Thanks for the encouragement Rogerbee1 ;-)

Okay I think I worked out what is going on with my jcm.
It looks like MorphingClothes multiples the morph target values by whatever value is in deltaAddDelta when transferring to the clothing item, so in the case of the Kick jcm it was multiplying by -0.016667 aka 1/60th. This tallies with what Cage said, that the jcm morph should be at 1.0 at -60 degrees. Sooooo my earlier jcm morph transfer was getting 1/60th of the correct value, add a bit of rounding error and I can see how things would work at low kick angles but deteriotate a higher angles, and explains why I was having to use a value of deltaAddDelta 1.
I didn't realise Morping Clothes did that, but now that I do know I can easily work around it. All I need to do is copy Antonia cr2 and set the deltaAddDelta to 1 or -1 for the jcm morphs, and then transfer to the clothing item through Morphing Clothes .

Ok I tried my suggested work around and the morphs come out 60 times bigger so it appears to be a valid theory. I now need set up the jcm in the shorts again, but I learned something, and that's got to be a good thing, 

Cheers

Ian

Ian,

There are primarily four settings you have to be concerned about in a JCM targetGeom channel (or valueParm, if you're chaining), other than the ERC controlling actor and joint.

forceLimits 1
min 0
max 1
deltaAddDelta

forceLimits causes the channel to respect max and min.  Max and min control how far deltaAddDelta can go.  Meaning, max and min are used to keep the morph from going too far in either direction.  Sometimes we adjust max higher than 1, if it is necessary.  Min is typically 0, to keep JCM from working when the joint is bent the opposite direction.

The deltaAddDelta formula is 1 / angle degrees.  If you have something like a an rThigh xrot bent to -60, the calculation is 1 / -60 = -0.016667.  So, set your deltaAddDelta to -0.016667 in order to have the morph reach a setting of 1.0 at -60 degrees.

If the rotation is a negative number, the deltaAddDelta will be a negative number, however, the min will still be zero and the max will still be a positive number.

Now you're getting somewhere.

Short and denim will do me, a la Daisy Duke!

CHEERS!

ODF,

I'm working on Antonia compatibility for a utility I'm writing and stumbled into something kind of surprising.  Are you aware that her Thumb1 actors cannot be rotated on the Z axis?  RotateZ zrot is set for min/max 0 with forceLimits 1.

Hrm.  I just realized there is more to it than that.  Maybe you are still working on the joints for that actor?  The Thumb1 joint order is YZX, whereas the rest of the hand and finger actors are XYZ, which is the same as Vicky.

• Mike

Quote -
Okay I think I worked out what is going on with my jcm.
It looks like MorphingClothes multiples the morph target values by whatever value is in deltaAddDelta when transferring to the clothing item, so in the case of the Kick jcm it was multiplying by -0.016667 aka 1/60th. This tallies with what Cage said, that the jcm morph should be at 1.0 at -60 degrees. Sooooo my earlier jcm morph transfer was getting 1/60th of the correct value, add a bit of rounding error and I can see how things would work at low kick angles but deteriotate a higher angles, and explains why I was having to use a value of deltaAddDelta 1.
I didn't realise Morping Clothes did that, but now that I do know I can easily work around it. All I need to do is copy Antonia cr2 and set the deltaAddDelta to 1 or -1 for the jcm morphs, and then transfer to the clothing item through Morphing Clothes .

Ok I tried my suggested work around and the morphs come out 60 times bigger so it appears to be a valid theory. I now need set up the jcm in the shorts again, but I learned something, and that's got to be a good thing,

It sounds like you're getting there.

Morphing clothes can transfer the morph part of the JCM's quite well, but I couldn't find a setting that handled the ERC portion correctly.  It either linked the morph to a Body FBM dial incorrectly, or did something completely wrong when linking to the correct joint rotation.  I found it most straightforward to just delete all of the JCM ERC added by Morphing Clothes, then copy the correct version over from the original Antonia .cr2.  Maybe someone with a better knowledge of Morphing Clothes knows how to get all of it right using just that program.  I found it useful only for transferring the morphs.

And it can definitely do that fairly well.  If the correct morphs have been transferred over, they should be set to reach 1.0 at -60 bend for both legs and 60/-60 for spread on left/right.  I'm probably only confusing matters with my efforts to explain.    As Rogerbee1 said earlier, keep at it.  Progress is being made.  ERC can be a tricky beast.  :D 

Ok,

I just tested her in Carrara 8 (64 bit) and... It worked (as can be seen :P)

The hair used is Posh Hair, resized to 95% or something like that and then placed (still not that happy about that :/)

The pose is from TGP Pose set 01 which is really nice ;)
The skin used was Toni skin (without VSS)

Anyway, lots of good work all ^^

Just a question though, I couldn't find the facemorphs and expressions in there, are these supposed to be somewhere?

Many thanks and keep up the good work ;)

Andro

Thanks Mike, Rogerbee1,Cage,

Well I fixed up the jcm's to use the -0.016667 but still getting the poke through at -35 degrees and onward exactly as before, so it looks like it isn't a rounding error caused by multiplying up a tiny morph. I think I'm going to cheat and fix it up with the Poser morphing tool, which is a total cop out I know ;-(, question is having corrected the kick morph at -60 degrees using the morph tool I now have the original jcm morph and a custom morph acting together at -60 degrees. How would I combine those into one morph that works through jcm and achives the same result at -60? I'm talking here about combining the two morphs rather than coding the jcm.

You are probably getting the impression, quite rightly btw, that I'm  unfamiiar with some of this stuff. I apologise if I'm taking the thread off topic, my only ( weak ) excuse is that it might help others who want to make clothing for Antonia.

Cheers

Ian

I just tested her in Carrara 8 (64 bit) and... It worked (as can be seen :P)

The hair used is Posh Hair, resized to 95% or something like that and then placed (still not that happy about that :/)

The pose is from TGP Pose set 01 which is really nice ;)
The skin used was Toni skin (without VSS)

Anyway, lots of good work all ^^

Just a question though, I couldn't find the facemorphs and expressions in there, are these supposed to be somewhere?

Many thanks and keep up the good work ;)

Andro

Quote - ODF,

I'm working on Antonia compatibility for a utility I'm writing and stumbled into something kind of surprising.  Are you aware that her Thumb1 actors cannot be rotated on the Z axis?  RotateZ zrot is set for min/max 0 with forceLimits 1.

Hrm.  I just realized there is more to it than that.  Maybe you are still working on the joints for that actor?  The Thumb1 joint order is YZX, whereas the rest of the hand and finger actors are XYZ, which is the same as Vicky.

• Mike

My fault there.  :)  When I rigged her I was not really considering compatibility. I was just going for a more realistic thumb movement. Sorry.

I have done much the same with Brads thumbs, perhaps I should consider changeing it.  Does it cause serious issues with compatibility?

Nice render Andro, I'm starting to think again that Antonia looks pretty good au-naturel and maybe it's wrong cover up such loveliness ;-)

I think I answered my own question regarding combining the morphs. 
Set all rotations to zero to eliminate
set jcm morph to 1 for right side kick
set custom morph to 1 for right side kick cop out
spawn combined morph
use this as the new jcm for -60 right side kick

repeat for left side

Cheers

Ian

Quote -
I think I answered my own question regarding combining the morphs. 
Set all rotations to zero to eliminate
set jcm morph to 1 for right side kick
set custom morph to 1 for right side kick cop out
spawn combined morph
use this as the new jcm for -60 right side kick

repeat for left side

Absolutely correct.  :woot:

Maybe we need to create a tutorial for creating JCM's for Antonia.  Information like this, potentially useful to other clothing developers, will get lost in a thread of this length.

Quote - I have done much the same with Brads thumbs, perhaps I should consider changeing it.  Does it cause serious issues with compatibility?

For the utility I'm writing, I don't think it makes much difference.  But, it is something content creators targeting her hands should be aware of, simply because it is different from most other human figures.

The Thumb1 actors being immobilized is very strange though.  I mean, if you're going to immobilize an actor, why not just weld it and have less actors to deal with?

Quote - I think I answered my own question regarding combining the morphs. 
Set all rotations to zero to eliminate
set jcm morph to 1 for right side kick
set custom morph to 1 for right side kick cop out
spawn combined morph
use this as the new jcm for -60 right side kick

repeat for left side

Exactly correct.

Something you should know about clothing JCM.  It's pretty unusual to simply transfer morphs with Morphing Clothes, set up the proper ERC and have that be enough.  I've done clothes for V4 where I had to add dozens of additional JCM to completely eliminate poke-through.  For instance, the morph may look great at -60, but not good halfway at -30, and terrible at -90.  Part of the reason for having to do stuff like that with V4 is fighting her magnet JCM.  Antonia bends pretty well and may be easier to work with (I haven't tried clothing her, yet).  But, don't be surprised if a problem area arises where transfering morphs from Morphing Clothes simply isn't getting it done.

Yea, I agree, rotation SHOULD be implemented for the first joint of the thumb.  A human thumb has that sort of moveability so should the mesh thumb so it works as expected.  Just my 2 ¢

The thumb rotation can be turned back on, it will just need some falloff zones set up for the rotation. In order to change the rotation order to match V4, it would need redone basically, but the actuall center locations could remain the same. 

After thinking about it a minute, it would not be too hard to swap the rotation order. Just would need to copy the fall off zones to the new order from the old. IE bend(grasp) to bend, and so on.  The same could be done with any JCM's that may be present. 

I'm at work now but I will check the viability of a swap when I get home. But then there are some great hand poses done for Antonia already too, have to think about that.

Manually altering joint order in a finished cr2 does not sound like fun to me.  I did not intend to create work for anybody.  I was just surprised and wondered if it was by design.

If you're using Poser 7 or 8 on Windows, you can use the Python script I've been using to make symmetrical or flipped morphs for Antonia.  These requirements hold because the script utilizes a special extension for PoserPython, written by Spanki.

The script is attached.  The .pyd can be found at this link.

If you use the script, just set up your morphed actor and run the script.  Select "Match Sym", then "Flip morph".  A flipped version of the morph will be created as "symmetry", possibly with a suffix if you already have a morph of that name.

Note that this will only work on actors which have symmetry on the X-axis.  Good for the torso and head.  Other methods are needed for collars or thighs.

I've found this to be a time-saver, myself.

I just patched this together for my own use with Antonia development.  Even so, my apologies that I'm veering off-topic, here....  :blushing:

I have been following this thread behind the scenes and I really must comment that Antonia has really blossomed into one beautifully constructed model.  I congratulate everyone involved. 

Just recently I downloaded and installed the most recent version of Antonia, and have found that her open mouth morph does not include her teeth and jaw.  I am wondering if this was an oversight or her head rigging was not completely finished due to pending changes.  It would be nice to have some facial articulation.

The upper and lower jaws, as well as tongue parts, are separate actors in Antonia.  The mouth morphs could be ERC-linked to the lower jaw, to make it rotate, but in the current version the lower jaw needs to be rotated by hand.

It's actually a very sensible setup, IMO.  It makes things easier, not having internal head geometry confusing things sometimes.  Magnet morphs, for instance, no longer run the risk of deforming the teeth along with the cheeks or lips.  I don't recall whether the .cr2 holds zero deltas in morphs.  If it does, separate actors for teeth and tongue cuts down on useless zero delta listings, reducing filesize.  It makes it easier to access the internal geometries, too, for morphing purposes.

Quote -
Just recently I downloaded and installed the most recent version of Antonia, and have found that her open mouth morph does not include her teeth and jaw.  I am wondering if this was an oversight or her head rigging was not completely finished due to pending changes.  It would be nice to have some facial articulation.

I'm aching to work on that, but unfortunately the body rig has the priority right now, since we need a stable basis for conforming clothes. There are still a number of things that bug me regarding her shoulders, elbows and knees, but I think we're getting there. I might do some work on her expressions before fine-tuning the hands and feet, though.

Ditto to Cage's post!

Odf, my apologies for fixating on this script, which is pretty much OT.  Once a bad script is posted, a fix needs to be posted so people don't have trouble if they use it.    Here's the fix, and I won't keep doing this to your thread.  Sorry.  Sorry-sorry.  :blushing:

Unless you'd like a script to memorize and zero FBM's, which I've found useful for morphing Antonia....  :laugh:

To be honest I only understand less than half of what you guys have been saying! I don't speak developer!

The only work I've ever done with .cr2's and the like was when I manually altered the V4 Basicwear ones (following DavidGB's instructions) as I got so sick of waiting for Daz to sort them out. I got everything working but when Daz's versions eventually came out I decided to use theirs as I knew I'd never lose them.

I admire all of you for being able to do what you do, I wouldn't know where to start with developing a figure from scratch. Antonia is a shining example of just what can be achieved and I can't wait for the final version.

CHEERS!

I don't know anything about rigging clothes, so I'm just following that discussion silently for the moment. If  anyone wants to go ahead and write Antonia-specific tutorials and such, I'll very happy to put them up on her site (or link to them if you prefer to publish them somewhere else). Just let me know. I'd like to put an FAQ up at some point, as well, so any help with the major project of sifting through this thread and collecting useful tidbits would also be very much appreciated.

I can tell, but, that said, no two knees are going to be entirely identical anyway. They look good though.

CHEERS!

Quote - Guess which side has the new JCMs applied in this picture?

I hope it's her left (our right).  IMO, her left looks more like a real knee.

Hmm, be nice if someone could develop some hair for her. She needs a signature style that's all hers.

CHEERS!

Quote - > Quote - Guess which side has the new JCMs applied in this picture?

I hope it's her left (our right).  IMO, her left looks more like a real knee.

I very much hope you're either joking or have a problem distinguishing left and right.

This was not supposed to be a hard question.

LOL! I think pointy was the big clue!

CHEERS!

I guess it's not so obvious from a single picture, particularly one that does not show very clearly how much the knees are bent. My bad! Her left knee would have the better shape for a 90 degree angle or so, but in fact, they are bent much more and thus need to be broader and flatter at the top.

Quote - I guess it's not so obvious from a single picture, particularly one that does not show very clearly how much the knees are bent. My bad! Her left knee would have the better shape for a 90 degree angle or so, but in fact, they are bent much more and thus need to be broader and flatter at the top.

Heh...  Sorry about that.  I guess I'm all about women with small knees.

It's just that every woman I've ever seen with her knees bent up under her has an appearance where the knee looks small compared to the rest of the leg.  Part of that is because a knee joint is typically smaller than the rest of the leg, but a lot of it has to do with the calf and hamstring being smashed together, causing them to flare out, making the knee look even smaller.

The version I've played with bends very well, in my opinion.  Considerably better than V4, especially in the shoulder and hip areas where V4 is notoriously difficult..  I love the way you did the hip actors.  I've suspected for a long time that the age-old habit of splitting the hips in half with right and left thigh or buttock actors might cause more problems than it helps and Antonia seems to confirm this.

The best areas of Antonia are her arms and shoulders. When certain morphs, in particular the Elite Sylph, were applied to V4 there were oddities in her forearms that just shouldn't have been there. Antonia is entirely natural in any pose I've seen. When you look at the shoulders and upper arms in odf's render they look just right, the same most certainly can't be said of V4, indeed it took a third party artist to correct her, which seems quite unnecessary to me in that the developer should have taken care of this before release.

Considering the stage of development Antonia is at, things can only get better

CHEERS!

Quote -
The version I've played with bends very well, in my opinion.  Considerably better than V4, especially in the shoulder and hip areas where V4 is notoriously difficult..  I love the way you did the hip actors.  I've suspected for a long time that the age-old habit of splitting the hips in half with right and left thigh or buttock actors might cause more problems than it helps and Antonia seems to confirm this.

You are absolutely right. I started Rigging Antonia with a right and left buttocks bone, which made the legs unstable, plus that you could not get smooth transitions in the hip with it broken up. For some things I liked the right and left buttocks, you can get a better side to side bend with it but the problems with it outweighed the benefits.

So I went with the bike short style hip and single right and left thigh actors. The bike style hip is my favorite, because it gives me a large hip group to work with for the thigh bends. You can get smoother bends and place the thigh centers in a more anatomically correct position.

I love rigging (about the only thing I am good at LOL) I would go on for hours, but I know its sort of booring.  :)

cheers.

Whatever you're doing, whatever you're working on, keep it up! I'll work on the finished product and see who I can turn her into.

CHEERS!

Cage, 
Your script as described sounds ideal for transfering a good right hand side morph target onto the left hand side( and vice versa ), unfortunately I haven't had much luck making it work. This will be entirely down to me being tired after a busy day at work I'm sure.
 
What I did was load Antonia and the shorts and set her right leg to -60 degrees then I corrected the poke through with a custom morph made with the morph tool.  I then ran your script ( the most recent one you posted here ) and selected mirror, selected the hip2, selected -ve to +ve, and selected the custom morph. I then saw a dialog box come up with a stream of 'No match  followed by coordinates. One this ended a new morph target was created on the hip2, called custom morph_1 but when I dialed this it acted on the right hand side, ie the same side as the original custom morph.

Like I said I'm sure it's something I'm doing wrong because I'm tired and not thinking clearly, but any advice on where I'm going awry would be much appreciated. Meanwhile I'm going to go get some sleep and hopefully be more alert tomorrow.

Cheers

Ian

I'm certainly too tired to understand that! LOL!

CHEERS!

Sorry, I should have posted some instructions for the script. 

The "no match" results should indicate that there are vertices which aren't symmetrical, in your base mesh.  The listing only comes up when vertices which aren't centered on X fail to find a matching vertex on the other side.  Or if the central line of vertices isn't zeroed on X, possibly.

All functions in this script use a simple lookup table to match vertices, and don't hunt through the mesh looking for a best match, if there isn't a match found at coordinate -X,Y,Z.  The script which requires the Spanki extension can handle that, however, correcting small deviations or asymmetries.

Mirror copies one side to another on an actor with a base geometry which is centered on X.  It creates a symmetrized morph which includes the source morph.  It sounds like your results could have been generated if the script failed to find any matched vertices on the left side of the mesh, so you just ended up with the source morph with no mirrored deltas mixed in.  (Mirror will also produce odd results if, for instance, you use "+x to -x" and then select a morph which is on the right side of the actor (or vice versa for left).)

The Flip function would actually be the function you'd want to use.  It will completely flip the morph on X, producing a morph with both sides of the original reversed.  Flip also flips a morph between right actors and left actors, such as with the thighs.

I guess the terminology is a bit confused in my usage.    I was thinking of the Mirror Morph utility, which only copies one side of the morph to the other side, when I chose to use "mirror" as I have.  Maybe I should call it "symmetrize".

I may be completely wrong about asymmetries in your mesh.  That's just my first guess, based on the information you've given me.  My apologies if this seems like criticism of the model.    Such is not my intent.  My tests after reading your post could only produce odd results if I used settings which mixed up the +x and -x, as described above.  I could be completely missing something.

Anyway, you'd want to select: Flip  -->  actor  --> morph.

I've been trying to work out a spandex morph for hip2, but I still can't seem to get rid of the camel toe effect when the thighs are bent.  I've done as you suggested this time, creating a simple morph which just pulls the inner labia area of the geometry down and, where necessary, forward.  The deltas have no X component at all: I trimmed them with a Python script.  But it still disagrees with the JCM's.

So I decided to try to create a simple catsuit from the Antonia geometry, with a smoothed hip2 as the only alteration to the base geometry.  There was, predictably, still the same problem, since I was still using my morph as the new base hip2.  Because I was working on a new figure, though, I decided to start tinkering with the bend JCM's on hip2.

That's when I noticed that part of the trouble is that the JCM's are set up for a -60 degree bend.  Beyond that, they slowly begin to distort, and by -90 degrees there's some funny stuff happening.  (Which isn't a criticism!  I think this is inevitable distortion.)  I started working on that, and ended up with what might work nicely as a second bend JCM for Antonia.

On the attached image, the current JCM's are applied on the left.  On the right, I have a second JCM which fires from -60 to -90 and which has been developed at the -90 setting.  It seems to offer some decent correction for inevitable distortion in the main bend JCM.

So I thought I'd suggest a second bend JCM for the thigh for Antonia, and see what you thought of the idea.

As far as the catsuit thing goes, man oh man.  The tiny little polys in the inner labia region are tough to clean up!  :lol:

Nudity-flagged for showing the private region of a naked wireframe mesh.  :lol:

Cage: I was considering a second JCM at some point, but for some reason got caught up in trying to get everything to work with just the one. I agree that this might be helpful, but I guess I'm a little bit afraid of the added complexity. In the knees, it was pretty much unavoidable to have multi-stage JCMs, but for the hips it's only quite extreme poses that don't work so well. Still, I might yet reconsider. :laugh:

Please bear in mind though that anything more than a -100 xrot or so on the thigh is unrealistic for an actual human. To get the thighs up any higher, one needs to include some hip2 action.

One thing I'm a bit worried about is the proverbial fact that perfect is the enemy of done. We could keep fixing problems in the rig til kingdom come ("language" ticked for gratuitous religious reference), but in the meantime no one can confidently rig any clothes for Antonia since they don't know what might yet be changing.

Quote -
One thing I'm a bit worried about is the proverbial fact that perfect is the enemy of done. We could keep fixing problems in the rig til kingdom come ("language" ticked for gratuitous religious reference), but in the meantime no one can confidently rig any clothes for Antonia since they don't know what might yet be changing.

Okay.  I hadn't thought of that.  :lol:  I think I came in late to the project and I'm still a couple of steps behind, in some of my thinking.    I guess I can soup-up my own copy of the figure with things like this.  Heh.

 Ya know,guys...I really hate to point out the obvious....but REAL women in catsuits get camel-toe-effect when moving around in a cat suit!  Cage, you might be trying to fix something which isn't broken.:laugh: Just my 2 cents.

Quote - I guess I can soup-up my own copy of the figure with things like this.  Heh.

I encourage that quite vehemently. That's why she's under a CC licence.

Edit: What I mean is, you can soup-up your copy and then share the results (if you want) without any special encoding tricks. Of course anyone can built their own versions of, say, V4, but the sharing is the slightly inconvenient part.

@JOELGLAINE: la-la-lala, I can't hear you. holds hands over ears

Quote - One thing I'm a bit worried about is the proverbial fact that perfect is the enemy of done. We could keep fixing problems in the rig til kingdom come ("language" ticked for gratuitous religious reference), but in the meantime no one can confidently rig any clothes for Antonia since they don't know what might yet be changing.

And every joint-slaved valueParm or targetGeom is another potential groan from clothing makers.  When doing JCM for underlying figures, one only has to worry about the looks.  When doing JCM for clothes, one has to worry about the looks and poke-through.

Quote -  Ya know,guys...I really hate to point out the obvious....but REAL women in catsuits get camel-toe-effect when moving around in a cat suit!  Cage, you might be trying to fix something which isn't broken.:laugh: Just my 2 cents.

This is a good point.  :lol:  It isn't a bad thing, really.  I'd just like to be able to control the effect and only use it when I want it.  I need to start thinking like odf.  Perfect is the enemy of Done.  (Feels like a Yoda quote, sorta....)

Aanascent: would it help if certain JCMs could be turned off selectively? I was considering that, but am not sure it would really be that useful, because often an important part of the JCM action is to remove unsightly bulges.

Quote - Aanascent: would it help if certain JCMs could be turned off selectively? I was considering that, but am not sure it would really be that useful, because often an important part of the JCM action is to remove unsightly bulges.

I made Stephanie Max so you could inject JCM in just the areas you wanted.  But, it wasn't optimal.  By not forcing clothing makers to deal with the JCM, it didn't progress Poser human figure development much.  IMO, better to find the right balance between perfect bending and clothing makers not supporting Antonia due to complexity.  This seems to be the direction you have gone.  I was just posting to back you up.

FWIW, V4 is not all that easy to clothe.  She was supposed to be easier than V3, but I don't think that's the case.

At the very least, give Antonia good base JCM (which she seems to have now).  Then if you want, create additional JCM for when she is unclothed, make it injectible by region for topless, bottomless, sleeveless, etc....  Leaving the additional JCM as an injection add-on allows you to get a final build ready, without slamming the door on improvements to bending.  Like I was saying, she looks pretty darn good now.  Human figure JCM can become an addictive quest for perfection.  If you aren't careful, you'll someday find yourself still working on it while DAZ releases a new Victoria that includes Antonia's better features (like the hip configuration, IMO).  Then you'll be playing follow the leader.