submitted by Alexander, changes/improvements are:
- Moved to new Panel in Scene buttons "Post Effects". Together with other
postprocessing options, such as Edge render. It is also not called DoF,
this because that's a bit pretending too much then. It's a zblur still!
- Made it render Alpha as well
- Made it use and deliver float buffers
- Huge cleanup of zblur.c code, was very messy. It was alling things in render
code without need even (win matrices, transform faces, etc)
- Fixed errors in using Z values (zbuffer is signed int)
- Removed very weird gamma corrections for front/back half
- Tweaked gaussian table, allow variable 'Sigma' to be set for gauss curve
- Didn't copy 'auto focus' yet. Use of this is very limited, and gives
false expectations, nor works for rendering anims with deamons well.
Main issue remains: it's not a very advanced feature... I still doubt
very much if this deserves to be released. Spent 2 days on trying to get
the key issues solved, with not much results.
- gauss filter code has weird side effects on large blur size
- having unsharp (blurred) in front also blurs what's around in back.
only blurred in back with sharp in front works a little bit
- severe aliasing errors... also due the code splitting in 2 halves
- doesnt work with unified yet
- won't work for halos, spot halos or transparant faces
Anyhoo... It was promised to be committed, so now artists can play with it.
Who knows it's useful after all, or some fixes can be implemented. :)
(Looks like big commit, but is mostly just change of API call!)
Particle emittors now can be parented to an armature Bone, and give the
correct path for each particle. Note that this doesn't work for deform!
And, for each particle the entire armature is evaluated, all actions and
NLA strips.
It used to work a little while BTW, but the code just called ALL armatures
and made ALL deforms again. Was quite slow... thats why the API call
change: do_all_actions() now accepts Object * to only do that one. With
a NULL it now does all still. Will disapppear in recode of course...
NOTE: BLI_winstuff.h was meant to be a wrapper around windows.h to handle
undefining various crap that windows.h defines. Platform specific headers
should only have to be included in a few places. This reduces the number
of inclusions of BLI_winstuff.h to 16 which is a much more reasonable
number (than the 144 or whatever it used to be)
Render:
- New; support for dual CPU render (SDL thread)
Currently only works with alternating scanlines, but gives excellent
performance. For both normal render as unified implemented.
Note the "mutex" locks on z-transp buffer render and imbuf loads.
- This has been made possible by major cleanups in render code, especially
getting rid of globals (example Tin Tr Tg Tb Ta for textures) or struct
OSA or using Materials or Texture data to write to.
- Made normal render fully 4x32 floats too, and removed all old optimizes
with chars or shorts.
- Made normal render and unified render use same code for sky and halo
render, giving equal (and better) results for halo render. Old render
now also uses PostProcess options (brightness, mul, gamma)
- Added option ("FBuf") in F10 Output Panel, this keeps a 4x32 bits buffer
after render. Using PostProcess menu you will note an immediate re-
display of image too (32 bits RGBA)
- Added "Hue" and "Saturation" sliders to PostProcess options
- Render module is still not having a "nice" API, but amount of dependencies
went down a lot. Next todo: remove abusive "previewrender" code.
The last main global in Render (struct Render) now can be re-used for fully
controlling a render, to allow multiple "instances" of render to open.
- Renderwindow now displays a smal bar on top with the stats, and keeps the
stats after render too. Including "spare" page support.
Not only easier visible that way, but also to remove the awkward code that
was drawing stats in the Info header (extreme slow on some ATIs too)
- Cleaned up blendef.h and BKE_utildefines.h, these two had overlapping
defines.
- I might have forgotten stuff... and will write a nice doc on the architecture!
The changelog is very long... it's on the web too:
http://www.blender3d.org/cms/Mesh_editing_rewrite.425.0.html
EditMesh refactor notes (user)
**** New selection modes
When entering Edit Mode for a Mesh, you now have the choice for three selection modes. These are shown as icons in the 3D header (hotkey is being searched for!).
- Vertex Select
Select vertices as usual, fully compatible with how previous version work
- Edge Select
Vertices are not drawn anymore, and selections happen by default on the edges. It is a true edge select, meaning that you can select three out of four edges in a face, without automatic having the 4th edge selected.
- Face Select
Instead of vertices, now selection 'points' are drawn in the face centers. Selected faces also get a colored outline, like for edges. This also is true face select, for each face individual regardless selection status of its vertices or edges.
While holding SHIFT, and press a selection mode, you can also combine the above choices. Now selection becomes mixed, and will behave as expected.
For example; in Edge+Face select mode, selecting the 4 edges of a face will select the face too.
The selection modes and optional drawing modes (like transparant faces, normals, or solid drawing) all work together. All of Blender's mesh editing tools now react to the correct selection mode as well.
Most noticeable it's in:
**** Extrude
Extruding in Edge or Face Select mode allows much more precise control over what's extruded and what should be excluded. Try for example a checker pattern selection, and extrude it.
New is the fixed translation when faces are extruded. This always follows the (averaged) face normal(s) of the old face(s), enabling much easier working in 3D views . A single 'G' (Grab) or 'R' (Rotate) or 'S' (Scale) will change transform modus as usual.
**** Other things to note
- Hiding edges/faces will also behave different based on Select Mode.
- while editing, normals of faces are updated always now
- Border select (BKEY) has 2 different rules for edges; when one edge is fully inside of the border, it will only select edges that are fully inside. Otherwise it selects each edge intersecting with the border.
- in face mode, adding vertices, edges or a circle is invisible...
- "Add monkey" now works as a normal primitive (rotated and on 3d cursor)
- Mesh undo was fully recoded, hopefully solving issues now with Vertex Keys and Groups
- Going in and out of editmode was fully recoded. Especially on larger models you'll notice substantial speed gain.
**** Todo
Add 'FaceSelect mode' functionality in EditMode, including zbuffered selection, display and editing of UV texture.
EditMesh refactor notes (coder)
**** Usage of flags in general
The "->f" flags are reserved for the editmesh.c and editmesh_lib.c core functions. Actually only selection status is there now.
The "->f1" and "->f2" flags are free to use. They're available in vertex/edge/face structs. Since they're free, check carefully when calling other functions that use these flags... for example extrude() or subdivide() use them.
**** Selection flags
EditVert: eve->f & SELECT
EditEdge: eed->f & SELECT
EditFace: efa->f & SELECT
- Selection is only possible when not-hidden!
- Selection flags are always up-to-date, BUT:
if selection mode >= SELECT_EDGE vertex selection flags can be incorrect
if selection mode == SELECT_FACE vertex/edge selection flags can be incorrect
This because of shared vertices or edges.
- use for selecting vertices:
eve->f &= SELECT
- use for selecting edges always:
void EM_select_edge(eed, 1) // 1 = select, 0 = deselect
- use for selecting faces always:
void EM_select_face(efa, 1) // 1 = select, 0 = deselect
- To set the 'f' flags in all of the data:
void EM_set_flag_all(int flag);
void EM_clear_flag_all(int flag);
- the old faceselectedOR() and faceselectedAND() are still there, but only
to be used for evaluating its vertices
**** Code hints for handling selection
If the selectmode is 'face'; vertex or edge selections need to be flushed upward. Same is true for 'edge' selection mode. This means that you'll have to keep track of all selections while coding... selecting the four vertices in a face doesn't automatically select the face anymore.
However, by using the above calls, at least selections flush downward (to vertex level). You then can call:
void EM_selectmode_flush(void);
Which flushes selections back upward, based on the selectmode setting. This function does the following:
- if selectmode 'vertex': select edges/faces based on its selected vertices
- if selectmode 'edge': select faces based its selected edges
This works fine in nice controlled situations.
However, only changing the vertex selections then still doesn't select a face in face mode! If you really can't avoid only working with vertex selections, you can use this call:
void EM_select_flush(void);
Now selection is flushed upward regardless current selectmode. That can be destructive for special cases however, like checkerboard selected faces. So use this only when you know everything else was deselected (or deselect it). Example: adding primitives.
**** Hide flags
EditVert: eve->h
EditEdge: eed->h
EditFace: efa->h
- all hide flags are always up-to-date
- hidden vertices/edges/faces are always deselected. so when you operate on selection only, there's no need to check for hide flag.
**** Unified undo for editmode
New file: editmode_undo.h
A pretty nice function pointer handler style undo. Just code three functions, and your undo will fly! The c file has a good reference.
Also note that the old undo system has been replaced. It currently uses minimal dependencies on Meshes themselves (no abuse of going in/out editmode), and is restricted nicely to editmode functions.
**** Going in/out editmode
As speedup now all vertices/faces/edges are allocated in three big chunks. In vertices/faces/edges now tags are set to denote such data cannot be freed.
ALso the hashtable (lookup) for edges uses no mallocs at all anymore, but is part of the EditEdge itself.
Moved get_con_target from editconstraint to constraint kernel file and renamed to get_constraint_target
Make scene.c use constraint functions from kernel instead of doing the switch itself (regrouping specific constraint code in one file).
Add LockTrack as an option in the Make Track menu (Ctrl-T)
Adding fun for vortex forces in 3D window, the representation now match the orientation and direction of the force (this is code will need optimisation). I'll probably be slapped for mixing two things in one commit but some function calls needed to be fixed for the new spelling and since this is really isolated code I didn't feel like wasting time getting a fresh copy, doing the changes and then readded this after commiting.
Fixed some warning regarding implicit casts and constants (this last one is a MSVCism where you need to explicitely tell the compiler to use floats with an f. AFAIK, it doesn't affect/bother other compilers).
Removed a useless commented piece of code in readfile (there was an uncommented copy a couple of lines before)
different objects shouldn't share flags this way (still sharing of
other mesh flags in renderer... ickity pickity, but I'm not fixing now)
- removed some unnecessary uses of DNA_mesh_types.h
* Armatures that had constraint targets inside other armatures
weren't transform()-ing correctly
* Issues with lattice deformed objects that are parents of bones
when rendering an animation. Seems to be mostly OK now with the
exception of the first rendered frame -- weird bugs like this have
plagued blender for ages, which leads me to believe that
RE_rotateBlenderScene() is a piece of garbage that nobody understands
(especially me).
* made a few helper functions to clean up some repeated code related
to clearing constraint status and rebuilding displists.
Materials are exported the best we can do by now. It will look almost as in
blender except for the missing procedural textures and some minor issues.
You have to tweak normal modulation amount to get the desired result cause
is not the same in yafray.
We added a panel in render space to adjust some yafray settings (GI and so)
Also we export transparency and reflection using new raytracing settings,
but that will be changed and improved soon.
Remember that you have to set YFexport path in user defaults and yafray must
be on path (version 0.0.6)
We added the "yafray" button to activate all this stuff in the render window.
Panel and settings are only shown when checked.
So now when activated the code calls yafray export instead of the internal
renderer and finally the resulting image is loaded back into render window's
buffer. So animation is also possible and results can be saved using blender
usual scheme.
Main target was to make the inner rendering loop using no globals anymore.
This is essential for proper usage while raytracing, it caused a lot of
hacks in the raycode as well, which even didn't work correctly for all
situations (textures especially).
Done this by creating a new local struct RenderInput, which replaces usage
of the global struct Render R. The latter now only is used to denote
image size, viewmatrix, and the like.
Making the inner render loops using no globals caused 1000s of vars to
be changed... but the result definitely is much nicer code, which enables
making 'real' shaders in a next stage.
It also enabled me to remove the hacks from ray.c
Then i went to the task of removing redundant code. Especially the calculus
of texture coords took place (identical) in three locations.
Most obvious is the change in the unified render part, which is much less
code now; it uses the same rendering routines as normal render now.
(Note; not for halos yet!)
I also removed 6 files called 'shadowbuffer' something. This was experimen-
tal stuff from NaN days. And again saved a lot of double used code.
Finally I went over the blenkernel and blender/src calls to render stuff.
Here the same local data is used now, resulting in less dependency.
I also moved render-texture to the render module, this was still in Kernel.
(new file: texture.c)
So! After this commit I will check on the autofiles, to try to fix that.
MSVC people have to do it themselves.
This commit will need quite some testing help, but I'm around!
Updated the Make environment to point to the correct location. The include
paths were still pointing to source/blender/bpython/include while it should be
source/blender/python.
I did not encounter the build problems because I'm always working with the
autoconf build environment.
This enables Blender to;
- have scenes with different codec settings. (same as avicodec)
- render directly without dialog.
- batch/background render to Quicktime movies.
Only tested on windows. ;)
