Jump to content
Hash, Inc. Forums

strohbehn

*A:M User*
  • Posts

    542
  • Joined

  • Last visited

Everything posted by strohbehn

  1. Great! And, out of curiosity again, what method of payment interests you most... cash, models, rigging, or etc.?
  2. Looks like fun, Zach. How long did it take you to camera-match this scene, just out of curiousity? Also, would you be interested in doing some camera-match-for-pay (or trade... i.e. models, rigging, etc.) work at some point? Some contract work would help you recoup some of your costs. I don't currently have any projects in mind, but am very interested in this type of work and don't have much interest in learning yet another piece of software (not to mention shelling out $399).
  3. I'll second your comments of thanks to all those who post helpful hints, rigs, project files, plugins, models, etc. on this forum. Without this resource, I would have quit long ago. Glad you found the info useful.
  4. Using version 6a, I'm still getting the odd finger and arm movement. The test action file I'm including has only 5 keyframes (on frames 0-4). Only the hand_left_spread/clench slider was moved. It's keyed at 0% on frame 0, then 20% on frame 1, then left untouched on frames 2-4. But you will see that the fingers continue to curl on frames 2-4 even though the slider value is unchanged. For this action, the "key constraints" filter was ON when I forced the keyframes. I did the same test using "key constraints" OFF, and the model behaved as expected... no finger movement on frames 2-4, so it has something to do with including constraint info while keying. I'm afraid this might be one of those times where a swift smack to the head is in order, but I'm willing to take my lumps and move on for the sake of learning if I can find to answer to this "problem" I'm having. I thought the "key constraints" button was meant to be left on if you are using constraints in the rig. Am I wrong? HandAction.act
  5. So far, this rig gives really nice control over body movements during the walk/run actions I've been playing with. I especially like the independent hip and shoulder rotation capability, and the knee and foot control. One odd thing I'm finding as the model is keyed is that the fingers over-clench and the arms kind of collapse at the shoulders after I've set a few keys - almost like the slider values for the hand clench are being added to with each new key set (even though the actual slider values don't change). It's possible I have the key filters set wrong or something. I have the "translate, scale, rotate, constraints, poses, etc." filters set to be keyed for the whole model. Take a look at the screenshot and see if you have any ideas for why this is happening. It very well may be just my inexperience in animating that's the problem. It's a fun rig to use, David. Keep it up!
  6. I've only had a couple of minutes to check it out, David, but so far SOOOOO good! Thanks a lot for the great foot control. The ball pivot looks and works great. I'll play more this weekend and get back to you.
  7. This quote from the Mike Brown puppet link I refered to previously: The feet are really the only part of your rig, David, that I struggle with sometimes. They work fine as is, but I think the flexibility to use both heal and ball rotation would be great. I'm still finding all sorts of nice features every time I use it. Great job! Keep us posted on your progress. Mark
  8. David, here's a great article written by Mike Brown regarding setups for good CG puppets. Take a look... www.anticz.com/puppets.htm
  9. David, thanks a lot for the great new changes and the info on how you did it. Amazing... I haven't abandoned you here. It's been really busy this week and I haven't had as much time to work with the rig as I'd hoped. By the end of the weekend I hope to have spent enough time with it to comment more. Mark
  10. Hey David, I feel like the guy who rubbed the magic lamp and had his wishes granted. You nailed it, buddy (Genie). It works great. I don't have time this morning to work with it more, but tonight I'll being playing like mad. Thanks for the awesome work! PS. I'm very curious how you added the extra hip and knee controls. I'll look into that tonight, too.
  11. Hi David, I took a look at your helpful descriptions... Knees first: What I had in mind was control of knee direction completely independent of the hips and feet. For example, some people walk with their knees almost touching in the passing pose, and others walk with the knees flared out. It's very handy to be able to move the knee without pulling the foot along with it. Feet: After testing your foot setup more, I think your foot solution is probably simpler and more predictable than what I've been using. Your balance beam example described your intent for the heal control well. Hips: Thanks. I'm interested to know what your thoughts are on this.
  12. Nice changes on locking the hands! Very good. Here are some observations/suggestions: I've spent the last few weeks learning how to animate walks. After trying several different rigs I've found that having pelvis movement completely independent of upper body movement is very valuable. What I'm thinking of is the functionality of John Keates' rig (and others) where the hips pivot from the belly button and the torso is completely unaffected. For example, in a sexy feminine walk the hips swing side-to-side and rotate, pivoting from the navel, while the upper body is swaying and twisting in the opposite direction. In John's rig the upper and lower body can be animated independently without messing up the other. Unless I'm missing something, there's really no way to do this with your rig. The "hips_IK" moves the upper body as well. "Chest-FK" does work as I would expect, however, which is nice. Also, By having the upper and lower body movement completely independent, it's much easier to change one type of walk style into another. It makes it less likely to have to unanimate unintended movement, too. How tough would it be to add knee target controls for independent knee movement? The "knee_squetch_targets" could be used for this, I guess, but there would be the "problem" of unnatural shifts because the squetch control allows movement in all directions. One other thing I've noticed is that the heel control only moves the heel itself, not the whole leg. If the heel is raised, shouldn't the whole leg lift? It's a great rig as is, David, so don't feel obligated to add or change anything. I'm amazed that this could be used for either a cartoony or realistic character because of how well you've laid out everything. I love the hide/show options. Great job! Mark
  13. As usual, David, awesome effort here. I haven't checked on your progress since the early versions, and I'm impressed with how far you've taken this. He's a fun guy to work with. The time you've spent is huge, I'm sure. Thanks for all of your hard work! Now I'm going to go put him through a good workout. Mark
  14. Here's the thread regarding "Weight Flipper": http://www.hash.com/forums/index.php?showt...11229&hl=weight And here's a link to the actual utility: http://am.tigerbean.dk/weightflipper/ TSM2 Flipper is great because it doesn't duplicate the unilateral bones in the center of the model, like the spine, neck, head, pelvis, etc. You get a beautifully flipped skeleton without having to delete any duplicate bones. One thing about this technique, though, is that you have to start with the "right" half of the skeleton for it to work. So for John's rig I had to use a text editor first to rename left to right. Well....
  15. Hey Vern, I ended up using TSM2's "flipper" for the skeleton, Text editor for the relationships, copy/flip/attach for the mesh, PHP Weight Flipper for weights, and Mirrorsplines to realign bones and CP's as needed (both have to be exactly mirrored for "mirror smartskin" to work correctly). Any FYI, I was not able to automatically flip the smartskins either. They were lost somewhere in the process each time I tried, so I redid them all. A great big thanks to John Keates for sharing this rig. It's a beauty! Mark
  16. Mark, I tested out your idea this morning and found that it was confusing having the controls slide around, independent of the face. I think that if the goal is to always have the interface square-on to the camera, then the technique you described or Lazlo's (in the post that started this thread) or the one Bill Gaylord suggested of tying the controls to a camera view would be better. For this on-face rig, I'm still in favor of the setup I described a couple of posts back.
  17. Thanks for the kind words. I'd like very much to see what others have been doing with the Osipa interface. Some video of their work with the interface in action would be nice. Come on everybody, let's see your stuff! I'm sure there are lots of good modifications out there we can oggle over
  18. Hey David, I've been following your progress here. This last model is really fun to work with. Very nice idea to combine the hand and elbow control. Your controls are organized well, making it an easy rig to use. Thanks for working on this project and posting your updates. I'll stay tuned... Mark
  19. Ben, I like the extra dimension you get from from using the bone's roll. I may have to give that a try for fun. This is pretty much the same way I do it. The individual sliders work independantly of the nulls, so I go back and fine-tune with the sliders where necessary.
  20. You likely have a lot more experience relating to this than I do, Bill. I just recently gave this character a body and haven't yet animated a scene with it. I've been experimenting with a 2-camera setup that seems to work very well with the on-face controls. There is a close-up camera that sits just behind the main camera and has "track to" and "orient like" constraints to it, and an "aim at" constraint to the head. The close-up cam has a zoom of 250. The perspective between the two cameras is nearly identical. If I can't grab the nulls easily with the main camera, I just switch to the close-up with the 1-key and it's much easier to work with. Occasionally I have to readjust the position of the close-up camera with the shift/mouse wheel a bit, but that's really fast too. This is blast to work with, so far. All of the controlling nulls are constrained to move in only the X/Y plane and work well as long as the character is not turned far toward a profile view. Tell me if you think this isn't practical.
  21. Lazlo, I just checked it out to be sure and here's how it goes: The mouth null has X & Y relationships that control three sliders/poses (mouth open, smile/wide, and narrow). The "mouth open" slider/pose does control the rotation of the jaw bone, as well as muscle (CP) sculpting. I don't remember any crashes related to nulls controlling sliders and bone movement. I'll think on it and get back to you if anything comes to mind.
  22. Thanks for the input, Mike. I hope it was helpful to you. If you, or anyone else, have any other questions feel free to fire away. Mark
  23. Mike, Here's a screenshot: And Here's a DivX 5.5.1 video (8.6 MB) of what each slider does (I forgot to show the one on the bottom left that moves the jaw left/right/forward): http://www.mstrohbehn.lunarpages.com/Video...OsipiaSetup.avi The photo and video shows both sets of controls visible at the same time. I use it this way because it's not confusing for me. But either set can be hidden. As far as a tute goes, here is a cut and paste from a previous post of mine on how the interface is made. 1. In the modeling window, make the outlines for the interface and add type if you wish. For the on-face rig you won't be making the outlines or adding type. 2. Add the nulls (modeling window), rename, resize, and position them in the proper places in the interface outlines. 3. Make a new On/Off pose and rename it Face Constraints. 4. In the relationship window (for the Face Constraint pose), click on a null and select New>constraint>translate limits. Set the "min" and "max" values which will restrict the x and y movement of the null. Repeat this process for each null. This completes the interface part. Now you need to make relationships for each null slider position. 4. In the PWS go to Model>Bones and select one of the nulls. Open its' properties tab and go to Translate X (or Y, depending on how your slider is going to work), and right-click, select New Relationship. 5. In the relationship window, make your pose slider window visible and set the relevent pose slider's values to zero (the sliders that will be affected by the null). Move the null to one of its' extreme positions and set the appropriate pose slider to 100%. Now, move the null to its' other extreme position, and change another pose sliders' value to 100%. For example, on the right side of a side-to-side null slider I put a wide mouth pose and on the left side I put the narrow mouth pose. Close the relationship window and repeat this process for each null sliders' x and y value you have. 6. Test the null slider in an action window. As you translate each null it should automatically control whatever pose sliders you attached to it. 7. Save, save, save.... repeat. Good luck!
  24. Here's a piece I did a while back that shows a lipsync with the Osipa-style onscreen interface I use for animating. About 90% of the lipsync was done using the "slider interface" and the remaining 10% was tweeking the individual pose sliders. (DivX 5.1.1, 2.1MB) http://www.mstrohbehn.lunarpages.com/Videos/Yosemite26.avi Since then I've moved the controls onto the face and have two sets of controls, one similar to the simple setup, and one that has unilateral controls. Each set of controls can be hidden to reduce clutter. I'm very happy with the intuitive feel of working with these on-face controls.
  25. There are no fast, easy ways to mirror cp weighting that I know of. Thanks a lot for the explanation of your shoulder rig and for posting quicktime examples. As soon as I get a bit more time I'm going to dig into this. Two thumbs up for your efforts to keep the bone-count down for transferability.
×
×
  • Create New...