Jump to content
Hash, Inc. - Animation:Master

someawfulbridge

Forum Members
  • Posts

    103
  • Joined

  • Last visited

Profile Information

  • Location
    Nashville

Previous Fields

  • Hardware Platform
    Windows

someawfulbridge's Achievements

Apprentice

Apprentice (3/10)

0

Reputation

  1. So I just had a thought which I figure I’ll throw out there, as my own cloth/particles skills aren’t up to testing the idea yet—has anyone tried using cloth to serve as (pseudo)fluids in animations? I’m talking about a really “thin,” semi-transparent cloth, “pulled” over a surface (I imagine it’s outward “flow” would have to be manually adjusted/keyframed, like pulling a blanket, but perhaps with a slightly “tilted” set the cloth could slide down and calculate the flow/slide itself), where the surface itself would serve as the deflectors and the cloth contour itself accordingly. I’ve heard tell of new fluid simulations in v15, but, alas, my wallet mandates that I still play around in v13 for now. Plus, depending on the functionality of the actual fluid simulations, if my idea isn’t completely hare-brained it might allow for a more stylized “fluid.” Also, forgive the gratuitous use of quotes. It just seems necessary. (As Paul Rudd says in “Role Models”: “Why did you put quotes around ‘presence?’ Are you implying that we’re not really here?” In this case, yes.) Cheers-- Mark
  2. It keeps the group names and everything. Brilliant--thanks!
  3. I'm trying to line up two different components in a model very precisely, one row of "parts" to a panel that's at a bizarre angle. I can do my trigonometry, so I've never had trouble matching angles on one or two planes, but since this panel is rotated on all three planes it's proving surprisingly hard to match up--I've been doing inverse tangents 'til my eyes bleed, and it's just not working. A cursory look at some other forum posts had someone mentioning lining things up in an action, and that kinda worked--I made a bone for the panel, a bone for the parts, and in an action window used orient like/translate to constraints and the "parts" and panel were right where they were supposed to be. My question is this, then--how do I make that change reflected/permanent part of the model proper? It seems like creating a pose for each of the parts/panels that will have to line up will cause me to end up with what should be a simple model cluttered up with a bunch of relationships and bones that are never otherwise touched. Thanks for your help! Mark
  4. Alright--I'm way ignurnt here, so ignore me if that shows, but could object transparency be linked to an expression that gives it a (pseudo?)random value, so the flickering is somewhat automated but not obviously cyclical?
  5. You gents are gods among men. The simple fixes you mentioned did the trick. Here is the finished walk cycle. He's just walking in place right now. I have to fine-tune his stride length. Since he comes down hard on his feet and the back foot kicks out and up a bit, I'm wrestling with getting the proper stride length to account for that. But, in the meantime, I present Georges' latest movement: georgeswalk.mov Incidentally, his birthday is Monday. (Or would be, if he weren't all dead.) Happy birthday, Georges! Thanks again for y'all's help! Mark
  6. Allrighty--I've attached the project with the walk cycle-in-progress, and here is the really screwy thing that messes everything up. If you move one of the foot controls down in the y axis, even just one click, the whole foot control goes flying out and away--but this only happens on frame 20! Unhiding the knee controls, it looks like they're misbehaving--but just sometimes. I am able to do a clumsy and still unreliable workaround by hitting frames 19 and 21 and then keyframing the resulting interpolation on frame 20, but that shouldn't be necessary. Also, rendering the test choreography of him walking along a path doesn't render right. He walks correctly once, goose-steps about 4 - 4-1/2 times, and then slides around with a leg kicked out. I'm using a quad-processor Mac with whatever the latest OSX iteration is. It's a powerful computer, so it shouldn't be a memory issue. I recently upgraded from 13.0r to 13.0t, and I had none of these problems (with the same model, too) on 13.0r. Are these problems intrinsic to 13.0t? Is there a way I can find an old installer for 13.0r? Something fundamental is misbehaving, and I hope it's something easily fixed that I did wrong, but the errors seem too inconsistent for comfort. Any help will be extraordinarily appreciated. bataille.zip
  7. I was about to post it so y'all could see what was going on, if necessary, but it turns out A:M was all crazy and had to be completely reinstalled (which took all of 2 minutes). Now it's working, even with Quat interpolation, just beautifully. Apparently, OSX, in all its majesty, decided to randomly corrupt at least 4 of my commonly-used applications today. Crashes less than a PC my butt. Anyway, thanks for your help!
  8. I'm using an EXTREMELY simple home-brew rig (I'm a control freak and want to learn to make my own from scratch--of course this one needs no fan bones or anything special (including toes!)). I figured it out, I think. I changed the default interpolation from Quat to Vector and that seemed to fix it, based on a quick test. Euler worked as well. Hopefully that's it. Thanks!
  9. No--it's pushed in. That was the only possible solution I could find in other forum searches, but that wasn't it.
  10. I'm working on a walk cycle--I did one forever ago, but haven't touched on any actions or choreographies in awhile, so I don't recall having had this problem before . . . . I make my first pose, click off the model (so I'm not just copying that one bone), go to the end of the cycle, copy the keyframe. Then I go to the halfway mark and paste mirrored, which does absolutely nothing. OK, no sweat, my guy's a little assymetrical, so that's probably why, I just manually make a mirrored pose. But now, on frames 1 and 24 (the beginning and ends), the poses that were there have gone all wonky. One of the legs is now goose-stepping and the other is somewhere else crazy. Fortunately, I still have my previous attempt at copying the keyframe in memory, so a quick paste keyframe restores it to order. But now my halfway, mirrored pose has gone all screwy! What's going on here? I just want to make a keyframe, paste it, and move to the next, without the next screwing up the one I just did. For what it's worth, I'm using 13t.
  11. I CAN'T DELETE THIS, BUT CONSIDER IT SCRATCHED. Eventually, by arranging their order (in reverse order, too, I believe) in the model's shortcut PWS window, I beat them into submission. Urgh. I'm BAAACK-- So I'm obviously back in the world of poses. I remember all this from previously, and I can still delete poses and sliders left and right, but I can't reorder the sliders. If I reorder the poses under the model in the PWS, that doesn't translate to the order of the sliders in the Pose Sliders window--those seem stuck in the order in which they were created. How do I move them in that window? I tried re-dragging and dropping the poses' respective folders in the User Properties Relationships folder, so even their they're in the order I want them--the Pose Slider window is the only one I can't reorder. Can anyone help me out? Thanks!
  12. A Mac pays my rent, and I have to work with it 40 hours a week (and I'm typing on it now, so don't tell it I said this), but my god do I hate them. They're like Fischer-Price computers, and it sure is nice to go home and cozy up to Windows XP, rather than Mac OSX. I've also noticed A:M working more smoothly on my PC. For one thing, spring systems crash it on the Mac.
  13. AHA! Thanks everybody! And especially you, John! I looked at your file and realized the the key was leaving some portion of the blanket NOT cloth for the other parts to follow. (Perhaps this is a "duh" moment for most, but it was "eureka" for me.) I removed all bones as being superfluous, applied the cloth material to all the blanket except for the one corner CP by which it's "dragged," keyframed the model bone to move over 3 secs., which in turn moved the non-cloth CP, which in turn dragged the rest of the simulated cloth. This (successful!) test render is large (even at 320x240), being 3 secs. long. (If anyone has any tutorials on how to more effectively render tests for posting, point me there!) blanketdrag.mov And, if anyone is interested, here is the corrected project. blankettest.prj.zip Thanks again!
  14. I did, and just to be sure I also clicked the "clear all simulation data" (though, since that had no effect, I suspect that is exclusively for rigid bodies). I'm reattaching the beastie again, just in case. Also--even though I can't get it to move at all for starters, how would one go about dragging it, but just by one corner, rather than moving the whole thing? blankettest.prj.zip
  15. Your treatment did what I'm looking for--made the blanket move. I cleaned up the project, removed the bones and the "attached" group, and from what I can tell made my project mimic your changes, keyframing the model bone moving so that even in the quickie scrub preview the blanket mesh moves like I want it (minus the cloth/deflecting simulation). But still, whereas in your choreography the blanket follows the model bone down the stairs, my model bone moves while the blanket stays still after its first fall. Adjusting friction and mass have no effect. How did you get it to follow the model bone?
×
×
  • Create New...