Commit 9d4084c82f skips conversion and
blending of (some) transparent source and destination pixels. When
`blend_out == blend_layer`, it banks on the fact that the alpha values
of `blend_out` would be the same as those of `blend_layer`, and hence
the same as those of `layer`; thing is, we only copy those values from
`layer` to `blend_layer` for the pixels that we *don't* skip, so this
assumption is just wrong :P This leaves us with bogus alpha values in
`blend_out` for the skipped pixels, when the above equality holds.
For composite modes that use the alpha values of `blend_op` (aka `comp`)
even for transparent input pixels (i.e., src-atop and src-in), this may
result in artifacts.
Fix this by simply initializing the alpha values of `blend_out` for
skipped pixels unconditionally.
from gimpdir/tool-options/ to gimpdir/filters/, and only if moving
fails try reading from the old location as fallback. We don't normally
move files around, but this one-liner nicely avoids cluttering
gimpdir.
Filters settings used to be serialized and deserialized only
when a filter tool's GUI was shown, too late for the code that
re-runs/re-shows filters with previous values.
Move the entire loading/saving code to gimp-operation-config.c, even
adding/removing the dummy separator item between timestamped automatic
history and manually saved settings. Load the settings automatically
when a settings container is requested, but still trigger saving from
the few places the container is changed in the GUI; could also
automate that later.
This commit also moves all settings of filters that have their own
tools from gimpdir/tool-options/ to gimpdir/filters/. Add compat code
to try the old filename if the new doesn't exist, so files are
migrated automatically.
WIP, but this step already fixes the bug.
which unlike HSL Lightness is actually physically meaningful and
also generally speaking much more useful than HSL Lightness.
Change "Lightness" to "Lightness (HSL)" to make it clear that
the "Lightness" in the Colors/Desaturate/Desaturate menu is not the
same as "Lightness" in LAB/LCH.
For completeness add the option to desaturate to "Value (HSV)".
Add links in app/operations/gimpoperationdesaturate.c
to the Wikipedia article with definitions of L/I/V in HSL/HSI/HSV.
... since that's the color space it actually works in.
Keep the legacy "Color (HSV)" mode's name as is, wrong as it is,
since, well, that's what it used to be called...
No need to do full back and forth RGB/HSV conversions.
Change the behavior such that fully desaturated values remain
desaturated, instead of saturating towards red.
Pixels whose source or destination alpha is zero are not blended, and
therefore do not need to be converted between the composite and blend
spaces (assuming a conversion is necessary to begin with.) When there
is a large enough segment of consecutive pixels that don't need
blending, split the conversion/blending process around it, so that
we don't convert too many unblended pixels unnecessarily.
For layers with lots of transparency, this can dramatically reduce
compositing time; for layers with no transparency, the added
overhead is rather negligible.
Merge mode lays the source layer on top of the destination, same as
normal mode, however, it assumes the source and destination are two
parts of an original whole, and are therefore mutually exclusive.
This is useful for blending cut & pasted content without artifacts,
or for replacing erased content in general.
Calculates the dot product of the two input colors, and uses that
as the value for all the output color's components. Basically,
a per-pixel mono mixer.
Useful for custom desaturation, component extraction, and crazier
stuff (bump mapping!)
Include erase mode in the menu for layers and general paint tools.
This makes the eraser tool somewhat unnecessary, but allows for
interesting use cases (e.g., airbrush eraser, etc.)
... possibly due to small win32 stack
Limit the number of samples processed in one go by gimp_composite_blend()
so that we don't overflow the stack when we alloca() buffers on it.
... and get rid of the dedicated op. This gives us support for all
the blend/composite options for this mode.
Rename COLOR_ERASE to COLOR_ERASE_LEGACY, with perceptual blending/
compositing and immutable everything, and add a new COLOR_ERASE
mode, defaulting to linear blending/compositing, with mutable
everything. Modify affected code.
These are more general, and more expensive, versions of the non-
subtractive compositing functions. They are used with modes that
specify the SUBTRACTIVE flag. This doesn't affect anything yet, but
the next commit ports color-erase mode to a blendfun.
Most modes only modify the *color* of overlapping dest/src regions,
however, erase and color-erase may also reduce their *alpha*, i.e.,
eliminate some of the overlapping content. Flag these modes with
the new SUBTRACTIVE flag, as they require more general compositing
code. The next commit adds the said code.
The mode group switching combo box is hard to discover, until we use the
default group instead of legacy group as default - it is better to make legacy
resemble the full old set to.
Need to provide the pixels in a format that matches the profile,
simply using "RGBA float" here was a brain bug of mine. Two profiles
and two formats are parameters the used GimpProfileTransform needs to
work correctly.
which returns an array of modes in the order they would appear in a
GimpLayerModeContext's UI (like tool options or the layers dialog),
without the separators.
Use it in context-commands.c and layers-commands.c instead of static
(and outdated) arrays for the actions that cycle through modes.
Instead, add a gimp_layer_mode_get_format() function, which takes
the layer mode, composite space, and blend space, and returns the
I/O format.
Currently, we always use the composite space format as the I/O
format. This simplifies gimp_composite_blend(), and gives us
composite-space support for the "special" layer mode ops for free.
Replace the 'with-behind' and 'with-replace' properties with a
single 'context' property, and use it to select the included
layer modes, according to their context mask.
Add a dummy GIMP_LAYER_MODE_SEPARATOR value to the GimpLayerMode
enum, and use it to explicitly mark the menu separators in the
layer-mode group arrays; add separators to the layer-mode menu
accordingly.
Update the rest of the code to use 'context' instead of 'with-behind'
and 'with-replace'. In particular, in the layers and layer options
dialogs, select the right context based on whether or not the
selected layer is a group.
A bitmask, specifying in which contexts a layer mode is applicable.
Can be a combination of:
- LAYER: usable as a layer mode for actual layers.
- GROUP: usable as a layer mode for layer groups. Currently, all
modes that specify LAYER also specify GROUP, and vice versa,
but the planned pass-through mode will be GROUP only.
- PAINT: can be used as a paint mode.
- FADE: can be used for fading.
Add a 'context' field to _GimpLayerModeInfo, and provide context
masks to all the modes.
Use the context mask for validation when setting a layer's mode.
The next commit will use the mask when populating the layer mode
menus.
Add a "paint_composite_mode" field to GimpLayerModeInfo and set the
mode of the eraser to SRC_ATOP. Defaulting to SRC_OVER for all
painting didn't quite do it for all modes.
set all legacy modes to completely immutable and the LAB modes'
blend mode to immutable. Change GimpLayer setters and the UI
accordingly. Remove the LAB color spaces from the GUI, they can
only be used with the LAB blend modes anyway and not changed.
They can be affected by the same problem described in
commit 4c3a772cd8, although in the
case of SRC_ATOP, the affected pixels are always fully transparent.
When the source alpha is zero, we don't calculate the blended color,
so `comp[b]` can be infinite or NaN, in which case the expression
`in[ALPHA] * (comp[b] - layer[b])` is NaN, rather than the expected
value of zero.
and to operations/layer-modes/, respectively.
Add gimp_layer_modes_init() which asserts on the correct order of the
GimpLayerModeInfo array, and switch to accessing the array directly in
gimp_layer_mode_info().
Similar to the Photoshop mode of the same name. Assigns
either 0 or 1 to each of the channels, depending on whether the
sum of source and destination channel values is less than, or
greater than (or equals to), one, respectively.
This is equivalent to inverting the source, and using it to perform
per-pixel, per-channel threshold against the destination, which is
useful for various effects.
Stuff like passing "input" directly if "aux"'s opacity is 0, etc.
Used to be partly handled by normal mode, even though it applies
to other modes too.
Adjust the logic for the new compositing modes.
Add a GimpLayerModeAffectMask enum, and a corresponding
get_affect_mask() function to GimpOperationLayerMode, which
specifies which of the op's inputs, if any, are affected by the
mode, apart from the overlapping regions. Most modes affect only
the overlapping regions, but dissolve and replace also affect the
rest of the input. This information is used for determining if
the optimizations are applicable.
Some of the generic files still contain SSE2 code, in particular
gimpoperationlayermode.c. The reason why it often works without is that
native gcc will usually pre-define SSE macros by default.
To check this: gcc -dM -E - < /dev/null | grep SSE
Yet I had a case on a small netbook where the SSE macros were not
pre-defined even though supported. Consequently the build failed.
Largely based on a patch by Ell, with the enum type renamed and
various small changes. Adds another axis of configurability to the
existing layer mode madness, and is WIP too.
These variations on darken only and lighten only have the advantage over the
componentvise versions that they always use the full triplet of either original
or new layer - meaning no new colors/hues will be introduced. This is similar
to how these modes operated/operates in picture publisher and photo-paint.
The porting from 8bit per component scaled some 8bit fractions up to huge
floating point numbers, this works for most values but causes trouble for near
transparent pixel values. This commit copies the inner blend loop from the new
divide layer mode, but keeps the old compositing logic.
... is fully transparent, instead of just src.
The blend func results only affect the intersection of dest and src.
Run time is currently dominated by the compositing step for most modes,
so the difference in performance is pretty negligible, but it does make
a difference for the more expensive modes, like the HSV ones.
For operations needing to override default behavior sub-classes should still be
used.
This commit also enables pinligh, vividlight and linearlight blend mode modes
Since CIE Lab is one of the supported color spaces for doing the blending -
this enum is not only about transfer functions/curves or gamma. This finishes
already started cleanups.
It is used both for blending and compositing, the repeated use of the word
BLEND in code made the logic involving both blending and compositing hard to
read.
Note that in some cases the alloca may be unnecessary, but this keeps
the code clean, and we have to be able to *potentially* do the alloca
anyway, so what the hell.
Implement a common utility function gimp_blend_composite that uses utility
functions for implementing layer modes, with separate (possibly SIMD) optimized
loops for blending and compositing, with configured linear TRC, perceptual
gamma TRC or even using CIE Lab as the space.
like all other function typedefs and add GimpBlendFunc typedef which
will be needed soon. Also rename get_layer_mode_function() to
gimp_get_layer_mode_function().
Ell