, but for monotonicity
1279cmsBool CMSEXPORT cmsIsToneCurveMonotonic(cmsContext ContextID, const cmsToneCurve* t)
1280{
1281 cmsUInt32Number n;
1282 int i, last;
1283 cmsBool lDescending;
1284
1285 _cmsAssert(t != NULL);
1286
1287 // Degenerated curves are monotonic? Ok, let's pass them
1288 n = t ->nEntries;
1289 if (n < 2) return TRUE;
1290
1291 // Curve direction
1292 lDescending = cmsIsToneCurveDescending(ContextID, t);
1293
1294 if (lDescending) {
1295
1296 last = t ->Table16[0];
1297
1298 for (i = 1; i < (int) n; i++) {
1299
1300 if (t ->Table16[i] - last > 2) // We allow some ripple
1301 return FALSE;
1302 else
1303 last = t ->Table16[i];
1304
1305 }
1306 }
1307 else {
1308
1309 last = t ->Table16[n-1];
1310
1311 for (i = (int) n - 2; i >= 0; --i) {
1312
1313 if (t ->Table16[i] - last > 2)
1314 return FALSE;
1315 else
1316 last = t ->Table16[i];
1317
1318 }
1319 }
1320
1321 return TRUE;
1322}
1323
1324// Same, but for descending tables
1325cmsBool CMSEXPORT cmsIsToneCurveDescending(cmsContext ContextID, const cmsToneCurve* t)
1326{
1327 _cmsAssert(t != NULL);
1328 cmsUNUSED_PARAMETER(ContextID);
1329
1330 return t ->Table16[0] > t ->Table16[t ->nEntries-1];
1331}
1332
1333
1334// Another info fn: is out gamma table multisegment?
1335cmsBool CMSEXPORT cmsIsToneCurveMultisegment(cmsContext ContextID, const cmsToneCurve* t)
1336{
1337 _cmsAssert(t != NULL);
1338 cmsUNUSED_PARAMETER(ContextID);
1339
1340 return t -> nSegments > 1;
1341}
1342
1343cmsInt32Number CMSEXPORT cmsGetToneCurveParametricType(cmsContext ContextID, const cmsToneCurve* t)
1344{
1345 _cmsAssert(t != NULL);
1346 cmsUNUSED_PARAMETER(ContextID);
1347
1348 if (t -> nSegments != 1) return 0;
1349 return t ->Segments[0].Type;
1350}
1351
1352// We need accuracy this time
1353cmsFloat32Number CMSEXPORT cmsEvalToneCurveFloat(cmsContext ContextID, const cmsToneCurve* Curve, cmsFloat32Number v)
1354{
1355 _cmsAssert(Curve != NULL);
1356
1357 // Check for 16 bits table. If so, this is a limited-precision tone curve
1358 if (Curve ->nSegments == 0) {
1359
1360 cmsUInt16Number In, Out;
1361
1362 In = (cmsUInt16Number) _cmsQuickSaturateWord(v * 65535.0);
1363 Out = cmsEvalToneCurve16(ContextID, Curve, In);
1364
1365 return (cmsFloat32Number) (Out / 65535.0);
1366 }
1367
1368 return (cmsFloat32Number) EvalSegmentedFn(ContextID, Curve, v);
1369}
1370
1371// We need xput over here
1372cmsUInt16Number CMSEXPORT cmsEvalToneCurve16(cmsContext ContextID, const cmsToneCurve* Curve, cmsUInt16Number v)
1373{
1374 cmsUInt16Number out;
1375
1376 _cmsAssert(Curve != NULL);
1377
1378 Curve ->InterpParams ->Interpolation.Lerp16(ContextID, &v, &out, Curve ->InterpParams);
1379 return out;
1380}
1381
1382
1383// Least squares fitting.
1384// A mathematical procedure for finding the best-fitting curve to a given set of points by
1385// minimizing the sum of the squares of the offsets ("the residuals") of the points from the curve.
1386// The sum of the squares of the offsets is used instead of the offset absolute values because
1387// this allows the residuals to be treated as a continuous differentiable quantity.
1388//
1389// y = f(x) = x ^ g
1390//
1391// R = (yi - (xi^g))
1392// R2 = (yi - (xi^g))2
1393// SUM R2 = SUM (yi - (xi^g))2
1394//
1395// dR2/dg = -2 SUM x^g log(x)(y - x^g)
1396// solving for dR2/dg = 0
1397//
1398// g = 1/n * SUM(log(y) / log(x))
1399
1400cmsFloat64Number CMSEXPORT cmsEstimateGamma(cmsContext ContextID, const cmsToneCurve* t, cmsFloat64Number Precision)
1401{
1402 cmsFloat64Number gamma, sum, sum2;
1403 cmsFloat64Number n, x, y, Std;
1404 cmsUInt32Number i;
1405
1406 _cmsAssert(t != NULL);
1407
1408 sum = sum2 = n = 0;
1409
1410 // Excluding endpoints
1411 for (i=1; i < (MAX_NODES_IN_CURVE-1); i++) {
1412
1413 x = (cmsFloat64Number) i / (MAX_NODES_IN_CURVE-1);
1414 y = (cmsFloat64Number) cmsEvalToneCurveFloat(ContextID, t, (cmsFloat32Number) x);
1415
1416 // Avoid 7% on lower part to prevent
1417 // artifacts due to linear ramps
1418
1419 if (y > 0. && y < 1. && x > 0.07) {
1420
1421 gamma = log(y) / log(x);
1422 sum += gamma;
1423 sum2 += gamma * gamma;
1424 n++;
1425 }
1426 }
1427
1428 // Take a look on SD to see if gamma isn't exponential at all
1429 Std = sqrt((n * sum2 - sum * sum) / (n*(n-1)));
1430
1431 if (Std > Precision)
1432 return -1.0;
1433
1434 return (sum / n); // The mean
1435}
1436