Jump to content


  • Content count

  • Donations

    0.00 CAD 
  • Joined

  • Last visited

Community Reputation

1 Neutral

About toothpaste

  • Rank

Personal Information

  • Name
    Edge Flow
  • Location
  1. [ MANTRA ] Strange striped noise in render

    Thanks for your help TobyGaines! I've followed your advice and achieved much cleaner renders!
  2. I'm getting thick striped noise patterns in my renders in mantra and I am at a loss as to why! This is what it looks like: This is in a clean scene rendering with even higher samples, you can still clearly see the striped noise pattern: I've included the test scene here: noisyCircle.hiplc Please let me know why this is happening so I can get some sleep! Thank you!
  3. I booted up Houdini this morning to do some more learning and when I put down an attribute blur node this error came up: So I dived into the node to see where it was occurring and this is what I found: I tried to problem solve it on my own by doing a Google search and this is the solution I came upon: https://www.sidefx.com/forum/topic/59400/?page=1#post-266007 I then followed the steps and added "HOUDINI_OCL_DEVICETYPE=GPU" to my houdini.env. Saved and restarted my computer. But it has not helped in my case as my attribute blur node is still giving me the same error. I then dropped in an OpenCL node and this is the error from it: This is where I am stuck now. This is the details of my Houdini openCL: And of my computer: Is my computer no longer able to work with openCL? I'm not technically sound to problem solve this on my own so I'm here to seek help! Thanks a bunch, Toothpaste
  4. Crowd Transitions not random

    I have 3 animation clips for my crowd and I've built transition graphs for each of them to each other: 1 -> 2 1 -> 3 2 -> 1 2 -> 3 3 -> 1 3 -> 2 Then I merged all of these graphs into one merge node and sourced it into my crowd object's transition graph. All states are created in the dop net. I've built out all the transitions in my dopnet and the crowd trigger is set to Time > Frames > Playback Time : 1 > Random Offset : 10. Everything's plugged in correctly into the crowd solver, and with these settings I assume that when I hit play my crowd agents will randomly go between their initial state and another random state. But the actual result is them going between clip 1 -> 2 and then begin to loop. I can prep a hip file tomorrow when I get home from work, but in the mean time if anyone has any idea why the result is not as I expected ( agents adopt random animation clips ), please let me know! Brush everyday!
  5. A question about crowd transitions

    Gave it some thought through out the day and it makes perfect sense now. Thanks Skybar!
  6. Vellum SOP - update pin group over time

    Cool! That's a real smart way to get the same result without coding! 8) Thanks for showing me another way to do it!
  7. A question about crowd transitions

    I've got a few animation clips ( clips 1 - 5 ) appended to my agent. Then I built transition graphics from clip 1 to all other 4 clips, and I see a little spider web formed while viewing the transition graph ( which to me means they can now freely travel from each other ). I then set up everything in the dopnet and randomized the initial clips on my source, plugged in the crowd transition graph into the crowd object and hit play, ready to view all my animations randomly transition between each other. But all the solver did was to loop each clip's initial state, and only the one crowd agent who started on clip 1 ran through all 5 clips ( I assume this is because I've actually set up all the possible transitions for clip 1). Which leads to my assumption that I need to build a transition graph for EACH clip to EACH OTHER clip for what I want to achieve to work ( crowd agents randomly transition between clips )? I might not have worded this question well, but I don't know how to make it more concise. Thanks!
  8. Vellum SOP - update pin group over time

    Thanks Tomas! That solution worked as I expected! To clarify what this piece of vex code does: inpointgroup(0, "pin", @ptnum) : return True or False ( 1/0 ) based on if the current pt in input [0] is in group 'pin' then assigning the value to the stopped attribute on each point in the vellum system if stopped = 1, hard constraint is maintained if stopped = 0, hard constraint is turned off PS. I really like this solution because it allows me to animate the bounding object back over the unpinned points to constraint them again in the same vellum system. Thanks!
  9. Vellum SOP - update pin group over time

    Thanks for your help Noobini! I also want to do a snapping off animation ( someone pulling on the shower curtain to snap off the points one by one / release from the 'pin' group ), and I am animating the bounding object to try to achieve that, while it is updating in SOPs ( I see the points entering and leaving the group in the spreadsheet ) it is not transferring the updates into the vellum solver, could you help me solve this problem as well please? Thanks!
  10. Making a shower curtain vellum SOP set up, having trouble updating the pin constraint group, need help. What I have done: 1. Group points to 'pin' group using animated bounding object to simulate curtain being pulled and snapping from the curtain rod. 2. Pipe cloth into vellum system, using the 'pin' group in a vellum constraint set too pin. 3. Here's where I need help, I don't know how/where to use a sop solver to update the 'pin' group frame to frame. Does this solution differ when using vellum in DOPs? 4. Is there an '@active'-like attribute in vellum SOPs I can use to tell them when to enter and exit simulation? I've attached a quick example file, please take a peek if you can. Thanks! vellum_update_pinConstraint_HELP.hipnc
  11. Mantra Renderer Problem

    I've found the reason, a bad alembic cache. /end
  12. Mantra Renderer Problem

    I'm trying to render out a sequence using Mantra's 'Render to Disk in Background'. When I try to render the entire sequence it will complete the job in 5 seconds without outputting any images. But when I try to render the current frame it will complete the job in 12 minutes and outputs the correct image. I don't understand why this is happening and would like to get some advice. Thanks!
  13. I'm having some trouble understanding how to set up activation frame on the convex proxy shelf tool. I tried placing it beneath assemble it generates and made sure it is passed through to the dop. But it only results in frozen sims however I tried to do it. Can someone shed some light on this please? Thanks!
  14. The Twist and Bend deformers in Houdini 17.5.173 seems to be exactly the same? Is it true for your as well or is something wrong with my Houdini? Or are there some subtle differences I'm not aware of?
  15. Support has gotten back to me. Turns out my render licenses got mysteriously taken away. They sent over an updated license file and all was good! For future reference if anyone encounters this issue.