New Wonder Woman renders

Mr. X · 11 years ago

ttb51m · 11 years ago

The entirity of this is classified nonya

Stemming from years of observation, digital conversations and graphic media collection,there are several options available but only one viable thread that will directly affect the decision point on who rules the poser realm in superheroine fetish. After exaustive measures and countless man hours of work there is only one significant conclusion that can be logically and evedenciery derived from this applicable set of viewpoints. All polling and analysis on trends and future courses of action aside on this subject, it seems apparant that all indicators point towards one simple and most likely endstate,

MR X IS FU*K!NG AWESOME!!!!!!!!!

Pending any questions thats the end of my brief, ill be followed by a legion of X's fans co signing with me

Mr. X · 11 years ago

Thanks!

cdrei · 11 years ago

Great stuff, as ever.

I've been an admirer of your Poser work ever since the late 90s, and you were the first person to teach me the arcane Poser art of modifying cr2 files. I'm glad you've kept at it. Wish I could bring myself to render more these days, but Poser projects always seem to get bogged down in endless tweaks to the same handful of figures and characters, for me. I rarely seem to get past test renders because of this. Sigh.

I don't know if you'll want to reveal such things, but I am interested in what base figures you're using here, when the renders were produced, and what version of Poser was used. Poser geek details.

Have you worked at all with weight-mapped figures or subsurface scattering, the Poser 9 stuff? It can be a PITA to set up and to convert all of the figures, but the results can be very nice.

Mr. X · 11 years ago

I use Vicky 4 now. Used Vicky 3 for quite a while. I use Poser 2012 for renders. No I have not set up any scattering or weighted maps. If I see any good examples of them maybe I'll do something with them.

cdrei · 11 years ago

Several of us collaborated on developing a weight-mapped V4, last year. It's a fairly easy conversion to copy the weight maps to any garments, but character conversion can still be rather involved when there are complicated morphs that might need to be ported. If you're ever interested in exploring the option, the links are in a stickied thread at 'Rosity.

Dunno if they'd be useful to you, but I have some scripts for auto-creation within Poser of ropes, chains, and cocoons.

http://www.the.cage.page.phantom3d.net/ ... loopy.html

Umm. Apologies. I become enthused when I encounter other Poser SHIB enthusiasts, you know.

Mr. X · 11 years ago

Hmm I've been trying to find or write my own path following script so I can have a tentacle wrap around a figure during an animation. It would effectively snake along a hidden path.

cdrei · 11 years ago

If you have a poseable tentacle to test, the chain posing script at the link may or may not be able to help. It has some problems, due to my mathematical limitations. Poser's rotations use euler angles, which are susceptible to gimbal lock. Even if I do the math using matrices or quaternions, the conversion back to eulers can frequently lead to trouble.

At any rate, the chain posing script was intended to address problems like the one you want to tackle. It isn't all there yet, unfortunately.

Mr. X · 11 years ago

I just haven't had time to do my own script. Yes in python the rotations are eulers but I', not sure of you can give an object a new matrix. The matrices will not suffer gimble lock for small delta euler additions. I programmed a whole pool game at one point so I've dealt with these issues.

It should be interesting to see what you did. How do you plot the points for the path? Are they all world space points?

cdrei · 11 years ago

It's been a couple of years since I hacked the script together, so my recollection of some specifics may be a bit rusty.

But, yes, I think I used the worldspace coordinates. There had to be a lot of conversion back and forth between world- and localspace though, IIRC. If you have a decent grasp of how 3D math works (almost anyone's must surely be better than mine

), then maybe you can make it all work. I wasn't quite able to do so.

The whole set of scripts is a hackwork job, once you really look inside. There's a lot of borrowed freeware code, with attribution and source URLs, making the whole thing work. I guess I'm bright enough to figure out how the pieces can fit together for a workable result (sometimes), but not really smart enough to construct all of those pieces myself. Alas.

If you can come up with a solution to the euler problem, I know there are several in the Poser community who would receive your planned script with excitement and enthusiasm, if you saw fit to share it with them.

Mr. X · 11 years ago

I assume the process works along the concept of moving the root node of the "tentacle" along the path then recursively pointing each successive node back down the path toward the start. I have to look at poser python but it it accepts matrices directly then one can simply form a matrix of the path segment direction and then translate that into local space and apply that to the node.

I'm curious what did you use for a path itself.

cdrei · 11 years ago

The path is defined using a series of marker points. The basic process, unless I miss your meaning somehow, is more or less as you suggest. The trouble I ended up encountering was with the orientation of the base actor. I was able to come up with a way to work around gimbal lock, IIRC, in most cases by using coordinate space conversions derived from a couple of Blender Python scripts. I'll have to look at the script again.

Unfortunately, the trick I applied wouldn't work on the base link, because it had no parent link with which to work the coordinate space conversion. The base link orientation had to be set in world space, and was sensitive to gimbal lock. But the subsequent links would end up compensating, eventually finding the correct path but developing an ugly snarl between a mis-rotated base link and the point where the path ends up being corrected. I didn't understand the borrowed Blender process (Ideasman is far smarter than I

) well enough to find a way around this. I'm probably not explaining this well. It's the first time I've thought about it for a couple of years....

I suspect you might have better luck, if you know your way around these things. Frankly, I was surprised to have gotten as far as I did, given the troubles I usually have when rotations are involved. Sadly, I peaked out with math after a couple of years of algebra and geometry, and could never get the hang of trigonometry or calculus.

Unless a new option has been added with Poser Pro 2012 SR3.1, Poser rotations can only be set using euler angles. I haven't looked at the updated Poser Python documentation recently, however.

I guess my earlier answer about using worldspace coordinates may not be fully accurate. Worldspace is used for the translation and the base link. The child links may be handled mostly using localspace rotations, relative to one another. I'll just go ahead and peruse the script now, I think.

cdrei · 11 years ago

Umm. Okay. It looks like I started with a rest matrix for the figure, which would be the same for all actors since they need to have the same joint orientation for the script. Each of the links uses the rest matrix and its inverse to strip the parent rotation from its orientation, perhaps creating something similar to the delta value you suggest. This process ends up being self-correcting, as I noted, but the process also fails miserably too much of the time, with that base link. I could never figure out why, but the base link to its BODY parent could never be handled the same as the other parent-child relationships, for the purpose of removing parent rotations to arrive at usable child rotations.

Yeah, I was really winging it, there, actually.

Perhaps I could try to dignify the result and suggest that it was arrived at heuristically. Umm. But I doubt it.

I also re-read your last post. The posing process does start with the base link, but then works forward from there, not back toward the base from the final link. The rotation for each child is calculated according to the orientation of its parent. There's probably a better way. I don't recall ever considering working backwards.

Code: Select all

                act = self.chain.ActorByInternalName(acts[i])
                pd = pathdata[i]                
                """
                The following was developed after studying the approach used by Cambo/Ideasman (Campbell Barton)
                in his 2006 BVH import script for Blender 2.42.                
                """
                act_localmat = rotmat_from_vecs(pd[0],pd[1],pd[2],pos=pd[3])  #pd is path data, from marker point path              
                act_localmat = matrix_multiply(restmat,act_localmat)                   #Strip child orientation data using restmatrix
                act_localmat = matrix_multiply(act_localmat,restmat_inv)
                
                par_worldmat = act.Parent().WorldMatrix()
                inv_par_worldmat = matrix_invert(par_worldmat)
                
                act_localmat = matrix_multiply(inv_par_worldmat,act_localmat)                
                
                # Derive euler angles from matrix.  Sensitive to gimbal errors, but these are
                # tolerably infrequent with a chained figure posed along a spline path.                
                xr,yr,zr,error = mat3ToEuler(act_localmat)
                if status:
                    if error:
                        if error == 1:
                            show_status("%s: Gimbal lock at north pole." %(act.Name()))
                        else:
                            show_status("%s: Gimbal lock at south pole." %(act.Name()))
                
                xparm = act.ParameterByCode(poser.kParmCodeXROT)
                yparm = act.ParameterByCode(poser.kParmCodeYROT)
                zparm = act.ParameterByCode(poser.kParmCodeZROT)
                xparm.SetValue(xr) 
                yparm.SetValue(yr)
                zparm.SetValue(zr)

viking · 11 years ago

I find your discussion fascinating. For me, a computer programing novice, it reads like an exotic foreign language. It give me additional admiration for the poser work that you and others produce. I appreciate all the effort you make for our entertainment.

tallyho · 11 years ago

Here, here Viking, well said! Thanks a lot guys.

hootersfiend · 11 years ago

Mr.X, those are some great images! I put you on my watch list at deviantart. (I'm genchang2112 there).

ChenSing · 11 years ago

These pics are just awesome!

11 years ago

Great stuff Mr X!