Quote -
I really hope that I am able to catch up with everything as fast as possible! Leo wants me to tell you that he found some boo-boo on the freebie page (there was an error in the Java script that made uploads disapear from their categories) and since we are back online now he was able to fix the code. This made Antonia slip into data-Nirvana, he's very sorry but put her (and everything else) back into her place now.

Welcome back, SaintFox! Glad you're feeling better.

Yeah, I meant to tell Leo about that problem, but decided it wasn't so important that it couldn't wait a little.

Cage and lesbentley: Thanks for making me aware of the foot problems.

I had noticed the problem when bending the foot upward. As it turns out, part of it was due to an excessive amount of bulge. There's also some deformation of the ankle bones, but I think that's bearable up to about 30 degrees or so, which is as much as the foot can and - IMO - should do, anyway. Les, could you try and reduce the 'right neg' bulge to 0.05 on the foot bend and tell me if that looks reasonable for the pose you had in mind?

I also think it's kind of okay-ish to have ankle bone deformation for larger values of twist or swing, because those would be kind of unrealistic. But that distortion at the foot-instep boundary (that apparently only happens when bend, twist and swing are used all at the same time) is definitely weird and not cool. My suspicion is that I've been getting a bit too close to the actor boundary with the inclusion/exclusion angles. I'll see what I can do to fix this.

General remark: at this point, I tend to favour solutions that do not change the center for an actor, because that would very likely mean that I'd have to re-sculpt the JCMs for that bend. If I just change the falloff zones, inclusion/exclusion angles or even bulges, all I have to do is re-load them.

I like to call this my strategic pre-launch laziness. :laugh:

Okay, I found the culprit: using inclusion/exclusion angles for the foot twist was no good. It needs falloff zones instead. That's cool, since there are no JCMs for that channel. With a little luck, I should be able to set thing up so that I won't need any, either.

Now, where is that relieved emoticon?

Again, many thanks to Cage! :thumbupboth:

Quote - Re the JMC dials. I'd suggest visible, but in a rolled up group (collapsed 1).

Yes, that sounds good. Of course, people with older Posers (I forgot whether dial groups were introduced in P5 or P6) would then still see those dials.

Quote -
Unless something has changed in recent versions of Poser, reading the deltas in a targetGeom channel when the figure loads will force the dial to show, even with a normal 'hidden 1'. You can partially get round that by by placing 'hidden 1' after the deltas, but it won't survive a save and reload.

I'm not sure I understand what you mean. I don't recall ever seeing a dial for a targetGeom that used "hidden 1". I just tested this again with Antonia-125.cr2 and also with my modified cr2 that has the new JCM control channels in the body. None of the hidden channels show up in Poser 6 or Poser Pro 2010.

Question regarding the JCM get lost when Bone center changes?
Why is that exactly? Does the body part move to a different position so that the sculpted JCM no longer fits the new position?

Quote - Question regarding the JCM get lost when Bone center changes?
Why is that exactly? Does the body part move to a different position so that the sculpted JCM no longer fits the new position?

Yep, that's what I meant. As long as the body part doesn't move around, I can just reload my sculpted mesh and thus compute the new deltas to go with the modified JPs.

Come to think of it, even if the body part moves around, I could probably still apply the existing deltas, export the result and sculpt away any problems that might have turned up. I haven't tried that in praxis yet, but I imagine it might still be less work in many cases than redoing the whole JCM from scratch.

SaintFox,

Love the last image of Antonia you posted, the back view. I think you should do a bigger version and Post it in the Gallery. I will take as look at your Anna morphs, and report back when I have something to say.

odf,

I'm a bit stretched for time at the moment, but I will test those foot tweaks you mentioned, hopefully tonight.

SaintFox, You made one small mistake with your INJ files. PMD injections can create new custom channels in the figure with unique internal names. Ordinary INJ files (which yours is) can only inject stuff into a channel that already exists in the figure.

Your AnnaBody-INJ.pz2 file is trying to inject into a channel with the internal name "targetGeom AnnaBody". As no such channel exists in the figure, this does not work. You need the change the internal name in the pz2 to be one of the PBMCC_##  channels. What you have at the moment is this:

                targetGeom <span style="color:rgb(255,0,0);">AnnaBody</span>
                        {
                        name AnnaBody

You should change it to this: targetGeom PBMCC_04 { name AnnaBody

The number I used above "04" is only an example, you can use any number between "00" and "09". All Antonia's body parts have channels with that range of numbers, the head has even more channels to chose from, "00" through "49". As the head morph has a separate msater FBM dial in the BODY, you should use a difrent number for the head morph. You will also need to edit the REM files to use the PBMCC_## channel names.

One additional point. Whilst it is not essential, it is a good idea to leave the figure number out of any slaving code that is being injected from a pose file, this ensures that the slaving will work correctly if you have multiple figures in the scene. So use:

                        Figure
                        BODY
                        PBMCC_04
                        deltaAddDelta 1

In prefference to: Figure 1 BODY:1 PBMCC_04 deltaAddDelta 1

I have fixed your files, so if you would like to IM me with an email address, I will send the fixed version to you.

Quote - General remark: at this point, I tend to favour solutions that do not change the center for an actor, because that would very likely mean that I'd have to re-sculpt the JCMs for that bend. If I just change the falloff zones, inclusion/exclusion angles or even bulges, all I have to do is re-load them.

I like to call this my strategic pre-launch laziness.

Avoiding backtracking at this stage is certainly desirable, I would think.  I wouldn't call it "laziness", myself.  What with "lazy" being kind of a "kick me" term, and all.  :lol:

I'm glad the foot handling can be easily corrected!  :woot:

Looking at the foot has made me wonder: why does the Twist parameter point on Z, and not Y?  I've seen this in other figures, too, and it's always baffled me a bit, given what the twist deformer actually does.  Does this rotation axis ordering for feet have something to do with the Walk Designer?

Les, you're my hero!! :laugh: As more as the heat here makes me dizzy in the head and I am hardly able to understand what I made wrong (but that's what notepad is made for - as soon as we reach normal temperatures I will read your advice closely and try it out!). I will send you my mail-adress via pm and of course I will post a large version of the image I posted!!

I also agree I'd prefer to see JCms in a collapsed group, instead of hidden.

That way its not readily available for new users, but IS readily available for advanced users.

SF, glad back.

I'm on vacation until next week and am using my brother's comp so it will be awhile before I am back to posting.

Thanks a lot for the warm welcome-back!!

I take the chance to say thank-you to Les, the morphs work fine now AND I am able to inject the blink-morph, LidSynch and whatever else my heart desires as well without creating a Pinochio-nose or a Jane Mansfield breast-shape! I am able now  to update the Anna-morph in some minutes on the freebie-site.

odf,

Reducing the right neg bulge does seems to fix the problem where a rige develops on the ankle. I tried a few diffrent values with the foot bent at -30.0. A  right neg bulge of 0.07 seemed about optimal to my mind. I think looking reasionable at -30.0 is the point to aim for with foot bend.

Quote -
Looking at the foot has made me wonder: why does the Twist parameter point on Z, and not Y?  I've seen this in other figures, too, and it's always baffled me a bit, given what the twist deformer actually does.  Does this rotation axis ordering for feet have something to do with the Walk Designer?

Which figures have you seen this in? I just had a quick look at Alyson and V4, both of which use the YZX rotation order for the foot, like Antonia does, but unlike Antonia use the name 'twist' for the Y axis, which is in agreement with Poser's internal conventions.

With that axis order, the Y axis is the one that gets the twist deformer. I just decided to change the external names because the Z axis is really the one that twists the foot, even if Poser doesn't see it that way. The Y axis twists the shin.

Quote - odf,

Reducing the right neg bulge does seems to fix the problem where a rige develops on the ankle. I tried a few diffrent values with the foot bent at -30.0. A  right neg bulge of 0.07 seemed about optimal to my mind. I think looking reasionable at -30.0 is the point to aim for with foot bend.

Excellent! I confess I hadn't really looked for the optimal value for that bulge before I posted, just noticed that reducing it seemed to help. I'll try 0.07, as you suggest.

I'll post a pose file with the new JPs for the foot bend and twist as soon as I find the time. Hopefully, this will fix the last major boo-boo in Antonia's rig.

Quote - Which figures have you seen this in? I just had a quick look at Alyson and V4, both of which use the YZX rotation order for the foot, like Antonia does, but unlike Antonia use the name 'twist' for the Y axis, which is in agreement with Poser's internal conventions.

With that axis order, the Y axis is the one that gets the twist deformer. I just decided to change the external names because the Z axis is really the one that twists the foot, even if Poser doesn't see it that way. The Y axis twists the shin.

I just checked V1, and the "Twist" dial handles the Z axis rotation, but the twist deformer for the actor is actually vertical.  Perhaps what I've seen before was only a matter of naming conventions.  The same is true of Antonia, when I check her in a zeroed pose.  So now I'm confused. 

I checked Antonia's foot joints earlier while she was posed.  I guess the posing pointed the twist deformer axis on Z, rather than Y.  Hmm.

So.  Umm.  Never mind.  :lol:

:blushing:

OT: Hey, let's hope the forum software doesn't store the page number in a single byte. :lol:

 I'm just hoping the software doesn't melt down when the page count on the Antonia thread reaches 300! It's acting weird. 

After months I finally starting getting ebots again out of the blue. I'm completely out of touch with what's going on here, and I'm still busier "than a one legged man at a butt-kicking contest!" :laugh:

I'm going to be insanely busy the next few weeks, but after the last of court things about my SSI, I should be back to testing and commenting and such. I'm ready for a vacation.:lol:

Quote - I'll post a pose file with the new JPs for the foot bend and twist as soon as I find the time. Hopefully, this will fix the last major boo-boo in Antonia's rig.

Good, I'm keen to try them!

Quote - > Quote - I'll post a pose file with the new JPs for the foot bend and twist as soon as I find the time. Hopefully, this will fix the last major boo-boo in Antonia's rig.

Good, I'm keen to try them!

I can do one better. Just download Antonia-126.cr2 from github. That's the latest CR2, which has the fixed foot JPs and the new control parameters for all the JCMs. The file references an Antonia-126.obj, but you don't need to download that. It's the same as Antonia-125.obj, so you can either make a copy and rename it or change the references in the CR2.

Edit: Needless to say, that is not the final word on Antonia 126. But there's no reason not to play with the snapshots I'm committing to github. I take care not to make any messes in the files I put up there.

This may or may not be another joint situation which could be improved.  If the thigh is twisted enough, the distortion seen in the attached image can result, at the shin weld.  I haven't tested to see if this only occurs in combination with bend and/or side-side.  I have seen lesser manifestations of this distortion at fairly normal thigh twist settings, however.

I just downloaded 126 and changed the obj info in it to reflect 125.

She loads fine, morphs work and textures work great.

Quote - Will Antonia 126 end up having any changes for the neck JPs?

Probably not. I looked at the changes you suggested, and even though they made the neck look better in some poses, I did not like the fact that the stretch ended up being so uneven. Unless we happen to come across a real solution for the scaling problems - which also happen on the legs and arms - very soon, I'm afraid I'll have to leave things as they are for now and try to come up with some more satisfactory workaround - possibly using SCMs - later. Unfortunately, that would then have to be an add-on, not part of the base figure.

Quote -
This may or may not be another joint situation which could be improved.  If the thigh is twisted enough, the distortion seen in the attached image can result, at the shin weld.  I haven't tested to see if this only occurs in combination with bend and/or side-side.  I have seen lesser manifestations of this distortion at fairly normal thigh twist settings, however.

Thanks! :thumbupboth: That's definitely not right, and it looks like something I might be able to fix.

As usual, I'm not sure whether this counts as a joint situation which needs to be fixed, but here's another case where the bending isn't up to the standards of the rest of the rig.  If the thighs are moved inward, using spread, there's a bit of pinching of the hip2 mesh, at the thigh weld.  I'm not sure how to fix that.

Quote - Oh, bummer.  I understand about the neck, but it means the 125 head-neck JCMs won't be able to work with neck scaling.  :sad:

Yes, that would be a bit unfortunate. I might do some experiments with SCMs pre-126 to see if I can fix this problem. But no promises!

Quote -
As usual, I'm not sure whether this counts as a joint situation which needs to be fixed, but here's another case where the bending isn't up to the standards of the rest of the rig.  If the thighs are moved inward, using spread, there's a bit of pinching of the hip2 mesh, at the thigh weld.  I'm not sure how to fix that.

Thanks! That's definitely not an outlandish pose, so I'll go check that.

By the way, is that pose you showed meant to be symmetric? Because it looks like the deformation isn't the same on both sides. I'm asking because I've noticed that the limits for the foot bend JCM were different on the left and right foot, which made the right foot deform much worse with negative bend values. I was going to check the limits for the other JCMs as well, but haven't got to that yet.

Quote - By the way, is that pose you showed meant to be symmetric? Because it looks like the deformation isn't the same on both sides. I'm asking because I've noticed that the limits for the foot bend JCM were different on the left and right foot, which made the right foot deform much worse with negative bend values. I was going to check the limits for the other JCMs as well, but haven't got to that yet.

The thigh spread pose isn't symmetric, no.  I'd noticed the problem in one of my WIP characters so I opened a fresh copy of 125 and just applied the thigh spread to verify that it was present in the current version.

It's really not a big deal, the neck.  I'll just leave those morphs out of any customization I develop which requires neck scaling.  No rush.  It'll get there.  :thumbupboth:

Cage, I checked both problems you mentioned regarding the thigh JPs, and it turns out that both are caused by inaccurately positioned falloff spheres.

Solving the twist problem was completely trivial. All I had to do was rotate the inclusion sphere a little bit. The "negative spread" one could be trickier, but I'll see. Since there's a JCM on that channel, I need to change the JPs, then reload the morph and see what the result looks like with the thigh is move inward. The thing is, the JCM was made for rotating the thigh outward, but happened to work reasonably well in the other direction, too. But it's not perfect, so the "fix" I have in mind may just end up making trouble somewhere else. I'm debating now whether I should just add another JCM instead for the inward motion and leave the falloff zones as they are. That would allow me more control and might actually end up being less work.

With all these little glitches that turned up, I should probably go and do a systematic check of the falloff zones before Antonia-126. It's easy to assume that the JPs are all fine when the bends look good in normal situations, but in many of the cases that you guys brought up, it was actually very easy to spot the problem once I looked at the spheres in wireframe.

I think for finding all these bugs, Cage and lesbentley should get some kind of special prize, along with phantom3D, SaintFox and bagginsbill. I don't know what it could be, though.

I hope the thigh spread can be fixed without too much trouble.  I was afraid it might require new JCMs.

No prize needed, here.  An even better Antonia is the prize.  :laugh:

(Unless the prize might be someone making some muscle morphs, or maybe a displacement map for muscle definition.  Preferably for the A mapping.  :lol:  I want to try those at some point, but I'm not sure my skill set will extend to that, even with new tools to make it easier.  :unsure:)

You've made repeated references to "recalculating" the JCMs.  What do you mean by this?  Just re-loading the morphs?  Or is there a special program which calibrates the deltas in some special way? 

Edit: Now begins the 2-bit part of the thread, apparently.  Do we need to do anything special?  :ohmy:

2 byte. Figuring that the page counter will make this necessary sometime soon, I'm doing an upgrade to 64-bit.

Quote - 2 byte. Figuring that the page counter will make this necessary sometime soon, I'm doing an upgrade to 64-bit.

So apparently the special thing I need to do is learn me the correct vocabu-mo-lary.  :lol:

Thanks, BB.

Quote -
You've made repeated references to "recalculating" the JCMs.  What do you mean by this?  Just re-loading the morphs?  Or is there a special program which calibrates the deltas in some special way? 

No, I just meant reloading. That does recalculate the deltas, which is kind of a big deal for JCMs. I so wish I had an oracle to tell me exactly how Poser's joint algorithm works so that I could do that myself and not have to use any gorram morph loaders.

Anyway, I think I'll just make a new JCM for moving the thigh inward. That should be fairly easy, and will allow me better control of her crotch area for a number of useful poses.

So my plan for version 0.9.126 looks more or less like this:

• fix the thigh problems you pointed me at
• do a visual check of all the JPs, particularly the falloff zones
• do an automatic check of the CR2, looking for asymmetries and other inconsistencies
• create a shitload of new poses (using troll counting, of course: one, two, many, lots, a shitload)
• add some temporary empty channels in the head for injecting expression morphs into

Quote - No, I just meant reloading. That does recalculate the deltas, which is kind of a big deal for JCMs. I so wish I had an oracle to tell me exactly how Poser's joint algorithm works so that I could do that myself and not have to use any gorram morph loaders.

So theoretically we at home could try changing JPs and re-loading the JCMs, if we export the morph and re-load it onto the figure after joint alterations?  I'm thinking about the neck, and whether there could be a way for me to use the new 125 joint morphs with my custom JPs.

The plan looks good.  That's things yet-to-be-done, rather than a list of what to expect on the 126 figure, right?

For the joint algorithm, I think a few people may have pretty good guesses, but I doubt anyone knows exactly how it works.  IIRC, Kuroyume had to sort of reverse-engineer Poser's joints while creating his Poser to C4D program.  CodeTwister also seems to have recreated Poser joint functionality for Morph Master Pro.  Maybe they could offer some hints? 

I'm not quite sure this could be called a rigging issue, but I'll mention it any way. In most figures the foot actor goes right up to the the toes. That is not the case in Antonia because of the Instep actor. It seems that when the Walk   Designer calculates a pose for the foot, it rotates the foot geometry down until the front of the foot is touching the floor. In most figures this is equivalent to touching the ball of the foot to the ground. In Antonia, because of the shorter foot, by the time the front of the foot is touching the ground, the ball of the "foot" (Instep) is below the ground and the bend angle of the foot is greater than would otherwise be the case.

There might not be any good fix for this, and I think we may just have to live with it, but I thought it worth mentioning so that you are aware of the issue.

The workaround I am using at the moment is to inject a different foot geometry, with one vertex pulled forwards to the toes. This kind of wrecks the foot geometry, but as it is the bottom of the foot, it does not really matter in a walk sequence where bottom of the foot is not usually visible.

Quote -
So theoretically we at home could try changing JPs and re-loading the JCMs, if we export the morph and re-load it onto the figure after joint alterations?  I'm thinking about the neck, and whether there could be a way for me to use the new 125 joint morphs with my custom JPs.

Yes, if you have a morph loader that has a reverse deformation feature, you can do that. At the moment, as far as I know, that's either MorphLoaderPro for DAZ Studio, or colorcurvature's new tool, if you have access to it. You would export the mesh to an OBJ file as it bends with the original JCM applied, change the JPs, delete the old deltas and then reload the morph from your exported versions of the mesh.

Quote -
The plan looks good.  That's things yet-to-be-done, rather than a list of what to expect on the 126 figure, right?

Yes.

Oh my, a lot has been going on while I've been here on vacation. lol.
I hope 126 cr2 will still be availabel or a newer one when I get back home to download.

I've got a few things I want to do on my texture sets so I'll be back soon. :)

Quote - **
**The workaround I am using at the moment is to inject a different foot geometry, with one vertex pulled forwards to the toes. This kind of wrecks the foot geometry, but as it is the bottom of the foot, it does not really matter in a walk sequence where bottom of the foot is not usually visible.

Les, I forgot to thank you for this tip. Does the modified geometry have to be in effect whenever the walk sequence is used, or only when it is created/constructed? Would it be possible to use a morph instead of a geometry switch?

The JCM will be integrated into the base figure for version 0.9.126.

Hello; I'm baaaack...lol.
Just got home today but will be leaving again in a few days.

I was wondering if 0.9.126 will be the official release?

If so I'll be ready to submit a retail character/texture set.

Hi BluEcho!

The official release will be Version 1.0. My plan is that Version 0.9.126 will be official from the neck down. I say "the plan is" because that was already my plan for Version 0.9.125, but then it turned out that there were lots of little problems with the JPs that I hadn't noticed before. At any rate, since I'll be working on the face quite intensively between Version 0.9.126 and the official Version 1.0, I cannot really recommend submitting any retail character sets just yet. Clothes, poses and body shaping sets should be fine, but anything involving the face might be a bit risky.

By the way, I've started going through all the JPs systematically in search of potential problems. So far I've done the head, neck and torso. I've found that the neck could use some minor adjustments, then everything down to and including the hip looked fine, but the hip2 JPs needed some more substantial changes.

I think I'll go through the arms and legs - including hands and feet - first and check the symmetry as well before I start including these adjustments in the main CR2. As usual, I will probably make a pose file for loading all the changes into the current version, so that you guys can have a peek at them in advance.

Quote - At any rate, since I'll be working on the face quite intensively between Version 0.9.126 and the official Version 1.0, I cannot really recommend submitting any retail character sets just yet.

Risky, how?  :scared:  Will the base shape be changing, or even the actual head geometry?

(If the head's base shape is to change, there does seem to be a slight problem with ridges showing up above and below the lips.  If the vertex count or order changes and the shape changes, can a head with the old shape and the new vertex structure be released, to aid in morph transfer?)

(I'm getting ahead of things, right?  :unsure:)

I'm really glad to hear about all of the improvements to the joints.  Antonia's joints are the best I've seen, but I'm eager to see how they can be better.  :thumbupboth:

Quote -
Risky, how?  :scared:  Will the base shape be changing, or even the actual head geometry?

It's mostly that if one makes a character morph, particularly a commercial one, they'll want it to work well with the expression morphs. Since the expression morphs don't exist yet, that will be a bit tricky to test.

Also, if I come across some serious problems while making the expression morphs, there is a tiny, tiny possibility that I might make small changes in the default shape, or rather, move vertices around while mostly keeping the shape as it is. There's an even tinier chance that I might make some changes in the mesh topology for the face. This shouldn't worry anyone. I don't expect any of this to happen, and if it does, there are ways to deal with it. But I'd rather not tell anyone to go ahead selling stuff for Antonia if there's the slightest chance that it won't work properly with the official version.

Quote - Les, I forgot to thank you for this tip. Does the modified geometry have to be in effect whenever the walk sequence is used, or only when it is created/constructed? Would it be possible to use a morph instead of a geometry switch?

I think it needs to be used whenever Antonia is used in the walk designer. There may be some other way round the problem, but if so I can't think what it is. Unfortunately you can't use a morph for this, as the walk designer does not "see" the morphed shape. A better alternative to my crude solution might be to add an extra polygon with transparent material.

I should probably not have sounded so certain about what I said concerning how the foot bend works in the walk designer. It's more of a theory than a proven fact, but it's the only way I can account for what I have observed, and extending the foot does seem to solve the problem. So that's strong evidence that the theory is correct.

sorry but just  tryin to get the bot to send me notifications again, somehow they stopped    :)

I got one telling me about your post! LOL!!

I'll be glad when we get to Antonia V1.

CHEERS!

Regarding my last post about hip asymmetry. I can't find anything that looks wrong in the rig, so perhaps it is just one of those weird things that Poser seems so fond of doing, rather than a problem with Antonia herself.

Quote - Regarding my last post about hip asymmetry. I can't find anything that looks wrong in the rig, so perhaps it is just one of those weird things that Poser seems so fond of doing, rather than a problem with Antonia herself.

I was half suspecting that. At any rate, I will try to come up with a script to test the symmetry of the rig and other potential problems I can think of.

At the moment, I'm half through my visual inspection of the JPs. Besides the ones already mentioned, I found a few more small problems with the falloff zones, mostly the inclusion/exclusion angles. If the influence zone for a bend reaches the boundary between the parent actor and its parent, one gets unsightly crinkles at those boundaries for large rotation values. The adjustments are usually quite small, and I didn't have to change any JCMs for them.

I made one more substantial change so far, namely at the hip2 xrot (Bend), which in Antonia 125 involves a pretty severe deformation of the buttocks that I found intolerable. As a result, that bend now gets a bit ugly at larger values, but is, I think, more usable at realistic values. A new pose I made using the improved hip2 bend can be seen in my latest Gallery submission, entitled Skating Dream.

I still have to do some more of those small fixes I mentioned above on the neck and collars, and maybe check the fingers once more. After that, it's time to get to work on that CR2 checker script. After that, many new poses...

Just checking posts to keep my ebots coming. :)