Is there a way to create a new hou.Tool that is embedded in a digital asset via HOM? You can obviously add your own in the Type Properties dialog - and I suppose you could hack the xml in HDADefinition.sections() - but that's… hacky.
Are there some methods I'm missing for creating and/or fetching tools that live in an HDA?
Thanks!
Found 18 posts.
Search results Show results as topic list.
Technical Discussion » Create a hou.Tool in a HDADefinition via HOM
- Brian Walters
- 155 posts
- Offline
Technical Discussion » queryWorldPositionAndNormal
- Brian Walters
- 155 posts
- Offline
Technical Discussion » hou.Node.outputNode
- Brian Walters
- 155 posts
- Offline
Is there a method that gets the “output” node for a given dive-able node with sop children? One that would return any “output” nodes inside, or else the render flag?
e.g.:
e.g.:
n = hou.node("/obj/geo1/subnet1") n.displayNode() # Node with display flag n.renderNode() # Node with render flag n.outputNode() # Doesn't exist. I want the same logic as what's output from an HDA.
Houdini Lounge » QuickMarks for SceneViewer?
- Brian Walters
- 155 posts
- Offline
I could have sworn I saw a comment about how there were quickmarks for the scene viewer like there are for the network editor… but for the life of me I can't find any documentation on it/guess the hot keys.
Am I crazy? Do quickmarks exist for the sceneviewer? (If not, that'd be ooooh so awesome.)
Am I crazy? Do quickmarks exist for the sceneviewer? (If not, that'd be ooooh so awesome.)
Technical Discussion » Faster launch times?
- Brian Walters
- 155 posts
- Offline
Hey Olaf,
Unforunately, no, I cannot post an example hip file. That's why I was asking for general tips and tricks in debugging slow launch times.
Launching a new houdini session and then opening a “heavy” hip file (with Manual cook mode enabled) is still quite slow. :-|
Unforunately, no, I cannot post an example hip file. That's why I was asking for general tips and tricks in debugging slow launch times.
Launching a new houdini session and then opening a “heavy” hip file (with Manual cook mode enabled) is still quite slow. :-|
Technical Discussion » Faster launch times?
- Brian Walters
- 155 posts
- Offline
The old stop-watch on my phone. Start as soon as the splash screen pops-up, and stop as soon as the GUI is visible.
Technical Discussion » Faster launch times?
- Brian Walters
- 155 posts
- Offline
I'm talking pure load times. Time to cook is obviously dependent on what the user is doing.
Technical Discussion » Faster launch times?
- Brian Walters
- 155 posts
- Offline
What are the best ways to diagnose Houdini hip file load times? Or does anybody have any tricks they've learned for speeding Houdini up?
I'm especially interested in actual hip file load times. We know the times it takes for our pythonrc.py and 123/456 startup scripts to run. But what about loading an actual .hip file (the time after the pythonrc.py and pre the 123.py)? How can we best determine what's making that go slow?
We're seeing hip files take 5~ minutes to load that are a mere 19 MB that when playing around in the network editor don't appear to be all that heavy.
I'm especially interested in actual hip file load times. We know the times it takes for our pythonrc.py and 123/456 startup scripts to run. But what about loading an actual .hip file (the time after the pythonrc.py and pre the 123.py)? How can we best determine what's making that go slow?
We're seeing hip files take 5~ minutes to load that are a mere 19 MB that when playing around in the network editor don't appear to be all that heavy.
Houdini Lounge » Enable Lock View to Camera in Python/Hscript
- Brian Walters
- 155 posts
- Offline
Technical Discussion » String evaluation in shaders
- Brian Walters
- 155 posts
- Offline
Is string processing in PBR shaders inefficient? I was told this kills arnold Arnold, but I wasn't sure about Mantra. For example, would it be bad to get my diffuse color map from a snippet:
texture_path = s@mtl_dir + “/” + s@asset_name“ + ”.rat";
Where both mtl_dir and asset_name are primitive/detail attributes?
texture_path = s@mtl_dir + “/” + s@asset_name“ + ”.rat";
Where both mtl_dir and asset_name are primitive/detail attributes?
Technical Discussion » What are your favorite shading and rendering optimizations?
- Brian Walters
- 155 posts
- Offline
What are your favorite tricks to optimize your shader and rendering networks? (I'm mostly interested in PBR).
For example…
* I always set motion factor to 1.0
* I always convert all textures to RAT files pre-render
For example…
* I always set motion factor to 1.0
* I always convert all textures to RAT files pre-render
Technical Discussion » Min/Max values of a COP
- Brian Walters
- 155 posts
- Offline
I'm really hoping to avoid scattering X number of particles on a grid, sampling the map, and then attribute promoting to a detail the min/max of the map…
Technical Discussion » Min/Max values of a COP
- Brian Walters
- 155 posts
- Offline
Is there a way to get the min/max values for a channel in a COP? (Min/Max of the ENTIRE image - basically, I want to know if a image is constant/uniform or not).
Edited by Brian Walters - 2016年8月29日 12:16:14
Technical Discussion » OIIO support?
- Brian Walters
- 155 posts
- Offline
Does Mantra support OIIO? Or are we still tied to the awesome (but proprietary) rat format?
Technical Discussion » HDA Representative Node + Visualizing other Geo
- Brian Walters
- 155 posts
- Offline
I've created an HDA that contains two nodes: a camera and a geometry for visualizing the title safe region.
/obj/my_hda_camera/internal_camera
/obj/my_hda_camera/viz_title_safe_region
If I look through the internal_camera node, I can see the geometry for the “viz_title_safe_region”. However, if I change the HDA's Representative Node to be the “internal_camera” - then the “viz_title_safe_region” geometry is no longer visible when looking through the camera.
What gives? Is there a way to still see geometry when using a representative node type?
/obj/my_hda_camera/internal_camera
/obj/my_hda_camera/viz_title_safe_region
If I look through the internal_camera node, I can see the geometry for the “viz_title_safe_region”. However, if I change the HDA's Representative Node to be the “internal_camera” - then the “viz_title_safe_region” geometry is no longer visible when looking through the camera.
What gives? Is there a way to still see geometry when using a representative node type?
Technical Discussion » Visualizing Instances + Cd + OGSL
- Brian Walters
- 155 posts
- Offline
Technical Discussion » Visualizing Instances + Cd + OGSL
- Brian Walters
- 155 posts
- Offline
I have a system where an asset is instanced onto a large # of scattered points. The asset has a “Cd” attribute, and the particles have a “CdMult” attribute. I have a simple shader that takes the Cd and the CdMult and multiplies the two for the final diffuse color. It's pretty simple.
My ultimate question is, what is the most efficient way to display the multiplication of the two point attributes in the viewport, one attribute coming from a particle, and the other the asset being instanced?
This is not an uncommon task. Even the Copy SOP has a built in feature to multiply attributes when copying assets. But, I have a lot of geometry to copy, and the Copy SOP isn't known for it's speed… So how would you go about doing this common task?
The most efficient way I can think to do this involves PolySoup and a Copy SOP. Sadly, this is still quite slow.
Is this something you could do with a OpenGL Shader? Or do we not have access to the scattered points attributes in OpenGL? If it IS possible, could I beg someone to show me a simple example?
My ultimate question is, what is the most efficient way to display the multiplication of the two point attributes in the viewport, one attribute coming from a particle, and the other the asset being instanced?
This is not an uncommon task. Even the Copy SOP has a built in feature to multiply attributes when copying assets. But, I have a lot of geometry to copy, and the Copy SOP isn't known for it's speed… So how would you go about doing this common task?
The most efficient way I can think to do this involves PolySoup and a Copy SOP. Sadly, this is still quite slow.
Is this something you could do with a OpenGL Shader? Or do we not have access to the scattered points attributes in OpenGL? If it IS possible, could I beg someone to show me a simple example?
Technical Discussion » Pass custom point or vertex attribute to GLSL shader
- Brian Walters
- 155 posts
- Offline
-
- Quick Links