dnrops
/
gimp
mirror of https://github.com/GNOME/gimp.git
1
0
Fork 0
Code Issues Packages Projects Releases Wiki Activity
gimp/libgimpcolor/gimpcolortransform.c

652 lines
21 KiB
C
Raw Normal View History

libgimpcolor: add new object GimpColorTransform which encapsulates a cmsHTRANSFORM and does all the pixel format conversion magic. It has API to create transforms and proofing transforms, and to convert pixels arrays and GeglBuffers. Before, each place which has a transform had to keep around the transform and its input and output Babl formats, and had to implement lots of stuff itself. Now all that lives in GimpColorTransform, removing lots of logic from many places, and pretty much removing lcms from the public API entirely. This removes including <lcms2.h>, LCMS_LIBS and LCMS_CFLAGS from almost all directories and potentially allows to replace lcms by something else.
2016-05-26 03:35:54 +08:00
/* LIBGIMP - The GIMP Library
* Copyright (C) 1995-1997 Spencer Kimball and Peter Mattis
*
* gimpcolortransform.c
* Copyright (C) 2014 Michael Natterer <mitch@gimp.org>
* Elle Stone <ellestone@ninedegreesbelow.com>
* Øyvind Kolås <pippin@gimp.org>
*
* 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 3 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
* Library 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, see
Change the license URL from http://www.gnu.org/licenses/ to https://
2018-07-12 05:27:07 +08:00
* <https://www.gnu.org/licenses/>.
libgimpcolor: add new object GimpColorTransform which encapsulates a cmsHTRANSFORM and does all the pixel format conversion magic. It has API to create transforms and proofing transforms, and to convert pixels arrays and GeglBuffers. Before, each place which has a transform had to keep around the transform and its input and output Babl formats, and had to implement lots of stuff itself. Now all that lives in GimpColorTransform, removing lots of logic from many places, and pretty much removing lcms from the public API entirely. This removes including <lcms2.h>, LCMS_LIBS and LCMS_CFLAGS from almost all directories and potentially allows to replace lcms by something else.
2016-05-26 03:35:54 +08:00
*/
#include "config.h"
#include <string.h>
#include <lcms2.h>
#include <gio/gio.h>
#include <gegl.h>
#include "libgimpbase/gimpbase.h"
#include "libgimpconfig/gimpconfig.h"
#include "gimpcolortypes.h"
#include "gimpcolorprofile.h"
#include "gimpcolortransform.h"
#include "libgimp/libgimp-intl.h"
/**
* SECTION: gimpcolortransform
* @title: GimpColorTransform
* @short_description: Definitions and Functions relating to LCMS.
*
* Definitions and Functions relating to LCMS.
**/
/**
* GimpColorTransform:
*
* Simply a typedef to #gpointer, but actually is a cmsHTRANSFORM. It's
* used in public GIMP APIs in order to avoid having to include LCMS
* headers.
**/
enum
{
PROGRESS,
LAST_SIGNAL
};
struct _GimpColorTransformPrivate
{
GimpColorProfile *src_profile;
const Babl *src_format;
GimpColorProfile *dest_profile;
const Babl *dest_format;
cmsHTRANSFORM transform;
Bug 795300 - ICC profile conversion artifacts and gamma problems Cache the fish in GimpColorTransform as suggested by Massimo.
2018-04-19 00:49:47 +08:00
const Babl *fish;
libgimpcolor: add new object GimpColorTransform which encapsulates a cmsHTRANSFORM and does all the pixel format conversion magic. It has API to create transforms and proofing transforms, and to convert pixels arrays and GeglBuffers. Before, each place which has a transform had to keep around the transform and its input and output Babl formats, and had to implement lots of stuff itself. Now all that lives in GimpColorTransform, removing lots of logic from many places, and pretty much removing lcms from the public API entirely. This removes including <lcms2.h>, LCMS_LIBS and LCMS_CFLAGS from almost all directories and potentially allows to replace lcms by something else.
2016-05-26 03:35:54 +08:00
};
static void gimp_color_transform_finalize (GObject *object);
G_DEFINE_TYPE (GimpColorTransform, gimp_color_transform,
G_TYPE_OBJECT);
#define parent_class gimp_color_transform_parent_class
static guint gimp_color_transform_signals[LAST_SIGNAL] = { 0 };
libgimpcolor: return a NULL transform if cmsCreate[Proofing]Transform() fails It doesn't really "fail", we check if an error was logged during transform creation and get rid of the transform even if lcms returned a pointer. This totally sucks but is better than crashing on a corrupt transform.
2016-05-30 18:48:54 +08:00
static gchar *lcms_last_error = NULL;
static void
lcms_error_clear (void)
{
if (lcms_last_error)
{
g_free (lcms_last_error);
lcms_last_error = NULL;
}
}
static void
lcms_error_handler (cmsContext ContextID,
cmsUInt32Number ErrorCode,
const gchar *text)
{
lcms_error_clear ();
lcms_last_error = g_strdup_printf ("lcms2 error %d: %s", ErrorCode, text);
}
libgimpcolor: add new object GimpColorTransform which encapsulates a cmsHTRANSFORM and does all the pixel format conversion magic. It has API to create transforms and proofing transforms, and to convert pixels arrays and GeglBuffers. Before, each place which has a transform had to keep around the transform and its input and output Babl formats, and had to implement lots of stuff itself. Now all that lives in GimpColorTransform, removing lots of logic from many places, and pretty much removing lcms from the public API entirely. This removes including <lcms2.h>, LCMS_LIBS and LCMS_CFLAGS from almost all directories and potentially allows to replace lcms by something else.
2016-05-26 03:35:54 +08:00
static void
gimp_color_transform_class_init (GimpColorTransformClass *klass)
{
GObjectClass *object_class = G_OBJECT_CLASS (klass);
object_class->finalize = gimp_color_transform_finalize;
gimp_color_transform_signals[PROGRESS] =
g_signal_new ("progress",
G_OBJECT_CLASS_TYPE (object_class),
G_SIGNAL_RUN_FIRST,
G_STRUCT_OFFSET (GimpColorTransformClass,
progress),
NULL, NULL,
g_cclosure_marshal_VOID__DOUBLE,
G_TYPE_NONE, 1,
G_TYPE_DOUBLE);
g_type_class_add_private (klass, sizeof (GimpColorTransformPrivate));
libgimpcolor: return a NULL transform if cmsCreate[Proofing]Transform() fails It doesn't really "fail", we check if an error was logged during transform creation and get rid of the transform even if lcms returned a pointer. This totally sucks but is better than crashing on a corrupt transform.
2016-05-30 18:48:54 +08:00
cmsSetLogErrorHandler (lcms_error_handler);
libgimpcolor: add new object GimpColorTransform which encapsulates a cmsHTRANSFORM and does all the pixel format conversion magic. It has API to create transforms and proofing transforms, and to convert pixels arrays and GeglBuffers. Before, each place which has a transform had to keep around the transform and its input and output Babl formats, and had to implement lots of stuff itself. Now all that lives in GimpColorTransform, removing lots of logic from many places, and pretty much removing lcms from the public API entirely. This removes including <lcms2.h>, LCMS_LIBS and LCMS_CFLAGS from almost all directories and potentially allows to replace lcms by something else.
2016-05-26 03:35:54 +08:00
}
static void
gimp_color_transform_init (GimpColorTransform *transform)
{
transform->priv = G_TYPE_INSTANCE_GET_PRIVATE (transform,
GIMP_TYPE_COLOR_TRANSFORM,
GimpColorTransformPrivate);
}
static void
gimp_color_transform_finalize (GObject *object)
{
GimpColorTransform *transform = GIMP_COLOR_TRANSFORM (object);
libgimp*: use some g_clear_object() and g_clear_pointer()
2017-09-03 21:14:51 +08:00
g_clear_object (&transform->priv->src_profile);
g_clear_object (&transform->priv->dest_profile);
libgimpcolor: add new object GimpColorTransform which encapsulates a cmsHTRANSFORM and does all the pixel format conversion magic. It has API to create transforms and proofing transforms, and to convert pixels arrays and GeglBuffers. Before, each place which has a transform had to keep around the transform and its input and output Babl formats, and had to implement lots of stuff itself. Now all that lives in GimpColorTransform, removing lots of logic from many places, and pretty much removing lcms from the public API entirely. This removes including <lcms2.h>, LCMS_LIBS and LCMS_CFLAGS from almost all directories and potentially allows to replace lcms by something else.
2016-05-26 03:35:54 +08:00
libgimp*: use some g_clear_object() and g_clear_pointer()
2017-09-03 21:14:51 +08:00
g_clear_pointer (&transform->priv->transform, cmsDeleteTransform);
libgimpcolor: add new object GimpColorTransform which encapsulates a cmsHTRANSFORM and does all the pixel format conversion magic. It has API to create transforms and proofing transforms, and to convert pixels arrays and GeglBuffers. Before, each place which has a transform had to keep around the transform and its input and output Babl formats, and had to implement lots of stuff itself. Now all that lives in GimpColorTransform, removing lots of logic from many places, and pretty much removing lcms from the public API entirely. This removes including <lcms2.h>, LCMS_LIBS and LCMS_CFLAGS from almost all directories and potentially allows to replace lcms by something else.
2016-05-26 03:35:54 +08:00
G_OBJECT_CLASS (parent_class)->finalize (object);
}
/**
* gimp_color_transform_new:
libgimpcolor: some doc fixes and updates
2017-06-04 03:03:25 +08:00
* @src_profile: the source #GimpColorProfile
* @src_format: the source #Babl format
* @dest_profile: the destination #GimpColorProfile
* @dest_format: the destination #Babl format
* @rendering_intent: the rendering intent
* @flags: transform flags
libgimpcolor: add new object GimpColorTransform which encapsulates a cmsHTRANSFORM and does all the pixel format conversion magic. It has API to create transforms and proofing transforms, and to convert pixels arrays and GeglBuffers. Before, each place which has a transform had to keep around the transform and its input and output Babl formats, and had to implement lots of stuff itself. Now all that lives in GimpColorTransform, removing lots of logic from many places, and pretty much removing lcms from the public API entirely. This removes including <lcms2.h>, LCMS_LIBS and LCMS_CFLAGS from almost all directories and potentially allows to replace lcms by something else.
2016-05-26 03:35:54 +08:00
*
* This function creates an color transform.
*
Issue 1917 - GIMP-2.99 uses sRGB instead of the user-selected monitor profile Since the space invasion commit, colors entering and leaving a GimpColorTransform were often implicitly converted up to three times, the code was simply not properly ported to babl formats with spaces. Fix GimpColorTransform to only ever transform colors between the specified src and dest profiles, ignoring the space of any babl formats involved. Also, always return a non-NULL transform, even if the transform could be done by a simply gegl_buffer_copy(), this way we can make sure in one central place that transforms are done correctly, no matter if babl or lcms is used. Added quite some docs and comments to make clear what happens.
2018-08-01 02:00:35 +08:00
* The color transform is determined exclusively by @src_profile and
* @dest_profile. The color spaces of @src_format and @dest_format are
* ignored, the formats are only used to decide between what pixel
* encodings to transform.
*
* Note: this function used to return %NULL if
* gimp_color_transform_can_gegl_copy() returned %TRUE for
* @src_profile and @dest_profile. This is no longer the case because
* special care has to be taken not to perform multiple implicit color
* transforms caused by babl formats with color spaces. Now, it always
* returns a non-%NULL transform and the code takes care of doing only
* exactly the requested color transform.
*
* Return value: the #GimpColorTransform, or %NULL if there was an error.
libgimpcolor: add new object GimpColorTransform which encapsulates a cmsHTRANSFORM and does all the pixel format conversion magic. It has API to create transforms and proofing transforms, and to convert pixels arrays and GeglBuffers. Before, each place which has a transform had to keep around the transform and its input and output Babl formats, and had to implement lots of stuff itself. Now all that lives in GimpColorTransform, removing lots of logic from many places, and pretty much removing lcms from the public API entirely. This removes including <lcms2.h>, LCMS_LIBS and LCMS_CFLAGS from almost all directories and potentially allows to replace lcms by something else.
2016-05-26 03:35:54 +08:00
*
* Since: 2.10
**/
GimpColorTransform *
gimp_color_transform_new (GimpColorProfile *src_profile,
const Babl *src_format,
GimpColorProfile *dest_profile,
const Babl *dest_format,
GimpColorRenderingIntent rendering_intent,
Argh, this was meant to be squashed into the previous commit...
2016-05-27 04:18:50 +08:00
GimpColorTransformFlags flags)
libgimpcolor: add new object GimpColorTransform which encapsulates a cmsHTRANSFORM and does all the pixel format conversion magic. It has API to create transforms and proofing transforms, and to convert pixels arrays and GeglBuffers. Before, each place which has a transform had to keep around the transform and its input and output Babl formats, and had to implement lots of stuff itself. Now all that lives in GimpColorTransform, removing lots of logic from many places, and pretty much removing lcms from the public API entirely. This removes including <lcms2.h>, LCMS_LIBS and LCMS_CFLAGS from almost all directories and potentially allows to replace lcms by something else.
2016-05-26 03:35:54 +08:00
{
GimpColorTransform *transform;
GimpColorTransformPrivate *priv;
cmsHPROFILE src_lcms;
cmsHPROFILE dest_lcms;
cmsUInt32Number lcms_src_format;
cmsUInt32Number lcms_dest_format;
libgimpcolor: use gimp_color_profile_get_format() in GimpColorTransform and use babl to convert between profiles if possible. With the default BABL_TOLERANCE this is about 5 times faster than lcms on my test image. Using babl can be disabled by setting GIMP_COLOR_TRANSFORM_DISABLE_BABL.
2017-09-03 02:32:50 +08:00
GError *error = NULL;
libgimpcolor: add new object GimpColorTransform which encapsulates a cmsHTRANSFORM and does all the pixel format conversion magic. It has API to create transforms and proofing transforms, and to convert pixels arrays and GeglBuffers. Before, each place which has a transform had to keep around the transform and its input and output Babl formats, and had to implement lots of stuff itself. Now all that lives in GimpColorTransform, removing lots of logic from many places, and pretty much removing lcms from the public API entirely. This removes including <lcms2.h>, LCMS_LIBS and LCMS_CFLAGS from almost all directories and potentially allows to replace lcms by something else.
2016-05-26 03:35:54 +08:00
g_return_val_if_fail (GIMP_IS_COLOR_PROFILE (src_profile), NULL);
g_return_val_if_fail (src_format != NULL, NULL);
g_return_val_if_fail (GIMP_IS_COLOR_PROFILE (dest_profile), NULL);
g_return_val_if_fail (dest_format != NULL, NULL);
transform = g_object_new (GIMP_TYPE_COLOR_TRANSFORM, NULL);
priv = transform->priv;
Issue 1917 - GIMP-2.99 uses sRGB instead of the user-selected monitor profile Since the space invasion commit, colors entering and leaving a GimpColorTransform were often implicitly converted up to three times, the code was simply not properly ported to babl formats with spaces. Fix GimpColorTransform to only ever transform colors between the specified src and dest profiles, ignoring the space of any babl formats involved. Also, always return a non-NULL transform, even if the transform could be done by a simply gegl_buffer_copy(), this way we can make sure in one central place that transforms are done correctly, no matter if babl or lcms is used. Added quite some docs and comments to make clear what happens.
2018-08-01 02:00:35 +08:00
/* only src_profile and dest_profile must determine the transform's
* color spaces, create formats with src_format's and dest_format's
* encoding, and the profiles' color spaces; see process_pixels()
* and process_buffer().
*/
priv->src_format = gimp_color_profile_get_format (src_profile,
src_format,
BABL_ICC_INTENT_RELATIVE_COLORIMETRIC,
&error);
if (! priv->src_format)
libgimpcolor: use gimp_color_profile_get_format() in GimpColorTransform and use babl to convert between profiles if possible. With the default BABL_TOLERANCE this is about 5 times faster than lcms on my test image. Using babl can be disabled by setting GIMP_COLOR_TRANSFORM_DISABLE_BABL.
2017-09-03 02:32:50 +08:00
{
g_printerr ("%s: error making format: %s\n",
G_STRFUNC, error->message);
g_clear_error (&error);
}
Issue 1917 - GIMP-2.99 uses sRGB instead of the user-selected monitor profile Since the space invasion commit, colors entering and leaving a GimpColorTransform were often implicitly converted up to three times, the code was simply not properly ported to babl formats with spaces. Fix GimpColorTransform to only ever transform colors between the specified src and dest profiles, ignoring the space of any babl formats involved. Also, always return a non-NULL transform, even if the transform could be done by a simply gegl_buffer_copy(), this way we can make sure in one central place that transforms are done correctly, no matter if babl or lcms is used. Added quite some docs and comments to make clear what happens.
2018-08-01 02:00:35 +08:00
priv->dest_format = gimp_color_profile_get_format (dest_profile,
dest_format,
rendering_intent,
&error);
if (! priv->dest_format)
libgimpcolor: use gimp_color_profile_get_format() in GimpColorTransform and use babl to convert between profiles if possible. With the default BABL_TOLERANCE this is about 5 times faster than lcms on my test image. Using babl can be disabled by setting GIMP_COLOR_TRANSFORM_DISABLE_BABL.
2017-09-03 02:32:50 +08:00
{
g_printerr ("%s: error making format: %s\n",
G_STRFUNC, error->message);
g_clear_error (&error);
}
if (! g_getenv ("GIMP_COLOR_TRANSFORM_DISABLE_BABL") &&
Issue 1917 - GIMP-2.99 uses sRGB instead of the user-selected monitor profile Since the space invasion commit, colors entering and leaving a GimpColorTransform were often implicitly converted up to three times, the code was simply not properly ported to babl formats with spaces. Fix GimpColorTransform to only ever transform colors between the specified src and dest profiles, ignoring the space of any babl formats involved. Also, always return a non-NULL transform, even if the transform could be done by a simply gegl_buffer_copy(), this way we can make sure in one central place that transforms are done correctly, no matter if babl or lcms is used. Added quite some docs and comments to make clear what happens.
2018-08-01 02:00:35 +08:00
priv->src_format && priv->dest_format)
libgimpcolor: use gimp_color_profile_get_format() in GimpColorTransform and use babl to convert between profiles if possible. With the default BABL_TOLERANCE this is about 5 times faster than lcms on my test image. Using babl can be disabled by setting GIMP_COLOR_TRANSFORM_DISABLE_BABL.
2017-09-03 02:32:50 +08:00
{
Issue 1917 - GIMP-2.99 uses sRGB instead of the user-selected monitor profile Since the space invasion commit, colors entering and leaving a GimpColorTransform were often implicitly converted up to three times, the code was simply not properly ported to babl formats with spaces. Fix GimpColorTransform to only ever transform colors between the specified src and dest profiles, ignoring the space of any babl formats involved. Also, always return a non-NULL transform, even if the transform could be done by a simply gegl_buffer_copy(), this way we can make sure in one central place that transforms are done correctly, no matter if babl or lcms is used. Added quite some docs and comments to make clear what happens.
2018-08-01 02:00:35 +08:00
priv->fish = babl_fish (priv->src_format,
priv->dest_format);
libgimpcolor: use gimp_color_profile_get_format() in GimpColorTransform and use babl to convert between profiles if possible. With the default BABL_TOLERANCE this is about 5 times faster than lcms on my test image. Using babl can be disabled by setting GIMP_COLOR_TRANSFORM_DISABLE_BABL.
2017-09-03 02:32:50 +08:00
g_printerr ("%s: using babl for '%s' -> '%s'\n",
G_STRFUNC,
gimp_color_profile_get_label (src_profile),
gimp_color_profile_get_label (dest_profile));
Bug 795300 - ICC profile conversion artifacts and gamma problems Cache the fish in GimpColorTransform as suggested by Massimo.
2018-04-19 00:49:47 +08:00
libgimpcolor: use gimp_color_profile_get_format() in GimpColorTransform and use babl to convert between profiles if possible. With the default BABL_TOLERANCE this is about 5 times faster than lcms on my test image. Using babl can be disabled by setting GIMP_COLOR_TRANSFORM_DISABLE_BABL.
2017-09-03 02:32:50 +08:00
return transform;
}
Issue 1917 - GIMP-2.99 uses sRGB instead of the user-selected monitor profile Since the space invasion commit, colors entering and leaving a GimpColorTransform were often implicitly converted up to three times, the code was simply not properly ported to babl formats with spaces. Fix GimpColorTransform to only ever transform colors between the specified src and dest profiles, ignoring the space of any babl formats involved. Also, always return a non-NULL transform, even if the transform could be done by a simply gegl_buffer_copy(), this way we can make sure in one central place that transforms are done correctly, no matter if babl or lcms is used. Added quite some docs and comments to make clear what happens.
2018-08-01 02:00:35 +08:00
/* see above: when using lcms, don't mess with formats with color
* spaces, gimp_color_profile_get_lcms_format() might return the
* same format and it must be without space
*/
src_format = babl_format_with_space (babl_format_get_encoding (src_format),
NULL);
dest_format = babl_format_with_space (babl_format_get_encoding (dest_format),
NULL);
libgimpcolor: add new object GimpColorTransform which encapsulates a cmsHTRANSFORM and does all the pixel format conversion magic. It has API to create transforms and proofing transforms, and to convert pixels arrays and GeglBuffers. Before, each place which has a transform had to keep around the transform and its input and output Babl formats, and had to implement lots of stuff itself. Now all that lives in GimpColorTransform, removing lots of logic from many places, and pretty much removing lcms from the public API entirely. This removes including <lcms2.h>, LCMS_LIBS and LCMS_CFLAGS from almost all directories and potentially allows to replace lcms by something else.
2016-05-26 03:35:54 +08:00
libgimpcolor: rename gimp_color_profile_get_format() to _get_lcms_format()
2017-09-03 00:47:04 +08:00
priv->src_format = gimp_color_profile_get_lcms_format (src_format,
&lcms_src_format);
priv->dest_format = gimp_color_profile_get_lcms_format (dest_format,
&lcms_dest_format);
libgimpcolor: add new object GimpColorTransform which encapsulates a cmsHTRANSFORM and does all the pixel format conversion magic. It has API to create transforms and proofing transforms, and to convert pixels arrays and GeglBuffers. Before, each place which has a transform had to keep around the transform and its input and output Babl formats, and had to implement lots of stuff itself. Now all that lives in GimpColorTransform, removing lots of logic from many places, and pretty much removing lcms from the public API entirely. This removes including <lcms2.h>, LCMS_LIBS and LCMS_CFLAGS from almost all directories and potentially allows to replace lcms by something else.
2016-05-26 03:35:54 +08:00
libgimpcolor: use gimp_color_profile_get_format() in GimpColorTransform and use babl to convert between profiles if possible. With the default BABL_TOLERANCE this is about 5 times faster than lcms on my test image. Using babl can be disabled by setting GIMP_COLOR_TRANSFORM_DISABLE_BABL.
2017-09-03 02:32:50 +08:00
src_lcms = gimp_color_profile_get_lcms_profile (src_profile);
dest_lcms = gimp_color_profile_get_lcms_profile (dest_profile);
libgimpcolor: return a NULL transform if cmsCreate[Proofing]Transform() fails It doesn't really "fail", we check if an error was logged during transform creation and get rid of the transform even if lcms returned a pointer. This totally sucks but is better than crashing on a corrupt transform.
2016-05-30 18:48:54 +08:00
lcms_error_clear ();
libgimpcolor: add new object GimpColorTransform which encapsulates a cmsHTRANSFORM and does all the pixel format conversion magic. It has API to create transforms and proofing transforms, and to convert pixels arrays and GeglBuffers. Before, each place which has a transform had to keep around the transform and its input and output Babl formats, and had to implement lots of stuff itself. Now all that lives in GimpColorTransform, removing lots of logic from many places, and pretty much removing lcms from the public API entirely. This removes including <lcms2.h>, LCMS_LIBS and LCMS_CFLAGS from almost all directories and potentially allows to replace lcms by something else.
2016-05-26 03:35:54 +08:00
priv->transform = cmsCreateTransform (src_lcms, lcms_src_format,
dest_lcms, lcms_dest_format,
rendering_intent,
libgimpcolor: use the new cmsFLAGS_COPY_ALPHA in GimpColorTransform and remove all manual alpha copying.
2017-09-02 23:11:45 +08:00
flags |
cmsFLAGS_COPY_ALPHA);
libgimpcolor: add new object GimpColorTransform which encapsulates a cmsHTRANSFORM and does all the pixel format conversion magic. It has API to create transforms and proofing transforms, and to convert pixels arrays and GeglBuffers. Before, each place which has a transform had to keep around the transform and its input and output Babl formats, and had to implement lots of stuff itself. Now all that lives in GimpColorTransform, removing lots of logic from many places, and pretty much removing lcms from the public API entirely. This removes including <lcms2.h>, LCMS_LIBS and LCMS_CFLAGS from almost all directories and potentially allows to replace lcms by something else.
2016-05-26 03:35:54 +08:00
libgimpcolor: return a NULL transform if cmsCreate[Proofing]Transform() fails It doesn't really "fail", we check if an error was logged during transform creation and get rid of the transform even if lcms returned a pointer. This totally sucks but is better than crashing on a corrupt transform.
2016-05-30 18:48:54 +08:00
if (lcms_last_error)
{
if (priv->transform)
{
cmsDeleteTransform (priv->transform);
priv->transform = NULL;
}
libgimpcolor: make gimp_color_transform_process_buffer() thread-safer Avoid gegl_buffer_set_format() and simply construct the formats differently to make sure reading/writing doesn't do any color space transform. This makes it possible again to run it on different areas of the same buffer from different threads.
2018-08-03 01:25:52 +08:00
g_printerr ("%s: %s\n", G_STRFUNC, lcms_last_error);
libgimpcolor: return a NULL transform if cmsCreate[Proofing]Transform() fails It doesn't really "fail", we check if an error was logged during transform creation and get rid of the transform even if lcms returned a pointer. This totally sucks but is better than crashing on a corrupt transform.
2016-05-30 18:48:54 +08:00
}
libgimpcolor: return NULL GimpColorTransform if creating cmsHTRANSFORM fails
2016-05-30 16:58:48 +08:00
if (! priv->transform)
{
g_object_unref (transform);
transform = NULL;
}
libgimpcolor: add new object GimpColorTransform which encapsulates a cmsHTRANSFORM and does all the pixel format conversion magic. It has API to create transforms and proofing transforms, and to convert pixels arrays and GeglBuffers. Before, each place which has a transform had to keep around the transform and its input and output Babl formats, and had to implement lots of stuff itself. Now all that lives in GimpColorTransform, removing lots of logic from many places, and pretty much removing lcms from the public API entirely. This removes including <lcms2.h>, LCMS_LIBS and LCMS_CFLAGS from almost all directories and potentially allows to replace lcms by something else.
2016-05-26 03:35:54 +08:00
return transform;
}
/**
* gimp_color_transform_new_proofing:
libgimpcolor: some doc fixes and updates
2017-06-04 03:03:25 +08:00
* @src_profile: the source #GimpColorProfile
* @src_format: the source #Babl format
* @dest_profile: the destination #GimpColorProfile
* @dest_format: the destination #Babl format
* @proof_profile: the proof #GimpColorProfile
* @proof_intent: the proof intent
* @display_intent: the display intent
* @flags: transform flags
libgimpcolor: add new object GimpColorTransform which encapsulates a cmsHTRANSFORM and does all the pixel format conversion magic. It has API to create transforms and proofing transforms, and to convert pixels arrays and GeglBuffers. Before, each place which has a transform had to keep around the transform and its input and output Babl formats, and had to implement lots of stuff itself. Now all that lives in GimpColorTransform, removing lots of logic from many places, and pretty much removing lcms from the public API entirely. This removes including <lcms2.h>, LCMS_LIBS and LCMS_CFLAGS from almost all directories and potentially allows to replace lcms by something else.
2016-05-26 03:35:54 +08:00
*
* This function creates a simulation / proofing color transform.
*
Issue 1917 - GIMP-2.99 uses sRGB instead of the user-selected monitor profile Since the space invasion commit, colors entering and leaving a GimpColorTransform were often implicitly converted up to three times, the code was simply not properly ported to babl formats with spaces. Fix GimpColorTransform to only ever transform colors between the specified src and dest profiles, ignoring the space of any babl formats involved. Also, always return a non-NULL transform, even if the transform could be done by a simply gegl_buffer_copy(), this way we can make sure in one central place that transforms are done correctly, no matter if babl or lcms is used. Added quite some docs and comments to make clear what happens.
2018-08-01 02:00:35 +08:00
* See gimp_color_transform_new() about the color spaces to transform
* between.
*
* Return value: the #GimpColorTransform, or %NULL if there was an error.
libgimpcolor: add new object GimpColorTransform which encapsulates a cmsHTRANSFORM and does all the pixel format conversion magic. It has API to create transforms and proofing transforms, and to convert pixels arrays and GeglBuffers. Before, each place which has a transform had to keep around the transform and its input and output Babl formats, and had to implement lots of stuff itself. Now all that lives in GimpColorTransform, removing lots of logic from many places, and pretty much removing lcms from the public API entirely. This removes including <lcms2.h>, LCMS_LIBS and LCMS_CFLAGS from almost all directories and potentially allows to replace lcms by something else.
2016-05-26 03:35:54 +08:00
*
* Since: 2.10
**/
GimpColorTransform *
gimp_color_transform_new_proofing (GimpColorProfile *src_profile,
const Babl *src_format,
GimpColorProfile *dest_profile,
const Babl *dest_format,
GimpColorProfile *proof_profile,
GimpColorRenderingIntent proof_intent,
GimpColorRenderingIntent display_intent,
Argh, this was meant to be squashed into the previous commit...
2016-05-27 04:18:50 +08:00
GimpColorTransformFlags flags)
libgimpcolor: add new object GimpColorTransform which encapsulates a cmsHTRANSFORM and does all the pixel format conversion magic. It has API to create transforms and proofing transforms, and to convert pixels arrays and GeglBuffers. Before, each place which has a transform had to keep around the transform and its input and output Babl formats, and had to implement lots of stuff itself. Now all that lives in GimpColorTransform, removing lots of logic from many places, and pretty much removing lcms from the public API entirely. This removes including <lcms2.h>, LCMS_LIBS and LCMS_CFLAGS from almost all directories and potentially allows to replace lcms by something else.
2016-05-26 03:35:54 +08:00
{
GimpColorTransform *transform;
GimpColorTransformPrivate *priv;
cmsHPROFILE src_lcms;
cmsHPROFILE dest_lcms;
cmsHPROFILE proof_lcms;
cmsUInt32Number lcms_src_format;
cmsUInt32Number lcms_dest_format;
g_return_val_if_fail (GIMP_IS_COLOR_PROFILE (src_profile), NULL);
g_return_val_if_fail (src_format != NULL, NULL);
g_return_val_if_fail (GIMP_IS_COLOR_PROFILE (dest_profile), NULL);
g_return_val_if_fail (dest_format != NULL, NULL);
g_return_val_if_fail (GIMP_IS_COLOR_PROFILE (proof_profile), NULL);
transform = g_object_new (GIMP_TYPE_COLOR_TRANSFORM, NULL);
priv = transform->priv;
src_lcms = gimp_color_profile_get_lcms_profile (src_profile);
dest_lcms = gimp_color_profile_get_lcms_profile (dest_profile);
proof_lcms = gimp_color_profile_get_lcms_profile (proof_profile);
Issue 1917 - GIMP-2.99 uses sRGB instead of the user-selected monitor profile Since the space invasion commit, colors entering and leaving a GimpColorTransform were often implicitly converted up to three times, the code was simply not properly ported to babl formats with spaces. Fix GimpColorTransform to only ever transform colors between the specified src and dest profiles, ignoring the space of any babl formats involved. Also, always return a non-NULL transform, even if the transform could be done by a simply gegl_buffer_copy(), this way we can make sure in one central place that transforms are done correctly, no matter if babl or lcms is used. Added quite some docs and comments to make clear what happens.
2018-08-01 02:00:35 +08:00
/* see gimp_color_transform_new(), we can't have color spaces
* on the formats
*/
src_format = babl_format_with_space (babl_format_get_encoding (src_format),
NULL);
dest_format = babl_format_with_space (babl_format_get_encoding (dest_format),
NULL);
libgimpcolor: rename gimp_color_profile_get_format() to _get_lcms_format()
2017-09-03 00:47:04 +08:00
priv->src_format = gimp_color_profile_get_lcms_format (src_format,
&lcms_src_format);
priv->dest_format = gimp_color_profile_get_lcms_format (dest_format,
&lcms_dest_format);
libgimpcolor: add new object GimpColorTransform which encapsulates a cmsHTRANSFORM and does all the pixel format conversion magic. It has API to create transforms and proofing transforms, and to convert pixels arrays and GeglBuffers. Before, each place which has a transform had to keep around the transform and its input and output Babl formats, and had to implement lots of stuff itself. Now all that lives in GimpColorTransform, removing lots of logic from many places, and pretty much removing lcms from the public API entirely. This removes including <lcms2.h>, LCMS_LIBS and LCMS_CFLAGS from almost all directories and potentially allows to replace lcms by something else.
2016-05-26 03:35:54 +08:00
libgimpcolor: return a NULL transform if cmsCreate[Proofing]Transform() fails It doesn't really "fail", we check if an error was logged during transform creation and get rid of the transform even if lcms returned a pointer. This totally sucks but is better than crashing on a corrupt transform.
2016-05-30 18:48:54 +08:00
lcms_error_clear ();
libgimpcolor: add new object GimpColorTransform which encapsulates a cmsHTRANSFORM and does all the pixel format conversion magic. It has API to create transforms and proofing transforms, and to convert pixels arrays and GeglBuffers. Before, each place which has a transform had to keep around the transform and its input and output Babl formats, and had to implement lots of stuff itself. Now all that lives in GimpColorTransform, removing lots of logic from many places, and pretty much removing lcms from the public API entirely. This removes including <lcms2.h>, LCMS_LIBS and LCMS_CFLAGS from almost all directories and potentially allows to replace lcms by something else.
2016-05-26 03:35:54 +08:00
priv->transform = cmsCreateProofingTransform (src_lcms, lcms_src_format,
dest_lcms, lcms_dest_format,
proof_lcms,
proof_intent,
display_intent,
libgimpcolor: use the new cmsFLAGS_COPY_ALPHA in GimpColorTransform and remove all manual alpha copying.
2017-09-02 23:11:45 +08:00
flags |
cmsFLAGS_SOFTPROOFING |
cmsFLAGS_COPY_ALPHA);
libgimpcolor: add new object GimpColorTransform which encapsulates a cmsHTRANSFORM and does all the pixel format conversion magic. It has API to create transforms and proofing transforms, and to convert pixels arrays and GeglBuffers. Before, each place which has a transform had to keep around the transform and its input and output Babl formats, and had to implement lots of stuff itself. Now all that lives in GimpColorTransform, removing lots of logic from many places, and pretty much removing lcms from the public API entirely. This removes including <lcms2.h>, LCMS_LIBS and LCMS_CFLAGS from almost all directories and potentially allows to replace lcms by something else.
2016-05-26 03:35:54 +08:00
libgimpcolor: return a NULL transform if cmsCreate[Proofing]Transform() fails It doesn't really "fail", we check if an error was logged during transform creation and get rid of the transform even if lcms returned a pointer. This totally sucks but is better than crashing on a corrupt transform.
2016-05-30 18:48:54 +08:00
if (lcms_last_error)
{
if (priv->transform)
{
cmsDeleteTransform (priv->transform);
priv->transform = NULL;
}
libgimpcolor: make gimp_color_transform_process_buffer() thread-safer Avoid gegl_buffer_set_format() and simply construct the formats differently to make sure reading/writing doesn't do any color space transform. This makes it possible again to run it on different areas of the same buffer from different threads.
2018-08-03 01:25:52 +08:00
g_printerr ("%s: %s\n", G_STRFUNC, lcms_last_error);
libgimpcolor: return a NULL transform if cmsCreate[Proofing]Transform() fails It doesn't really "fail", we check if an error was logged during transform creation and get rid of the transform even if lcms returned a pointer. This totally sucks but is better than crashing on a corrupt transform.
2016-05-30 18:48:54 +08:00
}
libgimpcolor: return NULL GimpColorTransform if creating cmsHTRANSFORM fails
2016-05-30 16:58:48 +08:00
if (! priv->transform)
{
g_object_unref (transform);
transform = NULL;
}
libgimpcolor: add new object GimpColorTransform which encapsulates a cmsHTRANSFORM and does all the pixel format conversion magic. It has API to create transforms and proofing transforms, and to convert pixels arrays and GeglBuffers. Before, each place which has a transform had to keep around the transform and its input and output Babl formats, and had to implement lots of stuff itself. Now all that lives in GimpColorTransform, removing lots of logic from many places, and pretty much removing lcms from the public API entirely. This removes including <lcms2.h>, LCMS_LIBS and LCMS_CFLAGS from almost all directories and potentially allows to replace lcms by something else.
2016-05-26 03:35:54 +08:00
return transform;
}
/**
* gimp_color_transform_process_pixels:
libgimpcolor: some doc fixes and updates
2017-06-04 03:03:25 +08:00
* @transform: a #GimpColorTransform
* @src_format: #Babl format of @src_pixels
* @src_pixels: pointer to the source pixels
* @dest_format: #Babl format of @dest_pixels
* @dest_pixels: pointer to the destination pixels
* @length: number of pixels to process
libgimpcolor: add new object GimpColorTransform which encapsulates a cmsHTRANSFORM and does all the pixel format conversion magic. It has API to create transforms and proofing transforms, and to convert pixels arrays and GeglBuffers. Before, each place which has a transform had to keep around the transform and its input and output Babl formats, and had to implement lots of stuff itself. Now all that lives in GimpColorTransform, removing lots of logic from many places, and pretty much removing lcms from the public API entirely. This removes including <lcms2.h>, LCMS_LIBS and LCMS_CFLAGS from almost all directories and potentially allows to replace lcms by something else.
2016-05-26 03:35:54 +08:00
*
* This function transforms a contiguous line of pixels.
*
Issue 1917 - GIMP-2.99 uses sRGB instead of the user-selected monitor profile Since the space invasion commit, colors entering and leaving a GimpColorTransform were often implicitly converted up to three times, the code was simply not properly ported to babl formats with spaces. Fix GimpColorTransform to only ever transform colors between the specified src and dest profiles, ignoring the space of any babl formats involved. Also, always return a non-NULL transform, even if the transform could be done by a simply gegl_buffer_copy(), this way we can make sure in one central place that transforms are done correctly, no matter if babl or lcms is used. Added quite some docs and comments to make clear what happens.
2018-08-01 02:00:35 +08:00
* See gimp_color_transform_new(): only the pixel encoding of
* @src_format and @dest_format is honored, their color spaces are
* ignored. The transform always takes place between the color spaces
* determined by @transform's color profiles.
*
libgimpcolor: add new object GimpColorTransform which encapsulates a cmsHTRANSFORM and does all the pixel format conversion magic. It has API to create transforms and proofing transforms, and to convert pixels arrays and GeglBuffers. Before, each place which has a transform had to keep around the transform and its input and output Babl formats, and had to implement lots of stuff itself. Now all that lives in GimpColorTransform, removing lots of logic from many places, and pretty much removing lcms from the public API entirely. This removes including <lcms2.h>, LCMS_LIBS and LCMS_CFLAGS from almost all directories and potentially allows to replace lcms by something else.
2016-05-26 03:35:54 +08:00
* Since: 2.10
**/
void
gimp_color_transform_process_pixels (GimpColorTransform *transform,
const Babl *src_format,
gconstpointer src_pixels,
const Babl *dest_format,
gpointer dest_pixels,
gsize length)
{
GimpColorTransformPrivate *priv;
gpointer *src;
gpointer *dest;
g_return_if_fail (GIMP_IS_COLOR_TRANSFORM (transform));
g_return_if_fail (src_format != NULL);
g_return_if_fail (src_pixels != NULL);
g_return_if_fail (dest_format != NULL);
g_return_if_fail (dest_pixels != NULL);
priv = transform->priv;
Issue 1917 - GIMP-2.99 uses sRGB instead of the user-selected monitor profile Since the space invasion commit, colors entering and leaving a GimpColorTransform were often implicitly converted up to three times, the code was simply not properly ported to babl formats with spaces. Fix GimpColorTransform to only ever transform colors between the specified src and dest profiles, ignoring the space of any babl formats involved. Also, always return a non-NULL transform, even if the transform could be done by a simply gegl_buffer_copy(), this way we can make sure in one central place that transforms are done correctly, no matter if babl or lcms is used. Added quite some docs and comments to make clear what happens.
2018-08-01 02:00:35 +08:00
/* we must not do any babl color transforms when reading from
* src_pixels or writing to dest_pixels, so construct formats with
* src_format's and dest_format's encoding, and the transform's
* input and output color spaces.
*/
src_format =
babl_format_with_space (babl_format_get_encoding (src_format),
babl_format_get_space (priv->src_format));
dest_format =
babl_format_with_space (babl_format_get_encoding (dest_format),
babl_format_get_space (priv->dest_format));
libgimpcolor: add new object GimpColorTransform which encapsulates a cmsHTRANSFORM and does all the pixel format conversion magic. It has API to create transforms and proofing transforms, and to convert pixels arrays and GeglBuffers. Before, each place which has a transform had to keep around the transform and its input and output Babl formats, and had to implement lots of stuff itself. Now all that lives in GimpColorTransform, removing lots of logic from many places, and pretty much removing lcms from the public API entirely. This removes including <lcms2.h>, LCMS_LIBS and LCMS_CFLAGS from almost all directories and potentially allows to replace lcms by something else.
2016-05-26 03:35:54 +08:00
if (src_format != priv->src_format)
{
src = g_malloc (length * babl_format_get_bytes_per_pixel (priv->src_format));
babl_process (babl_fish (src_format,
priv->src_format),
src_pixels, src, length);
}
else
{
src = (gpointer) src_pixels;
}
if (dest_format != priv->dest_format)
{
dest = g_malloc (length * babl_format_get_bytes_per_pixel (priv->dest_format));
}
else
{
dest = dest_pixels;
}
libgimpcolor: use gimp_color_profile_get_format() in GimpColorTransform and use babl to convert between profiles if possible. With the default BABL_TOLERANCE this is about 5 times faster than lcms on my test image. Using babl can be disabled by setting GIMP_COLOR_TRANSFORM_DISABLE_BABL.
2017-09-03 02:32:50 +08:00
if (priv->transform)
{
cmsDoTransform (priv->transform, src, dest, length);
}
else
{
Bug 795300 - ICC profile conversion artifacts and gamma problems Cache the fish in GimpColorTransform as suggested by Massimo.
2018-04-19 00:49:47 +08:00
babl_process (priv->fish, src, dest, length);
libgimpcolor: use gimp_color_profile_get_format() in GimpColorTransform and use babl to convert between profiles if possible. With the default BABL_TOLERANCE this is about 5 times faster than lcms on my test image. Using babl can be disabled by setting GIMP_COLOR_TRANSFORM_DISABLE_BABL.
2017-09-03 02:32:50 +08:00
}
libgimpcolor: add new object GimpColorTransform which encapsulates a cmsHTRANSFORM and does all the pixel format conversion magic. It has API to create transforms and proofing transforms, and to convert pixels arrays and GeglBuffers. Before, each place which has a transform had to keep around the transform and its input and output Babl formats, and had to implement lots of stuff itself. Now all that lives in GimpColorTransform, removing lots of logic from many places, and pretty much removing lcms from the public API entirely. This removes including <lcms2.h>, LCMS_LIBS and LCMS_CFLAGS from almost all directories and potentially allows to replace lcms by something else.
2016-05-26 03:35:54 +08:00
if (src_format != priv->src_format)
{
g_free (src);
}
if (dest_format != priv->dest_format)
{
babl_process (babl_fish (priv->dest_format,
dest_format),
dest, dest_pixels, length);
g_free (dest);
}
}
/**
* gimp_color_transform_process_buffer:
libgimpcolor: some doc fixes and updates
2017-06-04 03:03:25 +08:00
* @transform: a #GimpColorTransform
* @src_buffer: source #GeglBuffer
* @src_rect: rectangle in @src_buffer
* @dest_buffer: destination #GeglBuffer
* @dest_rect: rectangle in @dest_buffer
libgimpcolor: add new object GimpColorTransform which encapsulates a cmsHTRANSFORM and does all the pixel format conversion magic. It has API to create transforms and proofing transforms, and to convert pixels arrays and GeglBuffers. Before, each place which has a transform had to keep around the transform and its input and output Babl formats, and had to implement lots of stuff itself. Now all that lives in GimpColorTransform, removing lots of logic from many places, and pretty much removing lcms from the public API entirely. This removes including <lcms2.h>, LCMS_LIBS and LCMS_CFLAGS from almost all directories and potentially allows to replace lcms by something else.
2016-05-26 03:35:54 +08:00
*
libgimpcolor: optimize gimp_color_transform_process_buffer() for the case we're transforming the same buffer in place.
2016-05-30 02:24:37 +08:00
* This function transforms buffer into another buffer.
libgimpcolor: add new object GimpColorTransform which encapsulates a cmsHTRANSFORM and does all the pixel format conversion magic. It has API to create transforms and proofing transforms, and to convert pixels arrays and GeglBuffers. Before, each place which has a transform had to keep around the transform and its input and output Babl formats, and had to implement lots of stuff itself. Now all that lives in GimpColorTransform, removing lots of logic from many places, and pretty much removing lcms from the public API entirely. This removes including <lcms2.h>, LCMS_LIBS and LCMS_CFLAGS from almost all directories and potentially allows to replace lcms by something else.
2016-05-26 03:35:54 +08:00
*
Issue 1917 - GIMP-2.99 uses sRGB instead of the user-selected monitor profile Since the space invasion commit, colors entering and leaving a GimpColorTransform were often implicitly converted up to three times, the code was simply not properly ported to babl formats with spaces. Fix GimpColorTransform to only ever transform colors between the specified src and dest profiles, ignoring the space of any babl formats involved. Also, always return a non-NULL transform, even if the transform could be done by a simply gegl_buffer_copy(), this way we can make sure in one central place that transforms are done correctly, no matter if babl or lcms is used. Added quite some docs and comments to make clear what happens.
2018-08-01 02:00:35 +08:00
* See gimp_color_transform_new(): only the pixel encoding of
* @src_buffer's and @dest_buffer's formats honored, their color
* spaces are ignored. The transform always takes place between the
* color spaces determined by @transform's color profiles.
*
libgimpcolor: add new object GimpColorTransform which encapsulates a cmsHTRANSFORM and does all the pixel format conversion magic. It has API to create transforms and proofing transforms, and to convert pixels arrays and GeglBuffers. Before, each place which has a transform had to keep around the transform and its input and output Babl formats, and had to implement lots of stuff itself. Now all that lives in GimpColorTransform, removing lots of logic from many places, and pretty much removing lcms from the public API entirely. This removes including <lcms2.h>, LCMS_LIBS and LCMS_CFLAGS from almost all directories and potentially allows to replace lcms by something else.
2016-05-26 03:35:54 +08:00
* Since: 2.10
**/
void
gimp_color_transform_process_buffer (GimpColorTransform *transform,
GeglBuffer *src_buffer,
const GeglRectangle *src_rect,
GeglBuffer *dest_buffer,
const GeglRectangle *dest_rect)
{
GimpColorTransformPrivate *priv;
Issue 1917 - GIMP-2.99 uses sRGB instead of the user-selected monitor profile Since the space invasion commit, colors entering and leaving a GimpColorTransform were often implicitly converted up to three times, the code was simply not properly ported to babl formats with spaces. Fix GimpColorTransform to only ever transform colors between the specified src and dest profiles, ignoring the space of any babl formats involved. Also, always return a non-NULL transform, even if the transform could be done by a simply gegl_buffer_copy(), this way we can make sure in one central place that transforms are done correctly, no matter if babl or lcms is used. Added quite some docs and comments to make clear what happens.
2018-08-01 02:00:35 +08:00
const Babl *src_format;
const Babl *dest_format;
libgimpcolor: add new object GimpColorTransform which encapsulates a cmsHTRANSFORM and does all the pixel format conversion magic. It has API to create transforms and proofing transforms, and to convert pixels arrays and GeglBuffers. Before, each place which has a transform had to keep around the transform and its input and output Babl formats, and had to implement lots of stuff itself. Now all that lives in GimpColorTransform, removing lots of logic from many places, and pretty much removing lcms from the public API entirely. This removes including <lcms2.h>, LCMS_LIBS and LCMS_CFLAGS from almost all directories and potentially allows to replace lcms by something else.
2016-05-26 03:35:54 +08:00
GeglBufferIterator *iter;
gint total_pixels;
gint done_pixels = 0;
g_return_if_fail (GIMP_IS_COLOR_TRANSFORM (transform));
g_return_if_fail (GEGL_IS_BUFFER (src_buffer));
g_return_if_fail (GEGL_IS_BUFFER (dest_buffer));
priv = transform->priv;
if (src_rect)
{
total_pixels = src_rect->width * src_rect->height;
}
else
{
total_pixels = (gegl_buffer_get_width (src_buffer) *
gegl_buffer_get_height (src_buffer));
}
Issue 1917 - GIMP-2.99 uses sRGB instead of the user-selected monitor profile Since the space invasion commit, colors entering and leaving a GimpColorTransform were often implicitly converted up to three times, the code was simply not properly ported to babl formats with spaces. Fix GimpColorTransform to only ever transform colors between the specified src and dest profiles, ignoring the space of any babl formats involved. Also, always return a non-NULL transform, even if the transform could be done by a simply gegl_buffer_copy(), this way we can make sure in one central place that transforms are done correctly, no matter if babl or lcms is used. Added quite some docs and comments to make clear what happens.
2018-08-01 02:00:35 +08:00
/* we must not do any babl color transforms when reading from
* src_buffer or writing to dest_buffer, so construct formats with
libgimpcolor: make gimp_color_transform_process_buffer() thread-safer Avoid gegl_buffer_set_format() and simply construct the formats differently to make sure reading/writing doesn't do any color space transform. This makes it possible again to run it on different areas of the same buffer from different threads.
2018-08-03 01:25:52 +08:00
* the transform's expected input and output encoding and
* src_buffer's and dest_buffers's color spaces.
Issue 1917 - GIMP-2.99 uses sRGB instead of the user-selected monitor profile Since the space invasion commit, colors entering and leaving a GimpColorTransform were often implicitly converted up to three times, the code was simply not properly ported to babl formats with spaces. Fix GimpColorTransform to only ever transform colors between the specified src and dest profiles, ignoring the space of any babl formats involved. Also, always return a non-NULL transform, even if the transform could be done by a simply gegl_buffer_copy(), this way we can make sure in one central place that transforms are done correctly, no matter if babl or lcms is used. Added quite some docs and comments to make clear what happens.
2018-08-01 02:00:35 +08:00
*/
src_format = gegl_buffer_get_format (src_buffer);
dest_format = gegl_buffer_get_format (dest_buffer);
src_format =
libgimpcolor: make gimp_color_transform_process_buffer() thread-safer Avoid gegl_buffer_set_format() and simply construct the formats differently to make sure reading/writing doesn't do any color space transform. This makes it possible again to run it on different areas of the same buffer from different threads.
2018-08-03 01:25:52 +08:00
babl_format_with_space (babl_format_get_encoding (priv->src_format),
babl_format_get_space (src_format));
Issue 1917 - GIMP-2.99 uses sRGB instead of the user-selected monitor profile Since the space invasion commit, colors entering and leaving a GimpColorTransform were often implicitly converted up to three times, the code was simply not properly ported to babl formats with spaces. Fix GimpColorTransform to only ever transform colors between the specified src and dest profiles, ignoring the space of any babl formats involved. Also, always return a non-NULL transform, even if the transform could be done by a simply gegl_buffer_copy(), this way we can make sure in one central place that transforms are done correctly, no matter if babl or lcms is used. Added quite some docs and comments to make clear what happens.
2018-08-01 02:00:35 +08:00
dest_format =
libgimpcolor: make gimp_color_transform_process_buffer() thread-safer Avoid gegl_buffer_set_format() and simply construct the formats differently to make sure reading/writing doesn't do any color space transform. This makes it possible again to run it on different areas of the same buffer from different threads.
2018-08-03 01:25:52 +08:00
babl_format_with_space (babl_format_get_encoding (priv->dest_format),
babl_format_get_space (dest_format));
Issue 1917 - GIMP-2.99 uses sRGB instead of the user-selected monitor profile Since the space invasion commit, colors entering and leaving a GimpColorTransform were often implicitly converted up to three times, the code was simply not properly ported to babl formats with spaces. Fix GimpColorTransform to only ever transform colors between the specified src and dest profiles, ignoring the space of any babl formats involved. Also, always return a non-NULL transform, even if the transform could be done by a simply gegl_buffer_copy(), this way we can make sure in one central place that transforms are done correctly, no matter if babl or lcms is used. Added quite some docs and comments to make clear what happens.
2018-08-01 02:00:35 +08:00
libgimpcolor: optimize gimp_color_transform_process_buffer() for the case we're transforming the same buffer in place.
2016-05-30 02:24:37 +08:00
if (src_buffer != dest_buffer)
{
iter = gegl_buffer_iterator_new (src_buffer, src_rect, 0,
libgimpcolor: make gimp_color_transform_process_buffer() thread-safer Avoid gegl_buffer_set_format() and simply construct the formats differently to make sure reading/writing doesn't do any color space transform. This makes it possible again to run it on different areas of the same buffer from different threads.
2018-08-03 01:25:52 +08:00
src_format,
libgimpcolor: optimize gimp_color_transform_process_buffer() for the case we're transforming the same buffer in place.
2016-05-30 02:24:37 +08:00
GEGL_ACCESS_READ,
GEGL_ABYSS_NONE);
libgimpcolor: add new object GimpColorTransform which encapsulates a cmsHTRANSFORM and does all the pixel format conversion magic. It has API to create transforms and proofing transforms, and to convert pixels arrays and GeglBuffers. Before, each place which has a transform had to keep around the transform and its input and output Babl formats, and had to implement lots of stuff itself. Now all that lives in GimpColorTransform, removing lots of logic from many places, and pretty much removing lcms from the public API entirely. This removes including <lcms2.h>, LCMS_LIBS and LCMS_CFLAGS from almost all directories and potentially allows to replace lcms by something else.
2016-05-26 03:35:54 +08:00
libgimpcolor: optimize gimp_color_transform_process_buffer() for the case we're transforming the same buffer in place.
2016-05-30 02:24:37 +08:00
gegl_buffer_iterator_add (iter, dest_buffer, dest_rect, 0,
libgimpcolor: make gimp_color_transform_process_buffer() thread-safer Avoid gegl_buffer_set_format() and simply construct the formats differently to make sure reading/writing doesn't do any color space transform. This makes it possible again to run it on different areas of the same buffer from different threads.
2018-08-03 01:25:52 +08:00
dest_format,
libgimpcolor: optimize gimp_color_transform_process_buffer() for the case we're transforming the same buffer in place.
2016-05-30 02:24:37 +08:00
GEGL_ACCESS_WRITE,
GEGL_ABYSS_NONE);
libgimpcolor: add new object GimpColorTransform which encapsulates a cmsHTRANSFORM and does all the pixel format conversion magic. It has API to create transforms and proofing transforms, and to convert pixels arrays and GeglBuffers. Before, each place which has a transform had to keep around the transform and its input and output Babl formats, and had to implement lots of stuff itself. Now all that lives in GimpColorTransform, removing lots of logic from many places, and pretty much removing lcms from the public API entirely. This removes including <lcms2.h>, LCMS_LIBS and LCMS_CFLAGS from almost all directories and potentially allows to replace lcms by something else.
2016-05-26 03:35:54 +08:00
libgimpcolor: optimize gimp_color_transform_process_buffer() for the case we're transforming the same buffer in place.
2016-05-30 02:24:37 +08:00
while (gegl_buffer_iterator_next (iter))
{
libgimpcolor: use gimp_color_profile_get_format() in GimpColorTransform and use babl to convert between profiles if possible. With the default BABL_TOLERANCE this is about 5 times faster than lcms on my test image. Using babl can be disabled by setting GIMP_COLOR_TRANSFORM_DISABLE_BABL.
2017-09-03 02:32:50 +08:00
if (priv->transform)
{
cmsDoTransform (priv->transform,
iter->data[0], iter->data[1], iter->length);
}
else
{
Bug 795300 - ICC profile conversion artifacts and gamma problems Cache the fish in GimpColorTransform as suggested by Massimo.
2018-04-19 00:49:47 +08:00
babl_process (priv->fish,
libgimpcolor: use gimp_color_profile_get_format() in GimpColorTransform and use babl to convert between profiles if possible. With the default BABL_TOLERANCE this is about 5 times faster than lcms on my test image. Using babl can be disabled by setting GIMP_COLOR_TRANSFORM_DISABLE_BABL.
2017-09-03 02:32:50 +08:00
iter->data[0], iter->data[1], iter->length);
}
libgimpcolor: add new object GimpColorTransform which encapsulates a cmsHTRANSFORM and does all the pixel format conversion magic. It has API to create transforms and proofing transforms, and to convert pixels arrays and GeglBuffers. Before, each place which has a transform had to keep around the transform and its input and output Babl formats, and had to implement lots of stuff itself. Now all that lives in GimpColorTransform, removing lots of logic from many places, and pretty much removing lcms from the public API entirely. This removes including <lcms2.h>, LCMS_LIBS and LCMS_CFLAGS from almost all directories and potentially allows to replace lcms by something else.
2016-05-26 03:35:54 +08:00
libgimpcolor: optimize gimp_color_transform_process_buffer() for the case we're transforming the same buffer in place.
2016-05-30 02:24:37 +08:00
done_pixels += iter->roi[0].width * iter->roi[0].height;
libgimpcolor: add new object GimpColorTransform which encapsulates a cmsHTRANSFORM and does all the pixel format conversion magic. It has API to create transforms and proofing transforms, and to convert pixels arrays and GeglBuffers. Before, each place which has a transform had to keep around the transform and its input and output Babl formats, and had to implement lots of stuff itself. Now all that lives in GimpColorTransform, removing lots of logic from many places, and pretty much removing lcms from the public API entirely. This removes including <lcms2.h>, LCMS_LIBS and LCMS_CFLAGS from almost all directories and potentially allows to replace lcms by something else.
2016-05-26 03:35:54 +08:00
libgimpcolor: optimize gimp_color_transform_process_buffer() for the case we're transforming the same buffer in place.
2016-05-30 02:24:37 +08:00
g_signal_emit (transform, gimp_color_transform_signals[PROGRESS], 0,
(gdouble) done_pixels /
(gdouble) total_pixels);
}
}
else
{
iter = gegl_buffer_iterator_new (src_buffer, src_rect, 0,
libgimpcolor: make gimp_color_transform_process_buffer() thread-safer Avoid gegl_buffer_set_format() and simply construct the formats differently to make sure reading/writing doesn't do any color space transform. This makes it possible again to run it on different areas of the same buffer from different threads.
2018-08-03 01:25:52 +08:00
src_format,
libgimpcolor: optimize gimp_color_transform_process_buffer() for the case we're transforming the same buffer in place.
2016-05-30 02:24:37 +08:00
GEGL_ACCESS_READWRITE,
GEGL_ABYSS_NONE);
while (gegl_buffer_iterator_next (iter))
{
libgimpcolor: use gimp_color_profile_get_format() in GimpColorTransform and use babl to convert between profiles if possible. With the default BABL_TOLERANCE this is about 5 times faster than lcms on my test image. Using babl can be disabled by setting GIMP_COLOR_TRANSFORM_DISABLE_BABL.
2017-09-03 02:32:50 +08:00
if (priv->transform)
{
cmsDoTransform (priv->transform,
iter->data[0], iter->data[0], iter->length);
}
else
{
Bug 795300 - ICC profile conversion artifacts and gamma problems Cache the fish in GimpColorTransform as suggested by Massimo.
2018-04-19 00:49:47 +08:00
babl_process (priv->fish,
libgimpcolor: use gimp_color_profile_get_format() in GimpColorTransform and use babl to convert between profiles if possible. With the default BABL_TOLERANCE this is about 5 times faster than lcms on my test image. Using babl can be disabled by setting GIMP_COLOR_TRANSFORM_DISABLE_BABL.
2017-09-03 02:32:50 +08:00
iter->data[0], iter->data[0], iter->length);
}
libgimpcolor: optimize gimp_color_transform_process_buffer() for the case we're transforming the same buffer in place.
2016-05-30 02:24:37 +08:00
done_pixels += iter->roi[0].width * iter->roi[0].height;
g_signal_emit (transform, gimp_color_transform_signals[PROGRESS], 0,
(gdouble) done_pixels /
(gdouble) total_pixels);
}
libgimpcolor: add new object GimpColorTransform which encapsulates a cmsHTRANSFORM and does all the pixel format conversion magic. It has API to create transforms and proofing transforms, and to convert pixels arrays and GeglBuffers. Before, each place which has a transform had to keep around the transform and its input and output Babl formats, and had to implement lots of stuff itself. Now all that lives in GimpColorTransform, removing lots of logic from many places, and pretty much removing lcms from the public API entirely. This removes including <lcms2.h>, LCMS_LIBS and LCMS_CFLAGS from almost all directories and potentially allows to replace lcms by something else.
2016-05-26 03:35:54 +08:00
}
g_signal_emit (transform, gimp_color_transform_signals[PROGRESS], 0,
1.0);
}
libgimpcolor: add gimp_color_transform_can_gegl_copy() Which takes two profiles and returns TRUE if converting between them works correctly without a GimpColorTransform. Use it in gimp_color_transform_new() to return a NULL transform if none is needed. Took the code from gimp-gegl-loops.c.
2016-05-27 05:11:24 +08:00
/**
* gimp_color_transform_can_gegl_copy:
libgimpcolor: some doc fixes and updates
2017-06-04 03:03:25 +08:00
* @src_profile: source #GimpColorProfile
* @dest_profile: destination #GimpColorProfile
libgimpcolor: add gimp_color_transform_can_gegl_copy() Which takes two profiles and returns TRUE if converting between them works correctly without a GimpColorTransform. Use it in gimp_color_transform_new() to return a NULL transform if none is needed. Took the code from gimp-gegl-loops.c.
2016-05-27 05:11:24 +08:00
*
* This function checks if a GimpColorTransform is needed at all.
*
* Return value: %TRUE if pixels can be correctly converted between
* @src_profile and @dest_profile by simply using
* gegl_buffer_copy(), babl_process() or similar.
*
* Since: 2.10
**/
gboolean
gimp_color_transform_can_gegl_copy (GimpColorProfile *src_profile,
GimpColorProfile *dest_profile)
{
g_return_val_if_fail (GIMP_IS_COLOR_PROFILE (src_profile), FALSE);
g_return_val_if_fail (GIMP_IS_COLOR_PROFILE (dest_profile), FALSE);
if (gimp_color_profile_is_equal (src_profile, dest_profile))
return TRUE;
Initial space invasion commit in GIMP All babl formats now have a space equivalent to a color profile, determining the format's primaries and TRCs. This commit makes GIMP aware of this. libgimp: - enum GimpPrecision: rename GAMMA values to NON_LINEAR and keep GAMMA as deprecated aliases, add PERCEPTUAL values so we now have LINEAR, NON_LINEAR and PERCPTUAL for each encoding, matching the babl encoding variants RGB, R'G'B' and R~G~B~. - gimp_color_transform_can_gegl_copy() now returns TRUE if both profiles can return a babl space, increasing the amount of fast babl color conversions significantly. - TODO: no solution yet for getting libgimp drawable proxy buffers in the right format with space. plug-ins: - follow the GimpPrecision change. - TODO: everything else unchanged and partly broken or sub-optimal, like setting a new image's color profile too late. app: - add enum GimpTRCType { LINEAR, NON_LINEAR, PERCEPTUAL } as replacement for all "linear" booleans. - change gimp-babl functions to take babl spaces and GimpTRCType parameters and support all sorts of new perceptual ~ formats. - a lot of places changed in the early days of goat invasion didn't take advantage of gimp-babl utility functions and constructed formats manually. They all needed revisiting and many now use much simpler code calling gimp-babl API. - change gimp_babl_format_get_color_profile() to really extract a newly allocated color profile from the format, and add gimp_babl_get_builtin_color_profile() which does the same as gimp_babl_format_get_color_profile() did before. Visited all callers to decide whether they are looking for the format's actual profile, or for one of the builtin profiles, simplifying code that only needs builtin profiles. - drawables have a new get_space_api(), get_linear() is now get_trc(). - images now have a "layer space" and an API to get it, gimp_image_get_layer_format() returns formats in that space. - an image's layer space is created from the image's color profile, change gimpimage-color-profile to deal with that correctly - change many babl_format() calls to babl_format_with_space() and take the space from passed formats or drawables - add function gimp_layer_fix_format_space() which replaces the layer's buffer with one that has the image's layer format, but doesn't change pixel values - use gimp_layer_fix_format_space() to make sure layers loaded from XCF and created by plug-ins have the right space when added to the image, because it's impossible to always assign the right space upon layer creation - "assign color profile" and "discard color profile" now require use of gimp_layer_fix_format_space() too because the profile is now embedded in all formats via the space. Add gimp_image_assign_color_profile() which does all that and call it instead of a simple gimp_image_set_color_profile(), also from the PDB set-color-profile functions, which are essentially "assign" and "discard" calls. - generally, make sure a new image's color profile is set before adding layers to it, gimp_image_set_color_profile() is more than before considered know-what-you-are-doing API. - take special precaution in all places that call gimp_drawable_convert_type(), we now must pass a new_profile from all callers that convert layers within the same image (such as image_convert_type, image_convert_precision), because the layer's new space can't be determined from the image's layer format during the call. - change all "linear" properties to "trc", in all config objects like for levels and curves, in the histogram, in the widgets. This results in some GUI that now has three choices instead of two. TODO: we might want to reduce that back to two later. - keep "linear" boolean properties around as compat if needed for file pasring, but always convert the parsed parsed boolean to GimpTRCType. - TODO: the image's "enable color management" switch is currently broken, will fix that in another commit.
2018-07-21 20:23:01 +08:00
if (gimp_color_profile_get_space (src_profile,
GIMP_COLOR_RENDERING_INTENT_RELATIVE_COLORIMETRIC,
NULL) &&
gimp_color_profile_get_space (dest_profile,
GIMP_COLOR_RENDERING_INTENT_RELATIVE_COLORIMETRIC,
NULL))
libgimpcolor: add gimp_color_transform_can_gegl_copy() Which takes two profiles and returns TRUE if converting between them works correctly without a GimpColorTransform. Use it in gimp_color_transform_new() to return a NULL transform if none is needed. Took the code from gimp-gegl-loops.c.
2016-05-27 05:11:24 +08:00
{
return TRUE;
}
return FALSE;
}