Found 310 posts.
Search results Show results as topic list.
Technical Discussion » Vellum Wrinkle Tips
- Midasssilver
- 310 posts
- Offline
Anyone have any tips. I have a shirt that Im simming, and I'd like to be able to force some wrinkles in it. As far as I know, using a high compression stiffness is what is supposed to achieve that, but Im not making much progress by changing that parameter. Any tips would be appreciated.
Technical Discussion » Cloth Self Collision Interpenetration issue
- Midasssilver
- 310 posts
- Offline
I have a dress sim I am working on, and the model of the dress has a part folded up and "pinned" at the bottom, so its almost overlapping the other side of the dress. The issue Im having is, as soon as my sim starts, the folds that are so close together interpenetrate. Im not certain if I should use using a one sided model, or a way to work around these fine folds that need to be simmed. Ultimately, Im getting failed self collisions, and increasing the thickness isn't making a difference.
Anyone have any ideas?
Thanks!
Sorry no HIP I can share
Anyone have any ideas?
Thanks!
Sorry no HIP I can share
Technical Discussion » Dampen Collision Velocity
- Midasssilver
- 310 posts
- Offline
Here's a hip file where I tested with collisions, without collisions, and with collision vel. Even with my collision vel set to 0, we still get full velocity from the collisions. If I understand collisionvel correctly, it only happens when the fluid contacts the collision geometry. Since my collision geo is inside the source, thats my entire sim. We need collisions in order to have a collision vel, so how can I dampen the amount of velocity we are receiving from collisions?
Technical Discussion » Extract rotations for an alembic as a channel
- Midasssilver
- 310 posts
- Offline
Okay, after breaking down what's happening, I believe what I'm actually seeing is too much collision velocity. The collisions with the smoke is what is causing the look of inherited velocity, and without collisions, it isn't a problem. I believe that's why my workaround was dampening the movement of the alembic so we didn't get as much "collision velocity" for lack of a better term. I've attached a hip file where I'm able to scale the translation of the alembic, sim, then add that translation back. Ultimately what I need to do next, if possible, is scale the rotations as well, so that they remain in the right location/axis. The head turning alembic example I gave you is a good example of how, despite having essentially no translation, we still get the head twisting in each direction, pushing smoke further than I'd like. If I could get a channel for that rotation, scale the rotation down, then re-apply it, I'd be able to limit the "collision velocity" of any alembic I receive. Any help would be appreciated!
Technical Discussion » Extract rotations for an alembic as a channel
- Midasssilver
- 310 posts
- Offline
There's a type of inherited velocity from the movement of geometry, that can't be scaled by simply multiplying the velocity on the geometry before it goes into a sim. Ultimately what I want to do is be able to read the rotations off an alembic, scale them down, then re-apply them (post sim). That way, if my alembic is a head turning that turns very fast, I can dampen the amplitude of rotation, sim, and then apply the rotation back after the sim.
I was already able to grab the pivot for the intrinsic attributes, scale that down in chops, and reapply it. That took care of the translations. The rotations are more complex though.
I was already able to grab the pivot for the intrinsic attributes, scale that down in chops, and reapply it. That took care of the translations. The rotations are more complex though.
Technical Discussion » Extract rotations for an alembic as a channel
- Midasssilver
- 310 posts
- Offline
I've seen some posts on the forum regarding this, but I can't seem to get anything working regarding this.
I have an alembic, and I want to be able to extract the intrinsic rotations on the alembic, scale them down, then use that new channel to apply the scaled rotations to the original alembic. This should be able to work with translations as well.
Ultimately, Im trying to reduce the inherited velocity from an alembic, so reducing the movement of the alembic itself is my only option at the moment. If anyone can provide a sample file, or have advice, it would be much appreciated!
I have an alembic, and I want to be able to extract the intrinsic rotations on the alembic, scale them down, then use that new channel to apply the scaled rotations to the original alembic. This should be able to work with translations as well.
Ultimately, Im trying to reduce the inherited velocity from an alembic, so reducing the movement of the alembic itself is my only option at the moment. If anyone can provide a sample file, or have advice, it would be much appreciated!
Technical Discussion » Basic Liquid Rendering Issue - Flickering specular
- Midasssilver
- 310 posts
- Offline
Im working on an animation, and Im using a basic liquid shader to shade my ice. Im getting a bright specular highlight that looks like a firefly from a sun light in my scene, and I can't find the right combination of render settings to eliminate it. Any render settings or shader tips that might help me resolve the problem?
Image Not Found
Image Not Found
Technical Discussion » Smoke Cannot be seen in Simulation!?
- Midasssilver
- 310 posts
- Offline
I took a quick look at your scene. Honestly, everything was too slow to work with. Simplify it and see if the issues still persist. If you re-post a simplified scene, Ill check it out again.
Technical Discussion » Deforming Geometry + Glue Network
- Midasssilver
- 310 posts
- Offline
My last post with an included hip got lost in the nethers, so I wanted to see if anyone had a way they usually handle keeping pieces together after they become active after being deforming. It's very easy to just delete the glue constraints of active pieces, so they switch from deforming to active, but how would one keep those newly active pieces together, until the desired time to break them apart?
Technical Discussion » Deforming Geometry + Glue Network
- Midasssilver
- 310 posts
- Offline
I set up a simple scene, where I can't seem to get the glue network and the deforming geometry to behave together. Any idea why my glue constraints get so screwy when used with deforming geometry?
Technical Discussion » UDIM Scale Help!
- Midasssilver
- 310 posts
- Offline
Thanks. I’m currently filling 4 UDIMs with other faces, but I was told to put all my interior face uvs in one. Looks like I’m not doing anything wrong though. Thanks again
Technical Discussion » UDIM Scale Help!
- Midasssilver
- 310 posts
- Offline
Hello! I am struggling with an issue, where if I have a simple boolean fractured geometry, and I uv flatten it, I get my uvs in UDIM 1001, which is good, however at this point, the uvs are too small, and we can't see the resolution of the checkerboard from uv quickshade. If I scale up my UVs with a UVTransform by 10x, my UVs look great, but they are no longer in 1001. If I use a UV Layout again, then they are placed into UDIM 1001, but they are scaled down again and don't look great. I noticed on simpler geometry, with less fractures, there is enough space to fit all the UVs for the interior edges in 1 udim, and then I dont have any resolution issues, but what I am supposed to do when I have tons of geometry? Any help would be much appreciated!!
Technical Discussion » velocity loss in alembic export to maya
- Midasssilver
- 310 posts
- Offline
We noticed that velocity isnt getting interpreted correctly in vray when we exported our alembics. Turns out it depends on what we are exporting. With a shattered geometry, exporting the velocity as a vertex attribute v worked. With geometry copied onto points, we need v to be a point attribute for it work properly. Recording it here for anyone who experiences the same thing.
Technical Discussion » Crowd Transition Help
- Midasssilver
- 310 posts
- Offline
There appears to be a 1 frame difference between fbx loaded on disk, versus character rigs loaded from inside Houdini. The character rigs are more accurate, but also weigh the scene down. Try testing those two clips using character rigs in scene.
Technical Discussion » Crowd Transition Help
- Midasssilver
- 310 posts
- Offline
I have am testing the transition between 2 clips in a crowd sim. I only have those 2 clips enabled, and I can see that they have a connection in the Transition Graph. The problem occurs when I queue the transition on a specific frame. Houdini is telling me that the frame Clip B is similar to Clip A is on frame 435, but I know that the frames don't match up until 438, and I can't work around it. If I try and set up a custom transition at 438, the transition is just skipped. Anyone know a way to either force the transition to happen at the appropriate frame, or tell my why Houdini is interpreting my clip incorrectly. Thanks!
Technical Discussion » Stadium Agent Loop Help
- Midasssilver
- 310 posts
- Offline
SOLVED: Turns out the task was a lot more difficult than I initially expected. The bright side is that I was able to use the Test Transition DOP included in the Stadium Example to extract the code I needed to accomplish my task. That has a pre-built randomize transition setup. If anyone struggles with the same thing, that example file is the key.
Technical Discussion » Stadium Agent Loop Help
- Midasssilver
- 310 posts
- Offline
I have a crowd sim where I am transitioning from a sit state which contains multiple clips to the same sit state. I have variance in the clip selection, which works after the first transition, but after that, the clip transitions back into the first clip, as opposed to any other clip with a matching pose. Then the agent proceeds to continue this loop of switching between the same two clips, instead of choosing any new clips.
I saw other posts regarding the same thing, but with no response, so I was hoping to try again in case some gurus know how to easily break this loop.
Also, as an example: In the Stadium Crowd Shelf Tool, if you disable the Sit To Stand Transition, you will see that all the agents continue the same loop since they have no transitions queued. I would want it so that when they get to the end of a clip, it randomly switches to another motion with the same starting position. I tried by adding my own transition node with a sit state to sit state transition, and as I mentioned it did work, but then looped.
Grateful for any help!
Thanks
I saw other posts regarding the same thing, but with no response, so I was hoping to try again in case some gurus know how to easily break this loop.
Also, as an example: In the Stadium Crowd Shelf Tool, if you disable the Sit To Stand Transition, you will see that all the agents continue the same loop since they have no transitions queued. I would want it so that when they get to the end of a clip, it randomly switches to another motion with the same starting position. I tried by adding my own transition node with a sit state to sit state transition, and as I mentioned it did work, but then looped.
Grateful for any help!
Thanks
Technical Discussion » Crowd texture visualization bug
- Midasssilver
- 310 posts
- Offline
SOLVED: Having a condition in the material stylesheet broke the texture visualization in the viewport. Adding a material sop to the agents with a shader on it seems to have fixed the stylesheet bug, as now the stylesheet textures are correctly persisting throughout the timeline in the viewport.
Technical Discussion » Crowd Blend Duration Variance
- Midasssilver
- 310 posts
- Offline
Update: Upon further examination, I am successfully getting the attribute into dops, but it doesn't look like it is iterating through all my agents. It looks like it is just taking whatever value I have for (most likely) point 0. I am using the expression:
point(“../crowd_setup”, @ptnum, “duration”, 0)
I am calling the path to that node, I am iterating through all the points for the attribute “duration” and choosing the first index, which is a float. Not certain why it isn't looping through my agents and retrieving the duration value. Any ideas?? Thanks!
point(“../crowd_setup”, @ptnum, “duration”, 0)
I am calling the path to that node, I am iterating through all the points for the attribute “duration” and choosing the first index, which is a float. Not certain why it isn't looping through my agents and retrieving the duration value. Any ideas?? Thanks!
Technical Discussion » Crowd texture visualization bug
- Midasssilver
- 310 posts
- Offline
Note, this was actually not a workaround, or a solution. Caching the simulation did not resolve the bug. It is still a viewport visualization bug, regardless of whether the sim was cached or not. It does not effect the render, only playblasts.
-
- Quick Links