naming convention for Compositing:
- Render Result node -> Render Layers node (name only appears in Add menu)
- Compositor image -> Viewer Node image
I've also added a version patch (2.41 saved files only) to rename existing
"Compositor" Images.
Compositor:
Viewer Nodes inside of Groups now work too. To not frustrate interactive
speed, the following rules apply:
- Making a Group editable (or closing) doesn't signal recalculation of
composite
- clicking on an Input socket always checks changes and calulcates
- When there are Viewers inside an edited Group, Viewers in the main tree
are not executed.
Also added: a "hide unused sockets" icon in the header of Viewer nodes.
This allows cleanup of Groups, to prevent these sockets get reveiled.
This commit brings back:
- Field Render
- MBlur Render (old style)
- Border render with or without cropping
Note: Field Render is not supported in Compositor yet. Blurring or filter
will destroy field information.
Both MotionBlur as Field render are done before Compositing happens.
Fixes:
- The "Save Buffers" option only worked on single frame renders, not for
Anim render.
- Found an un-initalized variable in Render initialize... this might have
caused the unknown random crashes with render.
Code restructure:
Cleaned up names and calls throughout the pipeline, more clearly telling
what goes on in functions.
This is visible in the updated first image of the Wiki doc:
http://mediawiki.blender.org/index.php/BlenderDev/RenderPipeline
Material Nodes: The Texture node didn't do the standard "2d mapping" yet
in case an Image Texture is used. Caused wrong mapping for example for UV
coordinate inputs.
render option "save result to disk" and ESC from rendering, it crashed...
Note; reading partially saved exr files still crash... but that's an issue
within the openexr lib. I've mailed the openexr dev list for assistance
how to properly close a partial saved tile-file.
- Renderwin still used a thread-unsafe malloc, in the header text print
- Setting clipping flags in vertices for parts required a mutex lock after
all... I thought it would go fine, but noticed on renders with small
amounts of faces that sometimes faces disappear from a render.
(was doing movie credits, so all faces are visible! Otherwise it would
have hardly been noticable...)
same composit3_pixel_processor() function. gcc even gives nice warning:
node_composite.c: In function `node_composit_exec_mix_rgb':
node_composite.c:1437: warning: passing arg 9 of `composit3_pixel_processor' from incompatible pointer type
floats were read as pointer...
It still needs some work but at least now it doesn't crash.
Basically it changed composit3_pixel_processor so that its using full
vector instead of first element (fac variable in the code).
In reality though its just pushing where its looking at the first element
of the vector to another function (do_mix_rgb)
which is calling ramp_blend with fac[0].
Kent
filtersizes (below 2 pixels). This because Bokeh actually does 2 peaks...
/\ /\
/ \/ \
I've added some fixes in the filter calculus though, and made sure that
on size 1 at least the image gets copied straight away.
Also fixed error, Bokeh shifted image 1 pixel up.
Todo; make filters become real floats in size...
- Links now can be made between any socket type. The nodes recognize amount
of channels, and will convert types if needed.
Conversions from RGBA to 1 channel will use the 'RGB to BW' formula.
Also note that conversions only happen when required. So you can blur an
alpha channel, filter it, and put this in a 1-channel socket without any
conversion to happen, which saves memory & cpu time.
http://www.blender.org/bf/rt.jpg
The blur nodes don't accept Vector input yet... But filter does.
- RGB Curve Nodes now have the premultiply option resored, 2 x faster
- Fixed some confusing code in Node Group handling... much stabler now
- Texture Node
Allows to use any Blender Texture block as input for masks or color
blending. The texture node doesn't generate a real image, but adjusts to
the size as mapped with during an operation. So it won't work to use it
as Image input for Blur or Filter nodes.
Note; the Vector inputs for this node only work with manual input now!
- Translation Node
Give any image an offset in X or Y direction
For the Texture node to work, I needed to move the central 'pixel
processor' up one level... to allow differently sized images to merge
and allow 'procedural images' without size.
Temporal image of the day: http://www.blender.org/bf/rt.jpg
(NOTE: new include dependency in Render module, might need MSVC update!
It has to include the imbuf/intern/openexr/ directory in search path)
-> New Composite node: "Hue Saturation".
Works like the former 'post process' menu. There's no gamma, brightness or
multiply needed in this node, for that the Curves Node functions better.
-> Enabled Toolbox in Node editor
This now also replaces the SHIFT+A for adding nodes. The nodes are
automatically added to the menus, using the 'class' category from the
type definition.
Current classes are (compositor examples):
Inputs: RenderResult, Image
Outputs: Composite, Viewer
Color Ops: RGB Curves, Mix, Hue Saturation, AlphaOver
Vector Ops: Normal, Vector Curves, Map Value
Filters: Filter, Blur, VectorBlur
Convertors: ColorRamp, RGBtoBW, Separate RGBA, Separate HSVA, Set Alpha
Generators: RGB, Value, Time
Groups: the list of custom defined nodes
-> OpenEXR tile saving support
Created an API for for saving tile-based Images with an unlimited amount
of layers/channels. I've tested it for 'render result' now, with the idea
that this can (optionally) replace the current inserting of tiles in the
main result buffers. Especially with a lot of layers, the used memory for
these buffers can easily go into the 100s of megs.
Two other advantages:
- all 'render result' layers can be saved entirely in a single file, for
later use in compositing, also for animation output.
- on each render, per scene, a unique temp file can be stored, allowing
to re-use these temp files on starting Blender or loading files, showing
the last result of a render command.
The option is currently disabled, needs more work... but I had to commit
this because of the rest of the work I did!
-> Bug fix
The Image node didn't call an execute event when browsing another image.
they work ok in testing here and get done what I need, any checks or fixes are welcome.
* Separate RGBA: Separates an input RGBA image into its R, G, B and A channels
* Separate HSVA: Separates an input RGBA image into H, S, V and A channels
* Set Alpha: Takes an input RGBA image and an alpha value channel and combines them
into a single RGBA image channel. You can also set the alpha for the entire image
with the number field when there's no input alpha channel. TODO: Allow input alpha
channel with no input image, in order to output a solid colour, with alpha.
In Orange we've been fighting the past weeks with memory usage a lot...
at the moment incredible huge scenes are being rendered, with multiple
layers and all compositing, stressing limits of memory a lot.
I had hoped that less frequently used blocks would be swapped away
nicely, so fragmented memory could survive. Unfortunately (in OSX) the
malloc range is limited to 2 GB only (upped half of address space).
Other OS's have a limit too, but typically larger afaik.
Now here's mmap to the rescue! It has a very nice feature to map to
a virtual (non existing) file, allowing to allocate disk-mapped memory
on the fly. For as long there's real memory it works nearly as fast as
a regular malloc, and when you go to the swap boundary, it knows nicely
what to swap first.
The upcoming commit will use mmap for all large memory blocks, like
the composit stack, render layers, lamp buffers and images. Tested here
on my 1 GB system, and compositing huge images with a total of 2.5 gig
still works acceptable here. :)
http://www.blender.org/bf/memory.jpg
This is a silly composit test, using 64 MB images with a load of nodes.
Check the header print... the (2323.33M) is the mmap disk-cache in use.
BTW: note that is still limited to the virtual address space of 4 GB.
The new call is:
MEM_mapalloc()
Per definition, mmap() returns zero'ed memory, so a calloc isn't required.
For Windows there's no mmap() available, but I'm pretty sure there's an
equivalent. Windows gurus here are invited to insert that here in code! At
the moment it's nicely ifdeffed, so for Windows the mmap defaults to a
regular alloc.
- a Group has Curve node inside
- this Group was re-used more times
- with threaded render activated
- and both groups executed on same time
Then the premultipy optimize table was created twice... causing memory
to confuse.
This filter type uses a filter-image, and spreads color of current pixel
over all neighbour pixels based on this filter-image. That creates a
problem on borders... since there only parts get accumulated.
Solved by going over the to-be-filtered image pretending it is wider
exactly the amount of the filter-image size.
the Preview Panel, you got a crash... this case wasn't anticipated in
code... two cropping mechanisms on top. :)
Works OK now, but offset of cropping is drawn a bit weird... this case
needs recode a bit. On list for later.
I noticed still several cases where the Imbuf library was called within a
thread... and that whilst the Imbuf itself isn't threadsafe. Also the
thread lock I added in rendering for loading images actually didn't
work, because then it was still possible both threads were accessing the
MEM_malloc function at same time.
This commit nearly fully replaces ImBuf calls in compositor (giving another
nice speedup btw, the way preview images in Nodes were calculated used
clumsy imbuf scaling code).
I've also centralized the 'mutex' locking for threading, which now only
resides in BLI_threads.h. This is used to secure the last ImBuf calls
I cannot replace, which is loading images and creating mipmaps.
Really hope we get something more stable now!
- Appending Images now re-assigns relative paths, to match with the file
as currently being used
Bugfix:
- Memory error in creating preview render rects... gave no issues here,
but i guess that's not saying anything! Hope its more stable now. :)
You now can set a Preview panel in the Image window, to define a sub-rect
of an image to be processed. Works like the preview in 3D Window. Just
press SHIFT+P to get it activated. Very nice speedup!
This is how it works:
- The compositor still uses the scene image size (including % setting) for
Viewer or Composite output size
- If a preview exists, it calculates the cropped rect from its position
in the Image window, and stores that in the Scene render data
- On composite execute, it copies only this part from the 'generator nodes',
right now Images or Render Results. That makes the entire composite tree
only using small rects, so it will execute fast.
- Also the render window will only display the cropped rect, and on F12
only the cropped part is being executed
- On rendering in background mode, the cropping is ignored though.
Usability notes:
- translating or zooming view will automatically invoke a recalculation
- if you zoom in on details, the calculated rect will even become smaller
- only one Imagewindow can have this Preview Panel, to prevent conflicts of
what the cropped area should be. Compositing is on Scene level, not local
per image window. (Note; 3D Previews are local per window!)
- Closing the preview panel will invoke a full-size recalculation
- All passes/layers from rendering are nicely cropped, including Z and
vectors.
The work to make the compositor do cropping was simple, but getting the
Image window displaying correctly and get all events OK was a lot of work...
indeed, we need to refactor Image Window usage once. Sorry for making the
mess even bigger now. :) I've tried not to interfere with UV edit or Paint
though... only when you're in compositing mode the panel will work.
BUG fix:
3D Preview render didn't work when multiple layers were set in the current
scene.
vectors. It's actually shutter speed, but in this case works identical to
the old motionblur 'blur fac' button.
Note; the "Max Speed" button only clips speed, use this to prevent
extreme speed values. Max speed applied before the scaling happens.
After a couple of experiments with variable blur filters, I tried
a more interesting, and who knows... original approach. :)
First watch results here:
http://www.blender.org/bf/rt0001_0030.avihttp://www.blender.org/bf/hand0001_0060.avi
These are the steps in producing such results:
- In preprocess, the speed vectors to previous and next frame are
calculated. Speed vectors are screen-aligned and in pixel size.
- while rendering, these vectors get calculated per sample, and
accumulated in the vector buffer checking for "minimum speed".
(on start the vector buffer is initialized on max speed).
- After render:
- The entire image, all pixels, then is converted to quad polygons.
- Also the z value of the pixels is assigned to the polygons
- The vertices for the quads use averaged speed vectors (of the 4
corner faces), using a 'minimum but non-zero' speed rule.
This minimal speed trick works very well to prevent 'tearing' apart
when multiple faces move in different directions in a pixel, or to
be able to separate moving pixels clearly from non-moving ones
- So, now we have a sort of 'mask' of quad polygons. The previous steps
guaranteed that this mask doesn't have antialias color info, and has
speed vectors that ensure individual parts to move nicely without
tearing effects. The Z allows multiple layers of moving masks.
- Then, in temporal buffer, faces get tagged if they move or not
- These tags then go to an anti-alias routine, which assigns alpha
values to edge faces, based on the method we used in past to antialias
bitmaps (still in our code, check the antialias.c in imbuf!)
- finally, the tag buffer is used to tag which z values of the original
image have to be included (to allow blur go behind stuff).
- OK, now we're ready for accumulating! In a loop, all faces then get
drawn (with zbuffer) with increasing influence of their speed vectors.
The resulting image then is accumulated on top of the original with a
decreasing weighting value.
It sounds all quite complex... but the speed is still encouraging. Above
images have 64 mblur steps, which takes about 1-3 seconds per frame.
Usage notes:
- Make sure the render-layer has passes 'Vector' and 'Z' on.
- add in Compositor the VectorBlur node, and connect the image, Z and
speed to the inputs.
- The node allows to set amount of steps (10 steps = 10 forward, 10 back).
and to set a maximum speed in pixels... to prevent extreme moving things
to blur too wide.
- Scene support in RenderLayers
You now can indicate in Compositor to use RenderLayer(s) from other scenes.
Use the new dropdown menu in the "Render Result" node. It will change the
title of the node to indicate that.
The other Scenes are rendered fully separate, creating own databases (and
octrees) after the current scene was finished. They use their own render
settings, with as exception the render output size (and optional border).
This makes the option an interesting memory saver and speedup.
Also note that the render-results of other scenes are kept in memory while
you work. So, after a render, you can tweak all composit effects.
- Render Stats
Added an 'info string' to stats, printed in renderwindow header. It gives
info now on steps "creating database", "shadow buffers", and "octree".
- Bug fixes
Added redraw event for Image window, when using compositor render.
Text objects were not rendered using background render (probably a bug
since depsgraph was added)
Dropdown buttons in Node editor were not refreshed after usage
Sometimes render window did not open, this due to wrong check for 'esc'.
Removed option that renders view-layers on F12, with mouse in 3d window.
Not only was it confusing, it's now more efficient with the Preview Panel,
which does this nicely.
http://www.blender.org/bf/filters/
I found out current blur actually doesn't do gauss, but more did regular
quadratic. Now you can choose common filter types, but more specifically;
- set gamma on, to emphasize bright parts in blur more than darker parts
- use the bokeh option for (current circlular only) blur based on true
area filters (meaning, for each pixel it samples the entire surrounding).
This enables more effects, but is also much slower. Have to check on
optimization for this still... use with care!
- Enabled Groups to execute in Compositor. They were ignored still.
Note; inside of groups nothing is cached, so a change of a group input
will recalculate it fully. This is needed because groups are linked
data (instances use same internal nodes).
- Made Composit node "Viewer" display correctly input for images with
1/2/3/4 channels.
- Added pass rendering, tested now with only regular Materials. For
Material nodes this is quite more complex... since they cannot be
easily separated in passes (each Material does a full shade)
In this commit all pass render is disabled though, will continue work on
that later.
Sneak preview: http://www.blender.org/bf/rt.jpg (temporal image)
- What did remain is the 'Normal' pass output. Normal works very nice for
relighting effects. Use the "Normal Node" to define where more or less
light should be. (Use "Value Map" node to tweak influence of the
Normal node 'dot' output.)
- EVIL bug fix: I've spend almost a day finding it... when combining AO and
mirror render, the event queue was totally screwing up... two things not
related at all!
Found out error was in ray-mirror code, which was using partially
uninitialized 'ShadeInput' data to pass on to render code.
- Another fix; made sure that while thread render, the threads don't get
events, only the main program will do. Might fix issues reported by
people on linux/windows.
- Live scanline updates while rendering
Using a timer system, each second now the tiles that are being processed
are checked if they could use display.
To make this work pretty, I had to use the threaded 'tile processor' for
a single thread too, but that's now proven to be stable.
Also note that these updates draw per layer, including ztransp progress
separately from solid render.
- Recode of ztransp OSA
Until now (since blender 1.0) the ztransp part was fully rendered and
added on top of the solid part with alpha-over. This adding was done before
the solid part applied sub-pixel sample filtering, causing the ztransp
layer to be always too blurry.
Now the ztransp layer uses same sub=pixel filter, resulting in the same
AA level (and filter results) as the solid part. Quite noticable with hair
renders.
- Vector buffer support & preliminary vector-blur Node
Using the "Render Layer" panel "Vector" pass button, the motion vectors
per pixel are calculated and stored. Accessible via the Compositor.
The vector-blur node is horrible btw! It just uses the length of the
vector to apply a filter like with current (z)blur. I'm committing it anyway,
I'll experiment with it further, and who knows some surprise code shows up!
for compositing code.
Officially malloc/calloc/free is threadsafe, but our secure malloc system
requires all memory blocks to be stored in a single list, so when two
threads write in this list you get conflicts.
- Compositor now is threaded
Enable it with the Scene buttons "Threads". This will handle over nodes to
individual threads to be calculated. However, if nodes depend on others
they have to wait. The current system only threads per entire node, not for
calculating results in parts.
I've reshuffled the node execution code to evaluate 'changed' events, and
prepare the entire tree to become simply parsed for open jobs with a call
to node = getExecutableNode()
By default, even without 'thread' option active, all node execution is
done within a separate thread.
Also fixed issues in yesterdays commit for 'event based' calculations, it
didn't do animated images, or execute (on rendering) the correct nodes
when you don't have Render-Result nodes included.
- Added generic Thread support in blenlib/ module
The renderer and the node system now both use same code for controlling the
threads. This has been moved to a new C file in blenlib/intern/threads.c.
Check this c file for an extensive doc and example how to use it.
The current implementation for Compositing allows unlimited amount of
threads. For rendering it is still tied to two threads, although it is
pretty easy to extend to 4 already. People with giant amounts of cpus can
poke me once for tests. :)
- Bugfix in creating group nodes
Group node definitions demand a clear separation of 'internal sockets' and
'external sockets'. The first are sockets being linked internally, the latter
are sockets exposed as sockets for the group itself.
When sockets were linked both internal and external, Blender crashed. It is
solved now by removing the external link(s).
Here are some notes written by Ton:
There's several areas I still have to work on, so don't immediately
report bugs!
- Preview renders in buttons window only supports materials now (no
lamp, texture or world)
- Yafray code has to be checked, it might not work
- unified render doesn't work, might not come back even
- lens flares dont work yet
- motion blur and field render doesnt work yet
- exr libraries are static linked still, we look at making it dynamic
loadable
- the compositor is fully in development still
I really advise people to check on the orange cvs logs too (in case you
ignored it), this has most of the information in it.
At this moment I still have to code for Orange deadlines, that defines
my priority listing mostly. And since we don't use Windows there, doing
testing and bugfixing for that platform would be of great help!
Best features of this commit to checkout:
- Grouping
- full fixed library linking system
- custom bone drawing
- bone layers
- Material/Shading nodes
- recode of render engine
- preview renders
- Compositing options
When the work pressure is a bit lower, I'll make nice docs for all of
this!
-Ton-
