BL Render mentions in "Secrets of Figure Creation" that it is theoretically possible to bone a Poser figure as a single mesh. I might need to take him up on the theory and put it into practice. I have an object that needs to bend smoothly. This is sort of like bending a segmented cylinder, so there isn't much complexity involved in the mesh. At first, I was going to use only morphs to add curvature, but have decided that bones are more 'flexible' (hehe) and require less "fix" morphs for the other morphs. But, this introduces a problem in adding morphs. All of the morphs are setup for the entire mesh, not just the segments that would be required. Simple enough in other 3D apps with boning/morphing, but not Poser. Poser wants "parts" (groups, whatever) in a hierarchy and will only accept morphs on parts. To split up these morphs into the segments would not only be ridiculously time consuming, but probably nearly impossible to retain vertex order in the process as well as multiplying their number by 6, 8, 16, or 32. That's WAY too many morphs (possibly hundreds on one of the simple objects). I could use any help possible as I'm new to Poser rigging. If no solution is found, I could always stick to morphs only, but it is not my first choice. Thanks in desparation, Kuroyume

Are you using Poser 4 or 5. If you are using Poser 5, you can create your bones visually in the setup room, and then use the auto group to make your groups. Might save some time. Would help if we had a pic of the object to get a better idea.

I'm using Poser PP (I also have 5). But the issue is one of applying morphs created for an entire mesh onto an object that has been "grouped" into separate parts. Only parts can take morphs, not the Body. Splitting the morphs into itty-bitty pieces to apply to each segment is not going to happen. Most of these morphs effect the entire object. Imagine taking 30 morphs, splitting them into 8 morphs EACH (because of the grouping) and applying them. Yikes!! The "test" object is just a cylinder with 8 height segments along the X-axis. It is supposed to curve along its length (bones rotating about Z-axis) - think hose or rope here, but with morphs. ;)

For hose and ropes I think Easypose is a better solution than morphs. But I haven't done this myself, so I don't really know.

How's about "Body Handles" or "Poser Bones" as BL calls them in the aforementioned book? This seems to be exactly what I need, but there is no description on how it is applied. From the book: "'Poser Bones', as I called them, are good for adding posability, while preserving morphs,... . Jaw bones are a good example: instead of slicing off part of your creature's head to make the jaw [and thus destroying the head morphs], you can leave it all attaced and use a bone to pose it." [] my comment The idea is that you use unrendered, additional polygons as "bones" to pose an existing portion of a mesh without having to group/split it. Will have to check the accompanying CD for any hints or tutorials on this. I don't know if Easypose uses this same technology, but I'll look into it.

Easypose uses ERC, which relies on the internal geometry and still would require you to group the figure. The body handles system is essentially an extra bone/magnet that is used to move the objects vertices. I can't for the life of me remember where I first saw it, but check her website (3dmenagerie) for a mini-tute on it -- some of them made it into the book, other's didn't. I've always worked under the assumption that you cannot asign more than one bone to a group, so I wouldn't know how you could make an object posable and still retain a single object. You can bone it, which is one thing, but making it move is another -- that would require joints, which I thought required two or more bones to enable.

Here's an example: my Flo Hair is all grouped as "neck", which makes morph generation a doddle. Kozaburo does the same with his conforming hair - that's where I stole the idea from. ;-) This works because although the mesh has only one group, the CR2 has actors for head, neck and chest. The adjacent actors' joint parameters are arranged to have the appropriate effect on the neck group. Note also that, although Vicki 1/2 has an upper neck group, this can be skipped over entirely without too much trouble. However, you can't extend this principle beyond the 'children' of your single group. If I were to make the hair waist length, for instance, an abdomen group would be needed; but it wouldn't work, because although the abdomen affects the chest, that effect won't be propagated through to the neck. Body handles work in the same way - they are actors added to the CR2 to deform the mesh, but don't need any geometry associated with them. All the magic is done with the joint parameters. For your long, thin figure, you're right to choose bones. Morphs are no good for doing curvature unless it's a fixed amount; for example, a morph to make a skirt sit down is good for the target pose, but no use for just bending the legs slightly. I hope that made some sort of sense. By the way, to avoid future embarassment, BL Render is female. Read the back cover of your book. ;-)

Ahhhh!!! hey, now that's a nifty trick, EnglishBob -- I shall definitely add that to my list... Thank you!

Oh, you mean the back fold over. Well, sorry, I didn't read it. So, I'll refer to her as Ms. Render from here on in. :) As for the "body handles", my limited understanding so far is that all that is needed is to add polygons in the proper places (as the groups) to receive the bones and, as you mention, the JPs need to be set to do the rest. I can only imagine that since these polygons added to the original mesh, the bones can be made to deform the original mesh as well. My problem is this, though: I'm using C4D to model (including the morph targets). If I add these polygons to the model there and export the base mesh with them, it is almost guaranteed that the vertex ordering will not be the same as the exported morphs. There may be a way to text edit them into the file so as to retain the order, but it will be some interesting fudging. How to set the JPs to achieve this is still beyond me. Do the spherical falloff zones play a critical role?

You don't actually need any polygons for the body handles at all; but you may need them for other reasons. I put something in so that the user has something to click on, like a shadow part that has no connection to the rest of the mesh. These will be grouped as the body handle part, so they won't affect the morphability of the parent part. The new bones deform the original mesh because that's the way Poser works. If you imagine an arm bending, you may be altering the bend dial on the forearm; but the so-called shoulder will have to deform as well so that the joint won't tear apart. If you make the forearm invisible, and bend it, you'll see the effect on the shoulder. I'm not certain that bones are what you need, if you're making a snake-type figure. I may well have misread what you're trying to do. Can you post a shot of the bare mesh without giving too much away?

Here's the base mesh (in C4D). Without the "bend" morphs and fixes for other morphs because of them, there are now 25 morphs.

Here's what the "bones" are supposed to do in Poser while allowing morphs over the entire object. I'll give C4D one thing, boning a singular mesh is easy.

I can see by the plethora of responses that this is not so easy, now is it? The "Poser bones" have been figured out. It is just as simple as adding the polygons as parts and letting the bones form there to deform the entire mesh. BUT, only the first bone in the hierarchy works on the mesh. Subsequent bones, ala: 1 Base xyz ...2 Bend1 xyz ......3 Bend2 xyz etc. Bend2, have no effect on the mesh since the relationship is "grandparent" (just add "great" for each level below that). I tried just keeping them all at the same level in the hierarchy, but the results were not very useful. The bones are no longer interdependent (duh) and therefore deform the mesh incorrectly. Any ideas?

It all went quiet there. :) I was just trying to think of a method that will save you a lot of work, but sadly I don't think there is one. Hopefully someone else will contradict me. As far as I know, there is no way of keeping this as a single group that will give you the flexibility you want. It's either bend morphs on a single group (=crappy bends), or several groups + EasyPose (maybe) (= lots of work transferring the morphs). There has to be a way of transferring morphs from a single mesh to a grouped mesh, but I don't know of a general purpose method. I'd advise a few tests to see how few segments you can get away with and still retain the flexibility you want. Including the curve dial on each segment ought to help out, but apparently it doesn't always. :) Results have been rubbish whenever I've tried it, so don't ask me. ;)

Same here. Spent the day doing trials to see what could be worked out. Interestingly, I went back to the bones being at the same level in the hierarchy: 1 Base xyz ...2 Bend1 xyz ...2 Bend2 xyz ...2 Bend3 xyz This is how Ms. Render accomplished her "Flubber" figure with one mesh and bones to move legs, arms, and head. But these bones are not interdependent as mine should be. I have added the "poser bone" polygons to the mesh (this is still the "test" mesh cylinder divided into 8 height segments) and have created the figure using PhiBuilder. Because PhiBuilder doesn't specify JPs, the Setup room is then used to set them correctly (although not necessary, it at least gives visual feedback on the bones). This actually works except for some weird anomaly with the last bone being rotated (but not the mesh) by the other bones. When that last bone is rotated, it has been so distorted and rotated that it causes major distortions when it is rotated to deform the mesh. I have scoured the CR2 file and found nothing to indicate why this happens. A possible solution to this, untried, may be to add a "dummy" bone as the last one, outside of the base mesh with JPs set to avoid deforming it. This dummy bone can then be rotated to heck by the other bones' weird behavior and have no influence on the mesh. Will let you know the results.

Okay, found the reason for the anomaly. Simplistic, if not diabolical (muyahahahaha, emmm). The entire reason that that last bone gets rotated by the others is that one point of the polygon (triangle) coincides with a point on the base mesh. During weld of the figure (when loaded as a Figure), I suspect, these two points are merged. By moving that point off of the base mesh, the problem disappeared; no need for a "dummy bone" in the process. Under extreme settings, the mesh does get disfigured, but the expected settings are like 5 degrees or less which allows for mimicking of the real thing with minimal distortion. Finally on my way to completing this confounded project! If you're out there monitoring this, Thank You Very Much, Ms. BL Render for this workaround that allows bone deformation on a mesh without dividing it into parts so that it can take full morphs!! :0) Kuroyume (signing off for now)

Way to go. Please let us know how you get on.

Point your browser to the Product Showcase forum occasionally and you'll see the fruits of my hell, eh labour, :) in about a week. Thanks for your help and encouragement also!

And now a very dumb question of mine (I expect it's dumb, since no one asked it) : why didn't you use a magnet to deform the mesh in Poser ? (You can use the "affect only one group" option to make it deform only the blade, you can adjust the falloff Graph to give the curvature you need, and you can save it in the library along the cr2)

Because this isn't "one time only" curvature creation. This is part of the product: the ability to add curvature of differing amounts, centered at differing points along the blade and tang, in either the positive or negative direction, is on the onous of the user (or preconfigured figures). Additionally, avoidance of morphs for curvature (same as using magnets) is imperative. They do nasty things to the other morphs and require all sorts of fixes (none of them completely reliable), whereas bones do not suffer from this ailment. Multiplicatively, I abhor Poser magnets. I've tried them several times. I found them to be like modeling with a wrecking ball (no discredit to those who can use them well). Creating morphs externally to Poser is the only way it will ever be done by me.

OK, I get the idea. I'll peak in the product showcase to see what you'll come up with : it looks promising right now.