After testing and feedback, I've decided to slightly modify the way color
management works internally. While the previous method worked well for
rendering, was a smaller transition and had some advantages over this
new method, it was a bit more ambiguous, and was making things difficult
for other areas such as compositing.
This implementation now considers all color data (with only a couple of
exceptions such as brush colors) to be stored in linear RGB color space,
rather than sRGB as previously. This brings it in line with Nuke, which also
operates this way, quite successfully. Color swatches, pickers, color ramp
display are now gamma corrected to display gamma so you can see what
you're doing, but the numbers themselves are considered linear. This
makes understanding blending modes more clear (a 0.5 value on overlay
will not change the result now) as well as making color swatches act more
predictably in the compositor, however bringing over color values from
applications like photoshop or gimp, that operate in a gamma space,
will give identical results.
This commit will convert over existing files saved by earlier 2.5 versions to
work generally the same, though there may be some slight differences with
things like textures. Now that we're set on changing other areas of shading,
this won't be too disruptive overall.
I've made a diagram explaining the pipeline here:
http://mke3.net/blender/devel/2.5/25_linear_workflow_pipeline.png
and some docs here:
http://www.blender.org/development/release-logs/blender-250/color-management/
Brought back default effect fading: (adding a wipe effect makes it
wipe by default for the length of the strip)
First round in upgrading IPOs from older versions. (works for non-IPO
case now and sets at least the new "default effect fade"-flag)
Still non-working for old IPOs, since Sequence-Strips aren't real IDs!
And: non-frame-lock case should stretch the FCurve to the right length!
* Set default particleedit selection mode in add scene (TODO: not yet changed in the default blend)
* Corrected names for particleedit selection mode in RNA, added icons
* Added occlude geometry flag to view3d RNA
* Converted particleedit buttons to uiItems in view3d header
* Now support a Surface mode next to the existing Volume mode. This binds
the mesh to the cage mesh surface rather than it's volume.
* Implemented reusing the bone heat weighting code.
* Advantage is that it works for cage meshes that are not volumes and that
binding is much faster.
* Weak point is that disconnected components of a mesh are not guaranteed
to stick together (same problem exists with bone heat weighting).
* Bind weights could still be compressed better to use less memory.
Example file:
http://download.blender.org/ftp/incoming/cloth_mdef_surface.blend
* Increased the number of Armature and Bone Layers from 16 to 32. Please note that older versions of Blender may not correctly resolve the layers that bones are on when loading new files.
* Newly added objects are now made by default to allow locking of 4-component rotations using 4 separate locks (i.e. one by component) instead of requiring the obscure 'W' toggle (renamed '4L' now) to be enabled first. The objects in the default scene need modifying manually though.
This patch, by Elia Sarti (vekoon), simply adds the possibility to specify the final array component of the RNA path in the path itself,
e.g. using location[0] or location["x"] or even location.x,
instead of specifying this using an "array_index"
This should be easier for users to understand the driver system. The array-indices have been kept (but hidden from the UI under standard situations) since they are theoretically a tad faster than the in-path lookups still, and are easier for internal-tools to set for now...
- Drivers on added to the 'armature' datablock (i.e. keyframing some settings for a "Bone" as opposed to "PoseBone") now evaluate correctly. Added proper recalcs for this case too.
- Also fixed some memory leaks and loading problems I encountered with the test file provided. After having problems loading the test file, I ended up reproducing and finding the error.
The Animation System now respects the Transform Locks too (i.e. lock x-location, etc.) when writing settings. This means that it is no longer necessary to set up "constant drivers" to make sure some values don't get accidentally animated.
Internally, added a new callback for properties in RNA, which is responsible for checking if the item at some array-index is editable. This needs to be manually called for each place which uses rna to set settings for arrays (see the code changes in anim_sys.c for changes how to do this; the same thing needs to be done in the UI code too, and probably in py-api too)
* Fixed a few compile warnings for scons+mingw
* Driver variables are now added with the ID-type set to ID_OB (objects) by default since this is more convenient
duplicate_pose_channel_data(), the code to do this was inline in editarmature.c
duplicating editbones now duplicates posebone id-props
also removed an if test for &channew->constraints since it will always be true.
This should make animations in sequencer work again at most
places.
It removes facf0 and facf1 and replaces them by
* effect_fader (the fader position for transition effects
limited 0-1 value range)
* speed_fader (full range fader for speed effect)
Also: default transitions should work again.
Still not working: non-IPO-locked curves. (Don't really know,
if we can / want to bring them back in new animation system.)
Rather than applying the modifier to the object data, it will create a new shape
with the deformed vertices in there. Only mesh at the moment, other object
types on the todo.
sb->lcom : Center Of Mass .. might be used to create loc IPO
sb_>lrot : is a matrix[3] esitmates the roatation in world coordinates .. might be used to create rot IPO
sb_>lscale : is a matrix[3] esitmates the scaling in world coordinates .. might be used to create scale IPO
(no python for that yet .. but may be matt has mercy on me )
can be cropped direclty in soft body module by function
static void SB_estimate_transform(Object *ob,float lloc[3],float lrot[3][3],float lscale[3][3])
The targets lloc,lrot,lscale should work to be NULL, just in case you don't need it.
However i'd prefer if they were accessed via properties
which should be calculated automagically if
sb->solverflags & SBSO_ESTIMATEIPO
is set, like they do in draw_sb_motion(..) in drawobject.c
added static void draw_sb_motion(Scene *scene, Object *ob) to drawobject.c
for debuggering (had a hard time with destructive matrix operations )
if it causes any trouble with your build on any OS make sure to comment that away.
softbody.c and DNA should compile fine in any case.
GLEW update to version 1.5.1 [11-03-08]
this opens room for Geometry Shader support.
* - Brecht, Campbell told me you did some local changes in order to make it right in Linux. I get to you in order to know what those changes are (or feel free to commit them directly)
when changing curve twist I assumed 2D curves would not need dir or quat values set, however these are used when getting a path from a curve.
also added interp_v3_v3v3v3v3
- A single undo removes all 4 inserted spaces at once.
- Indent and unindent now use spaces.
The number of spaces is still hardcoded as 4. PS: Campbell sharing a couple of the minor todo's with me, thanks :).
* Fixed a nasty bug which meant that it was impossible to set an active node. Was caused by ntreeCopyTree() getting called when compo updates were done and clearing the active flags. The active flag clearing is only really needed for the "internal_select" case which is only used for duplicating selected nodes (from Shift-D duplicate).
* Recoded click-selection code. Was a mess of old code, bad exceptions from the old code half ported, duplicate operators, unnecessary flags/modes.
* Fixed bug #19927: compositing node groups can't be access via "tab" or ungrouped via "alt+g". Was probably related to the active group not being able to be set.
* Made resizing nodes work again. Again, this was due to the active node bug.
* Made adding a new group with Ctrl-G correctly update the views
* Added tablet pressure support (size/strength) for weight paint and vertex paint
* Added tablet eraser support for weight paint and vertex paint (inverts the effect of the current tool)
* Removed the old 'soft' option, now weight paint and vertex paint use the influence curve
* Made the default brush use a smooth influence curve, rather than sharp
To insert keys, use I key while hovering over the button for now, rmb clicking on the property to insert a key doesn't work (general bug for all regions except property editor - will investigate).
Doesn't convert over from old fac0 ipos on opening old files though for the time being.
* Made sequence strip names unique while I was at it, to allow strip properties to be animated properly.
only implimented min/max precision & step.
at the moment there is no way to edit these other then via python
example of setting UI limits...
>>> C.object['foo'] = 0.5
>>> C.object['_RNA_UI'] = {'foo': {'step': 0.5, 'soft_max': 10.0, 'soft_min': 0.0, 'precision': 2, 'description': 'Some setting'}}
Also fixed typo's: precission -> precision
* structural can be set to 0
* pre-roll now available through GUI and works like following:
a) Pre rolled frames are NOT cached
b) reset cache + cloth on pre roll setting change
