Jump to content

Doktor Gonzo

Members
  • Posts

    529
  • Joined

  • Last visited

Everything posted by Doktor Gonzo

  1. Fix was actually a lot simpler than what I originally had in mind: Each of your 2 leg IK handles has an attribute called "Twist". If your handles are anything like mine (set up for battroid jointing, connected at the knee to a Locator via a Pole Vector constraint) then a value of either 120 or -120 on Twist ought to convert the forward-jointed battroid configuration leg into a reverse-jointed GERWALK IK joint. Just set-drive Twist to have those values when you're in gerwalk, 0 when you're in Battroid, and you should be there...
  2. Rod, Oddly enough, I have gotten really good use out of the ART OF MAYA books - though they are mainly art galleries, they have loads of sidebars and margin text offering in-depth exploration of the tools and techniques used to produce each piece. If you need something more tutorial-y though, the LEARNING MAYA series is excellent. Not sure if LEARNING MAYA 6 is out though....
  3. KN, Your set-up pretty much EXACTLY mirrors my own - set driven transform, FK arms and IK legs (I hit the same "backwards leg IK snag" and gave up on getting IK to work for GERWALK, I just have a script driven IK/FK switch that activates the IK constraints/handles only when transformation to Battroid is complete. For GERWALK leg animation, I use the FK handles.) My transform works in arbitrary time as it is set driven key driven. The transform slider is attached to a master control object which is also the master "mover" control for the whole character. The arm and leg joint chains are attached to the body, just like yours. The body has no skeltal structure, it is just insanely complex parenting and properly placed pivot points (though I have FK-driven waist pivot and some limited "tilt", so it can lean or hunch a little). As for the legs, all I can tell you is that scripting is your friend. Placing the pivots properly, locking axes on hinge and rotator joints, and setting limits will also pay off. And remind me to share with you the crazy math I've got going for proper animation of the intakes and the kneepads soon... But I have to tell you -- I just got an idea while typing this, one that might solve that GERWALK IK leg problem. Let me try it, I'll get back to you later....
  4. Got a pretty crazy Frankenrig going in Maya 6 for my Valk - rigged for transform, battroid-mode IK, and various other bells and whistles. I'll try to get some samples up here soon. Very nice job on Dat's Valkyrie, by the way.
  5. Chan Fan, Looks great! In answer to your question, I, poor crazed soul, did try to go with a single, hideously complicated rig that handles all 3 modes. it's a partial skeleton - essentially joint chains for each limb, without any joints in the body (too much reconfiguration during transformation for that to work, and besides, there really aren't any flexible areas in the torso that would benefit from it). The limb joints have IK solvers driven by scripting to only activate in battroid mode. That way, I can for example have legs with properly bending knees for the battroid that are nevertheless animatable in their reverse-jointed GERWALK configuration. My transformation sequence is driven by a slider control, so that I can just set keys on that to transform and not worry about redoing the whole animation. This all sounds great and, on the whole, works great. But it was insanely complicated (I mainly did it as a learning exercise) and necessitated a great many compromises in the battroid-mode character rig. And it still wound up not working as an all-in-one solution - I have lately gained occasional access to a motion capture stage and wanted to use my Valkyrie with data from that, but could not make it work with the extant rig - so I wound up building a separate battroid-only motion capture-friendly rig. If I had it to do over again, I would probably do it as you suggest. Tweaking of parts scale in different modes, though, is something I consciously avoided - Aztek, myself, and some of the other guys on these boards had decided at the outset that we wanted to attempt to model "real" Valks, with no "anime magic" involved in their mode changes. Nevertheless, I love what you're doing, and will be interested to see how your swap-out works. My guess, though, would be that the toon shader will make the swap harder, not easier - everything looks so crisp, and there are sharp demarcations between color areas - that'll make any sudden "pops" stand out. Whereas with a realistic render, shadow, blur and fuzziness hide a multitude of sins
  6. Nope. They stay at their usual constant length.
  7. Manu, Good analysis! That behavior is intentional - I want the code inert save for the first time the controlling variable hits/departs from the value of 10 - otherwise it'll throw an error, and I'm not adroit enough with Maya's try/catch scripting to work past that yet. And the flag is, as you note, unique to this bit of scripting and is not modified elsewhere. As for what it does, it allows the knee part to be driven by animation keys linked to the transformation slider during transformation, then decouples it from those and attaches it to a computed value determined by the pose of the thigh and lower leg once transformation is complete and character animation has commenced. Oh, and sorry for the random digression into geekery, all.
  8. OKay, for the TRUE Maya geeks out there - tell me what this code snippet is doing, and win the eternal admiration of your peers: if (VF_1_Master_Control.BATTROID==10 && l_knee.keyableFlag==1) { disconnectAttr blendWeighted263.output l_knee.rotateX; connectAttr joint3.Kneepad_rotate l_knee.rotateX; l_knee.keyableFlag=0; } else if (VF_1_Master_Control.BATTROID<10 && l_knee.keyableFlag==0) { disconnectAttr joint3.Kneepad_rotate l_knee.rotateX; connectAttr blendWeighted263.output l_knee.rotateX; l_knee.keyableFlag=1; } -DG, going gaga with midnight programming...
  9. Gandcroft was pretty damn scary...
  10. KingNor, On the right track, but probably not quite the right tool for the job. Blend shapes are used for complex nonuniform deformations - i.e. where the constituent vertices of the object move nonproportionally in relation to each other. Their other big advantage is that they are additive, meaning you can establish multiple blend shape deformers for a single object and appply them combinatorially. These two cool traits are much of the reason why these are so useful for facial animation: faces twist and warp while moving, and it is useful to be able to, say, apply a smile deformer in combination with, say, an open-mouth deformer to arrive at, say, a grimace of pain. Neither of these conditions really apply to valkyrie parts, though: the "animagic" of the transformation seems limited to a proportionate scaling of certain parts, and there is generally only one possible transformation to which they are subject. As such, they really don't necessitate the time and cumbersomeness that blend shapes entail. I'd just set keys on the X/Y/Z Scale attributes instead.
  11. Hi Assaultor! Are you the Sean Soh who did the VF-19 model posted at CG Talk? If so, beautiful work!
  12. Ditto. B5 season sets for $9.95 apiece??!? I MUST know more!
  13. Personally, I vote for a TROOPS Star Wars series.....
  14. You know, I watched the TiVo'd pilot last night with the family, and that is EXACTLY what I said to 'em. Let's cross our fingers that the "otherworldly" element here, when revealed, is reasonably compelling...
  15. Do me a favor, elaborate on this please! I own the MR Anakin EL and am pretty satisfied - it looks cool, though I wish the durability was better (I've already had to do more than one bit of post-duel repair). What exactly makes Mr. Parks' work so superior? Durability is wonderful and I crave it, and the detachable blade is really cool, but are there other differences? And how much do you Parks owners miss the sound chip from the MR blades? To me, it's a major part of the experience! Just looking for some guidance, I am determined to get a Vader saber and really want the best one possible...
  16. What, you mean THIS??!? What can I say, lads? There's only one way to get to the truth here. And that's to enter the belly of the beast and see for myself. That theater is around the corner from where I work. Come next month, I think I just might pony up my $20 and view this trainwreck firsthand. I'll report back....if I survive....
  17. i see little random bit popping out of that, was it superimposed over an existing craft? Good eye! Yeah, this was a clip of an F-18 taxiing, I garbage-matted it out before rendering my own craft in. Obviously, for a "real" shot rather than a quick test, I'll shoot my own plate, but for this I just grabbed and modified something I found on the internet - and unfortunately nobody bothers to shoot and put up pans across an empty tarmac!
  18. I'd have to say, "sorta." In earlier editions, the unrotoscoped stick-topped-by-a-lightbulb was plainly visible. For this one, it looks like somebody took the frames in question into Photoshop and airbrushed a translucent blue glow over them - so there's an effect of some kind there now, but you can still see the ol' stick-topped-by-a-lightbulb through it. Ah well, it's still something of an improvement I guess. Just odd, given that many of the other saber shots have been completely rerotoscoped to look more like the Prequel saber effects (a pity, I kinda dug the wider, brighter, cartoonier-lookin' blades from ANH...) One would have thought they would've placed an actual blade element over the gaffe..... Oh, and on the dialogue issue, with dejr8bud and others saying the voices now sound different.... my guess is this: we're hearing these old audio tracks in much higher fidelity than ever before.... I think the "difference" that we're hearing (and it's clearest I think in Tarkin's dialogue.... his voice sometimes changes from "clear" to "muffled" in mid-sentence) is the difference in sound quality between dialogue recorded "live" on the set and dialogue that was ADR'd in postproduction. It's probably always been there, the mix just wasn't clear enough to hear it before. Now, I personally believe they could have gotten the different clips to blend much more seamlessly had they gone back and done a thorough digital cleanup and restoration of the audio masters..... let me guess, that's one more enhancement being withheld for the HD-DVD release....
  19. Dat, Forget the cameo.... we'll get you a co-starring role! Remember, the poster we're slowly slouching towards on this thread is allegedly only a warm-up.... the final goal is a collaborative CG Macross short!
  20. Hey gang, Sorry to be essentially absent lately - been sidelined by real-life concerns. Ought to be able to pick things up soon. Meanwhile, here's a brief test clip to show I haven't been completely idle. It's not much by itself, but it's proof-of-concept for some more involved things I hope to get started on soon: Taxiing for Takeoff
  21. Kornholio: nice!, and quite differently styled from the already extant valk models in this thread - so it certainly gets my vote! One quick improvement you could make is to the hips - the legs need to be a little more widely spaced, so they don't clip through the nose when the valk assumes that stance. They don't connect directly to the nose anyway - there should be some kind of a connecting strut, could be just a cylinder or whatever. Just making this small change would help a lot with the apparent proportioning of the battroid...
×
×
  • Create New...