Jump to content


  • Content Count

  • Joined

  • Last visited

Everything posted by Spritely

  1. I agree with TechCrunch: this needs to be in Photoshop NOW! BTW today TC reported that Adobe hired the dude. http://www.techcrunch.com/2007/08/28/adobe...zer-technology/
  2. I'm sure you've all seen this, but whatheheck. Best spec work ever, in this fool's mind. YouTube: http://www.youtube.com/watch?v=oWfBzuix_sY CgSociety article, HQ video requiring free membership: http://features.cgsociety.org/story_custom.php?story_id=3302 I also want to ask this board to remind me what is the name of the AE plug-in that does this. - A. A.
  3. <nitpicking> if all you care about is whether c>0 or not, there really is no point in dividing by length(. Just slows things down. </nitpicking> But cool stuff :-)
  4. You're comparing the position[1] value without multiplying it by time*time. This value refers to the original y coordinate, which stays put, so your expression always evaluates TRUE. Try: yaccel = thisComp.layer("Circle").position[1]*time*time*12; ystop = thisComp.layer("GroundPlane").position[1] if (yaccel < ystop) [100, yaccel, 3000] else [100,ystop, 3000] (Warning - I'm not in front of AE right now so can't test this, but I think this should work better. It's also better to have an else part to explicity say what happens if the condition is NOT satisfied).
  5. One of the cleanest I've seen recently. I love it.
  6. I wouldn't take it so far. I was "that crowd" for a while, and fairly quickly I learned from the buzz (even before trying anything myself) that fluid simulations should only be used as a very last resort, owing to fantastic render times, impossibly long tweak cycles, and software cost. Most cgtalkers I know would work long and hard with particle blobs before they give up and take the realflow route. Here's one of my favorite tutorials for simulating water surface (no pouring). It's XSI but should be fairly portable as it uses not much more than standard perlin noise, specularity control and displacement. - AA
  7. If you must do it in AE, you'll save yourself a lot of keyframes and trouble if you set up proper hierarchies for the characters, possibly with some expressions to hold things in place. South Park characters may not need a lot of inverse kinematics if I remember correctly, but anyway Dan Ebberts has some useful expressions here. If lip-synching is to be involved, here's a Creative Cow tutorial. Three minutes sounds to me like quite a lot. AE really isn't the proper platform for this kind of stuff, but frankly you don't need Maya for that either. Look at Moho and Toon Boom, they're relatively easy to pick up and they're really meant for such animations. AA
  8. Again, I'm not at all sure, I don't have the mograph module yet... I remember seeing something about geometry masking but it may be irrelevant.
  9. If C4D can generate 2D geometry from image textures I'd love to hear about it, but as far as I know it cannot do so directly. The simplest method for your situation, it seems to me, would be to use the animated alpha texture to drive a displacement map, or use a relief object with the texture as the height map. Some pre-blurring would probably be necessary. See, for instance: http://www.creativemac.com/2004/02_feb/tut...lace0402244.htm AA EDIT: if you start with a displaced plane, you can then also re-apply the alpha texture to the displaced plane so that the "unextruded" regions vanish. It may look better if you use a slightly expanded (negatively choked) version of the alphas for the mask. Also, as an alternative, the mograph module has some gemoetry masking options which could be relevant here but I'm not sure...
  10. I was craving for this for such a long time - you da man!! THANKS!
  11. I precomp ALL the time, and like Sao said, the purpose is not just to organize your layers. That said, I find it very helpful in keeping organized, I've no idea what layer sets are but I can't say I feel I'm missing something big in that respect. My only gripe is that when precomping a bunch of layers that do not span the whole timeline, the precomp ignores that and creates a composition across the full timeline. There's a script somewhere that overcomes this but it's not really convenient to reach for the script menu every time I need to precompose. I regard this as a silly oversight on the programmer's side and I hope it gets fixed... BTW nice moniker you got there, spritelyjim
  12. Have you searched TurboSquid? Good models are not cheap though.
  13. Beautiful stuff. One of the exciting things about this industry is the speed in which inventions migrate from the research bench to the market - hopefully we'll get to see some of those algorithms implemented in an accessible app before long. I'm curious how long those movies took to simulate and render...
  14. What is it you don't get? What have you tried? Where do you get stuck? It's nicer to provide some more details and explain what you did than just state "I don't get it, need more help". Honestly you may find this is not the right forum for such discussions (the Cow works better for most people). I for one will be glad to help but I just need to understand what the problem is.
  15. Spritely

    Psyop AT&T

    This is one damn handy technique. Thanks for sharing, Harry.
  16. Agreed, but I still think this opens up some interesting creative possibilities - especially if it's fast. I don't think anyone will actually use it to simulate caustics.
  17. This was posted on the Cow yesterday, looks like an interesting plug-in. I haven't tried it yet.
  18. Not in motion? Could have fooled me. Those are so dynamic they're just alive.
  19. There's a bunch of dedicated halftone plugins (search), but here's a quick fake: Apply Card Dance to a gray solid, with say 100 rows and 150 columns (depending on the proportion of your comp). Let the intensity of your source image drive the X scale (-0.5 mult), Y scale (-0.5 mult) and Z position (0.5 mult), and use a point source light in the effect settings with an intensity of 2.0 and positive depth. This'll break down your solid into small rectangular dots, which get larger as your source image is brighter, and they also get brighter on account of the Z positions and light. The only problem is that the dots are not circular, as they should be. One easy way around this is to start with a static image of an array of small dots, prepared in PS or AI. Make sure the array has the same number of rows and columns as your card dance, and you're done. (EDIT: of course the whole point of half tone is that the dots all have the same brightness, but you can get fancier results this way. Experiment...)
  20. Of course you can animate a 2-point open mask and 3d-stroke it, I guess you know that. You can also manually animate the stroke characteristics to have it thin out as the points drift apart, for instance. What you can't do, to the best of my knowledge, is: * READ the position of the mask point and use it in an expression, for example to have automatic control over the stroke's parameters. * WRITE the position of the mask point so that you can control where it goes in a more precise manner than just moving it about and keyframing the mask shape. This is very unfortunate, but as far as I know it can't be done, with or without scripting. There's a workaround of sorts for the READ part, using a separate tiny stroke and motion tracking, but it's VERY cumbersome. In some scenarios it could help you, though - you should be able to parent a null to your animated mask point and then use this null to drive whatever. EDIT: Wow Rothermel, interesting script. Thanks!
  21. I caught a glimpse of it earlier this morning and the really liked the new look. The press release seems to have vanished though - is it just me?
  22. OK people. Since I'm coming out here as the blood thirsty, torture-lovin', bunny-hatin' sadist, I thought I'd make one last chance to clarify my point of view. I think there's currently way more torture and abuse going on than is justified. I'm a long-time supporter of Amnesty, for those and other reasons. But to effectively defend this position, I believe the worst kind of argument is the naive "let's just be nice and humanitarian and everything will be handy dandy". The fact of the matter is that in our far-from-perfect world, there are scenarios were the morally justified position is to inflict some level of uncomfort on certain individuals. Dozens of suicide bombings were thwarted just before they occurred using information gathered in this way - I'm so sorry, but those are facts, not my wishful thinking. The goal should be to minimize those events to an absolute minimum, and then minimize them some more just to be on the safe side. But if you want to argue that position, you're not well served by the absolutist, self-righteous position. If you really want to counter over-zealous administrations, make a reasonable case, not something that sounds like red riding hood meets peter pan. It's better to set clear guidelines that to chuck it all off as irrelevant. What, exactly, counts as "torture"? I've asked this before but no-one really seems to want to think about that. Is keeping a person awake for 24 hours "torture"? That's "sleep deprivation", sure, but most designers I know have inflicted that torture upon themselves willingly. How about shaking a person's body? Yelling at them for an hour? If you stand firm and disallow ANYTHING, I'm sorry, nobody's going to take you seriously, and you'll be harming the very cause you're out to advance. Having said all that, I repeat: More should and can be done to minimize prisoner abuse. Over and out.
  23. Who's "we"? You, I presume, haven't. I have. I'm not happy about it, but I have. And, btw, I'm talking about the milder forms of coercion, not the kind of "torture" that comes to mind when you think of WW II prisoner camps. As this is the second time I'm being branded "stupid" here, I think I quit. Next time, try to argue with the position, not the person.
  24. Thanks for the kind words. Why are those hypotheticals bogus? You mean, it never happens that a person is captured who may have direct information on a crime that's about to be commited? I beg to differ. This has actully happened in England not so long ago, and it happens elsewhere pretty frequently. If torture isn't an effective tool, I'm glad to make it totally illegal. But its sounds strange to me that the best way to get information out of a fanatic of any kind is to ask them politely for it, without making them the least bit uncomfortable. Is it at least OK to yell at them? Scare them? Keep them up for 2 hours past their bedtime? That may well be true, I don't know one way or another. But what do you suggest be done with the minority who aren't just collateral? That's the question here, I believe, not the overzealous campaign.
  • Create New...