Even though the ACES RRT is a filmic mapping, and the current code is only an
approximation of the luminance aspects (ignoring saturation and hue sweeteners)
the intent of this display filter will be to implement a non-LUT based close to
standard conformant parametric ACES RRT in C for use in ACES workflows.
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.
It seems I forgot to also install python plug-ins in a subdirectory
(because they are deactivated on master for the time being). I even
found a year-old patch from myself rotting in the tracker, and which was
doing exactly this.
I redid it, improved to fit changes.
And connect to this property in the extension manager to allow dynamic
reload when starting/stopping an extension. It doesn't work yet for
extensions with plug-ins but works with other kinds of data.
When a user extension with same id as a system extension is installed,
set the system extension unsensitive in the GUI (as well as extensions
with missing requirements, though it is not implemented yet).
The reason why I broke this change in 2 commits (with the previous
commit) is because it will allow to easily cherry-pick the core code
commits whereas the GTK+ code will have to be reimplemented for
gimp-2-10 since this widget uses GTK+3 API.
Add gimp_extension_manager_can_run() and rename function:
s/gimp_extension_manager_is_active/gimp_extension_manager_is_running/
A system extension cannot be run if it has been overrided by a
user-installed extension of same ID. Moreover extensions can have GIMP
version requirements as well as have dependency on other extensions
(though these are not implemented yet).
We only save the active state of extensions so that we can reload all
extensions same as they were at previous exit. All other data are saved
as per-extension metadata and should not be saved in the rc file.
If an extension is not listed in extensionrc, we run it by default if
this is a system extension (so that new core extensions by the GIMP team
are run when installed after an updated), but not when they are
user-installed extensions.
Add a display filter that can be used in scene-referred imaging workflows,
this is using a luminance-only approximation of the ACES filmic HDR to SDR
proofing mapping - see link in source comment for details.
In the dynamics editor, use the neutral curve color, which depends
on the current theme, for the "fade" curve, instead of hard-coding
it to dark gray, which is hardly visible with dark themes.
... depending on UI language
Use gimp_register_file_handler_priority(), added in the previous
commit, to increase the priority of the merged PSD loader, so that
the non-merged loader is always preferred over it (unless
explicitly chosen), regardless of the UI language.
Add a gimp-register-file-handler-priority procedure, which can be
used to set the priority of a file-handler procedure. When more
than one file-handler procedure matches a file, the procedure with
the lowest priority is used; if more than one procedure has the
lowest priority, it is unspecified which one of them is used. The
default priority of file-handler procedures is 0.
Add the necessary plumbing (plus some fixes) to the plug-in manager
to handle file-handler priorities. In particular, use two
different lists for each type of file-handler procedures: one meant
for searching, and is sorted according to priority, and one meant
for display, and is sorted alphabetically.
In GimpToolCompass (and, as a consequence, in the measure tool),
update the measured angle when the shell is scaled, rotated, or
flipped, so that we always satisfy the compass's constrains, and
render correctly.
When using a GimpToolCompass in 3-point mode, add a small gap after
the angle arc to the line corresponding to the "second" non-origin
point, so that it's visually distinguishable from the line
corresponding to the "first" point. This has significance for the
measure tool, since it determines the direction of the rotation
when straightening the image (the first point is rotated toward the
second point.)
... does not restore the measure points
Halt the measure tool after straightening, thus removing the
expectation that undoing the operation should restore the original
points.
Halting the tool, rather than making undo work "as expected",
sidesteps several issues:
- Implementing undo correctly is tricky, since image-undo and
tool-undo are handled separately.
- In fact, the measure tool doesn't provide tool-undo, so that
image edits can be undone while the tool is active without
affecting the tool, and it's not clear that we want to change
this behavior. However, this makes undoing a straighten
operation a special case, and it's not entirely clear what the
behavior should be when undoing other kinds of transformations,
or when the measure points had changed since the straighten
operation.
- Perhaps most importantly, measure tool points are restricted to
the pixel grid, which means that when measuring an angle
against an orientation that's not fully horizontal or vertical
in image space (either using a 3-point angle, or when the
canvas is rotated), the resulting transformed point after
straightening doesn't generally land on the pixel grid, causing
it to be rounded, which can result in a non-zero angle after
the rotation. This is especially ugly, since clicking
"straighten" again at this point would cause another non-
trivial rotation.
Bind the "orientation" property of the measure-tool options to the
tool's compass widget's "orientation" property, instead of manually
synchronizing their values.
Rename XCF property PROP_SAMPLE_POINT to PROP_OLD_SAMPLE_POINT and add
new PROP_SAMPLE_POINT.
The new property saves the sample point's pick mode plus some padding
for whatever else we might want to add. Always save the old property
too so nothing changes for older GIMP versions, and avoid loading the
old property if the new one was loaded before.
Add an "orientation" option to the measure tool, corresponding to
the "orientation" property of GimpToolCompass (i.e., it controls
the orientation against which the angle is measured, when not in 3-
point mode.) The orientation is "auto" by default, so that the
angle is always <= 45 deg. Note that the "orientation" option
affects the tool's "straighten" function, so that the layer is
rotated toward the current orientation.
Use the "pixel-angle" and "unit-angle" properies of
GimpToolCompass to read the measured angle, instead of duplicating
the angle-measurement logic, in particular, so that we benefit from
the improvements/fixes of the previous commit.
Add an "orientation" property to GimpToolCompass, which can be one
of "auto", "horizontal", or "vertical", and which controls the
orientation of the line against which the angle is measured, when
not in 3-point mode (previously, the line would always be
horizontal.) When "orientation" is "auto", the orientation is
automatically set to either horizontal or vertical, such that the
measured angle is <= 45 deg.
Keep the line horizontal, or vertical, in display-space, rather
than in image-space, so that the compass works correctly even when
the canvas is rotated and/or flipped.
Fix the compass's behavior when the image's horizontal and vertical
resolutions are different, both with and without dot-for-dot.
Add "pixel-angle" and "unit-angle" read-only properties, which
return the measured angle either with or without taking the image's
resolution into account, respectively. These properties will be
used by the measure tool in the next commit, instead of having it
implement its own angle calculation.
Fix gimp_constrain_line() and friends to properly constrain line
angles when the image's horizontal and vertical resolutions are
different, and dot-for-dot is disabled.
Remember the sample point's GimpColorPickMode in the sample point
itself, so it is remembered across switching between images.
Not persistent in the XCF yet tho...
Øyvind Kolås