Up late and brwing some coffe and have been thinking about this idea for a very long time. The idea: To create a figure what has no mesh groups. The theory: To use bodyhandles/"Ghost Bones" and/or magnets to make the mesh move Application: Just a fun wayto try something new and figures different. Execution: You take a figure that is make of cyclinders and balls and give it a second BODY child. Basically a second hip that is seperate from the standard body hierarchy. Now we know the problem is that the joints won't affect this top skin. But... we know magnets are very similar to joints i many ways and can be paranted to parts. So you parent the magnests to the skeleton and then tell the magnets to also effect the top skin. And example I saw earlier which is similar is a thread awhile back where magnets made a figure indent things it sat on. Anyway just an idea floating in my mind I wanted to sahre. Not neccesarily practical, but interesting to maybe talk about. --------------------another idea------------------- Dynamic skin. We all know peeps have used the cloth room to caputre shapes of other figures and to make clothes. Well why not clothify skin over a muscle form? Or even a muscle figure made out of promitives. Elbows, knees etc are achored and then the skin can shift over the bone. Again not practical at this time but still an interesting idea. Could be used in conjuction witht he idea above. Okay just some fun ideas in the middle of the night. I have ahd success with a few scattered experiments and I think it would work. Especially the "Ghost bones",bodyhandles or whatever term you want to use. Regards, Anton

Forgeive the typos as usual. I really need a lamp for my keyboard. :)

Quote - Well why not clothify skin over a muscle form?

Interesting idea, but wouldn't that take forver to do on a whole figure?

And no worse than doing clothing sims now. However, if you are using 60K + poly figures it isn't practical. Well right now anyway, but at some point hardware/software will allow it. In the mean time it can be used to help shape/model human/animal forms for standard Poser rigging.

Wow Anton you brain is really cookin. Sounds like a cool new approach. I wish my Coffee did that for me. 8 )~ Tom

Uh Anton, Are we both wired the same way??? I was thinking of something similar to your first post earlier. Ironically, I was going to send you a PM about my idea, but my internet connection went fubar, and I decided to have dinner. Fast forward a few hours, and I see that you have posted pretty much what I had envisioned. We must have that Poser microchip in our heads. -The Starkdog

Stardog, look to your right. See that little black dot on the wall? That is my hidden camera. I have one in everyone's houses. :) PS: Tom is wearing a blue T-Shirt.

Amazing...lol I really am wearing a blue shirt. Well...it has coffee all over it now. Thanks Anton 8 ) Tom

heheh Seriously though. I love Poser because she can suprise you and do new things if you know how to show her. I think we, as users, have such a personal releationship with the program, we start to see patterns.

One of the rasons I was thinking about this is looking at my anatomy books. Everyone has different muscle shape/ devlopment, etc.

If the underlying muscle spheres are flexed/rotated, etc the skin will adjust to fit.

There was a post recently where I was trying to explain why, even with weight mapping, polygon skin will always be like a wetsuit crazy-glued to your skin. It will always pose like rubber.

But with a dynamic skin covering, anchors would maintain control(Contrained or rigid), allowing skin to shift over the bone and fat layer.

:) Maybe in a future version of Poser when hardware advances far enough.

I do think it would be fun to see if Magnet properties can be merged into the Poser joint system.

Message edited on: 06/27/2005 03:09

I'm still investigating how to avoid certain effects tho, to get it to work as it should u can try the following: 1.- Load AM and rename the figure to any other that the original name so u dun get confused(u'll understand why later) 2.- Delete the bicep bone of the figure in the setup room, let's work with lBicef for example 3.- Make a chain as it fits ur needs, in the image above the chain configuration is: Left shoulder Left Upper Arm Left Fore Arm Left Hand (Goal) but u can try the old 3 bone setup as well from the Upper Arm to the Hand 4.- Now conform the unchanged AM to the one with the chain and use the hierarchy editor to hide the mesh 5.- Zero pose both and if u want to work with the legs chains activated just deactivate the ones of the visible mesh to avoid unwanted glitches (END) Now u have to work the poses with the hiden AM and morphs with the visible one . . . Anton, maybe u can figure out a better way to acomplish this , what do u think? Regards

Cool. That was quite a pet project. Industrious. :) I can tell you are one of those tinkerers.

blush

I like the idea of IK. I just wish it wasn't so unpredictable. I did a head Ik on one of my dragons once so it could bite something like a tree and then thrash. Ripped the shoulders apart sometimes but was fun to watch.

Gosh ... historical thread? Sounds as if here youre talking about the near future of 3D characters

Very nice idea!

Someone else has mention this much time before and was working on a figure too...I remember I've downloaded a manequin using this technique...

Maybee the figure could be sepperated on at least 6 basic parts(head,torso,R-L leg,R-L arm) and every part could use this technic,those parts could be connected altogether with the current rigging system and fingers also...

pitklad

I'm sure I saw something here about using the Cloth Room in P5 for a dynamic breast model. Sometime in the last couple of years, I think. Elbows and knees seem to be the biggest problem, they're certainly where I notice things going wrong in a bend. I wonder if these methods could be applied to those joints? I reckon, myself, that having a straight knee/elbow as the zero-position it making things difficult. Current designs seem to be good for a little over 90 degrees of bend, with a bit more being tolerable sometimes. Start from a 45-degree bend in the mesh, and have the straight limb at -45, and it could make a big difference.

Ascii, glad to see you back here again! Anton's new model will be a good project for you to work on. Perhaps there's some news of the Anime Base figure to come?

Attached Link: http://www.getjolly.com/modules.php?op=modload&name=phpbb2&file=viewtopic.php&t=68

We created a test movie a while back in the cloth room which contains the above theory and the results were quite smooth. Check out the link attatched for the movie.

Hel@Jolly
:-)

opps edited 'cos I put in 2 http://

Message edited on: 06/27/2005 15:38

Ghost parts work fine as long as they are the last group in the chain. They work so well that the only way you can tell they area ghost part is you can't click on 'em (unless they have handles, which I seldom put in). My shoes have had ghost toes for years, Glamorous Jessi has ghost outer toes and a ghost upper head, for example. They work fine for skirts too, I do all mine that way, including one with both ghost thighs and buttocks, with some clever coding. The problem is when you go further up the chain, how are you going to get the child part to allow for the rotations of the parent? Poser does it for you if the parent is mesh, but if the parent and child are both ghosts you get into a world of hurt. I try and stay away from below-the-knee dresses for just this reason. ;-)

Additionally ghost bones and magnets positioned poorly can have a cumulative transformation effect - for example two overlaping zones with a weighting of 1 will give a resulting transform of 2 on the vertices involved. Alternatively any vertex not being influenced fully by at least 1 bone/magnet will be left behind if the character moves. With the Dynamic Skin rigging method above, modelled internal muscles and bones are controlled through Joint controlled morphs (JCM) with the external skin containing no morph information. The dynamics are applied on the Skin performing much like a sheath or sock over the internals. Giving a nice natural shaping to the elbow joint etc. The simulation has so far responded with great speed in tests. The advantages are that the character itself contains no JCM's or joints etc. and is 'driven' by the musculature. This would provide for much more flexibility with clothing but then this needs a bit more testing. Helena

I think JCM is always a bad idea. Like lense flair in photoshop. I think is is over used for what it can do. My idea was the whole point of the dynamic aspect is to get rid of the whole need for JCM in the first place. Did you mean erc. That would be feasible.

A matter of opinion, both can lead to powerful results if used correctly! As the Skin contains no controls either JCM or ERC there is no conflict. The feasibility of said method has been proven as per the test movie and our additional tests. Helena

:) I didn;t mean to offend you. Just my feelings.

Message edited on: 06/28/2005 15:40

I agree, JCM is a cludge. I tried to originally set Glamorous Jessi up with JCJ, ghost joints to do the things that the morphs do in JCM. My idea was that these could be easly added to the clothing. I actually got it to work, up to the point I tried putting it into the clothing, where the who scheme fell apart. Oh well, I tried. Most of GJ's joints are designed around JCM, I convinced myself that you could go places that you couldn't get to with straight Poser joints. Some of the bending is "wrong" in itself, but it gets the mesh to a place that the JCM can correct easier, for a better end result. There are other approaches, though.

No offence taken - I am of mild disposition :-) JCM's don't necessarily only fix dodgy joints, in combination with ERC they can create skin wrinkleling around joints, muscle bulging, tendons flexing etc. I see the only problems with JCM's being for the developer, this can sometimes be a time-consuming nightmare, but for a User the resulting effect can be marvellous. The most ideal situation would be for an official internal muscle driven system within Poser, but any such additions would exclude users importing Poser content into other applications without a great deal of work. I reckon it all comes down to whatever works, and anything goes! Helena