Jump to content

bentway23

Members
  • Content count

    58
  • Donations

    0.00 CAD 
  • Joined

  • Last visited

Community Reputation

2 Neutral

About bentway23

  • Rank
    Peon

Personal Information

  • Name
    Mark Reynolds
  • Location
    NYC
  1. Fireflies in Mantra render direct reflection

    You are my favorite person of the day. The problem was indeed that those "hotspots" were actually reflections of the infinitely small spotlight, as I saw when I replaced it with an area light or moderate size. The solution--find a better HDR and kill all the other lights. Going to revisit the roughness map, but changing up the light did it. Thanks for your time!!
  2. Fireflies in Mantra render direct reflection

    There's no sampling quality option on the distant and spotlights--it's greyed out. A pixel sample boost was the first thing I tried, and didn't change anything. Also, perhaps worth noting--if all the non-key lights are turned off, and the key light exposure cranked waaay down to almost complete darkness, the fireflies are still there. This is the same wither rendering through PBR or ray tracing.
  3. I'm having a devil of a time trying to get rid of fireflies in a render (files and an .exr frame attached). I've isolated it to the direct reflection pass (the hotspots along the nose in the pic), and Nuke tells me they are way hot superbrights. Doing the usual Google fixes (working with diffuse limit, boosting reflection quality and overall render quality, &c.) haven't changed anything. In the scene I have three lights and surprisingly it's not the environment light that's causing the problem (I isolated the lights and also boosted the sampling quality on the environment light), but the key light (a spot light, but I tried distant and area lights with the same result). This is my first time dabbling with Substance Painter, which I used for a quick not-too-boring look, and it looks like the problem might be with the roughness or metallic maps. They're .tgas, and open fine in Photoshop. If I turn off all maps and boost the roughness the hotspots are less hot and spotty, but I don't think that's actually causing the problem. I'm sure this is something easy and stupid. Thanks for any assistance! (Credit where due: the original bust is from a photoscan by Ryan Baumann on Sketchfab.) caligulaFireflies.zip
  4. Aha! Actually, calculate varying IS the default (if using the "configure cloth" rather than just a bare vellum contraints node), but changing it to unchanged did the trick! Thanks!
  5. I'm tinkering with vellum and have a weird issue--when I use a plain grid as the cloth object and run it through a remesh, random points are/act as if they are pinned (they are given a @mass of 0), even though there is nothing set as pinned in the vellum constraint node. I've emptied the cache, restarted, even tried a new grid with slightly a different initial subdivision count, and the result is the same--various points stay in space, pinned (although with differing initial geo point counts, the points that don't move are different). A grid without remesh works fine. A planar patch actually works fine. I've tried various other objects, all run through remesh nodes (and without), and they all work fine. I could even assign those points a new mass through a wrangle, and that worked fine, but it seems like for something as basic as a remeshed grid, this shouldn't be happening. What am I doing wrong here? Or is this just why god and SideFX made planar patches? vellum_pinning_problem.hiplc
  6. I finally upgraded my Macbook to Catalina and now Houdini 18 isn't opening. It will think for quite a while and then give me a crash warning without ever completely spinning up. Also, the license server opens, but just hangs, with no successful connection. I have completely uninstalled and reinstalled, but that didn't help it. Any tips for getting H18 to work on Catalina?
  7. This is fantastic--thanks for the pointers!
  8. That gets the shape, but not the changes in velocity--what I'm looking for is the acceleration at the perigee and deceleration at the apogee. I tried basic trig early on, but couldn't translate that into varying velocities.
  9. What I'm looking for is an object to follow an elliptical curve, accelerating near the perigee and decelerating near the apogee. Theoretically I could wrestle with magnet forces and do a sim for this, but I feel that (1) that'd be unnecessarily complicated, and (2) that this is going to be a collider for another sim, and that might muck it up. I've tried follow path and using primuv, but I have not been able to find a way to accelerate and decelerate them as they follow the curve. I've tried something, attached, that is MOSTLY working. I get velocities from the tangent normals of the orbit curve, and a velocity multiplier based on the min and max x positions along the curve, so at the perigee the velocity multiplier is 1 (or whatever), and at the apogee it's 0.1. Using a solver SOP, each frame I transfer the curvevel attribute to the ball orbiting object at its current location, and then use @P += @curvevel. This actually works for the most part, except the orbiting object keeps going further and further out, creating a spiral orbit. Furthermore, even though I have a wrangle setting its initial position to be at point 0 of the path curve, starts farther away.(Also, I will need to figure out a way to orient it so it's always looking forward, but that will be easy enough, methinks.) I assume I'm missing something obvious within the solver SOP. This might be easy and dumb, so thanks for any help. (Also, if there's just a way better way to do this, I'm open to suggestions!) orbit_setup.hiplc
  10. Cool. In this case I didn't actually have a grid, but a very lumpy thing so there were few actual inline points. That's a good thing to keep in mind, though!
  11. Perfect. I sort the points, use a wrangle to delete all but the first and last, and I'm good to go. This requires it to be even with the grid(/surface) in y, but it's easy enough to throw a transform after to raise it up afterwards. Thanks!
  12. That's exactly what I want! What is happening with the intersection analysis and add?
  13. That's actually the reverse of what I want--I want the points to match the edges of the grid (and preferably not move in y). That moves the line in y but doesn't match the length to the point of the grid where it drops down.
  14. I know how to use that to, well, detect intersections, but I don't see any options on it useful here. Is it one of the output attributes?
  15. I have a line hovering above another object (in this example, it's just a grid, but it's more irregularly shaped in what I need). I would like it so that the line length matches the width of the object directly underneath, without reorienting the line, and preferably without actually snapping it to the new point (in y). Ideally, if I could extract a bbox D_XMIN (/MAX) (or a getbbox min.x), but only at the point in z where the line is, then I could just translate the points to that location. I'm sure this is super-simple, but I'm not quite sure how to go about it. A ray SOP gets me in the ballpark, but presents some problems, as in the pics. Thanks! lineray.hiplc
×