Body part visibility sounds like a good idea, although it would not be so easy to just load a material collection or mat pose to switch stockings in and out.

Why would you want separate JPs? Don't you want the stocking vertices to move exactly where the toe vertices moved to?

No reason for the JPs... I'm just not clear how much separate toe rigging there is now, big toes saparateand other toes grouped I guess? Of course you are right about MATposes. Maybe we should leave odf alone about this and someone else can create a geometry switch INJ for stocking toes and maybe socks, high heel, shoes etc.

Oh - I'm all for leaving odf alone. Just wanted to get the idea on the table - in the queue, so to speak.

Also, sometimes, what seems hard at first turns out upon some reflection is easy. I get that all the time at work. I tell somebody "that's hard - we're not doing it", then the next morning I wake up and how to do it easily pops into my head. Must have worked on it in my sleep. But I wouldn't have if the request didn't come up.

I didn't mean anything vis you BB :) I just get the feeling odf is anxious to get this wrapped up.

Quote - Oh - I'm all for leaving odf alone. Just wanted to get the idea on the table - in the queue, so to speak.

Also, sometimes, what seems hard at first turns out upon some reflection is easy. I get that all the time at work. I tell somebody "that's hard - we're not doing it", then the next morning I wake up and how to do it easily pops into my head. Must have worked on it in my sleep. But I wouldn't have if the request didn't come up.

That's quite true. I've had that happen many times as well.

The thing is that there is not one, but three toe groups, and we have JCMs in place to get the curling right. So it's a little bit more complicated than what you described, but I imagine we could indeed add the toe cap geometry to those groups. For some reason I thought we'd need geometry switching or something like that.

Anyway, I think I can make the geometry fairly easily,  and I'll throw in some masks to drive this and support stocking texturing in general. I don't really have a workflow set up for morphing and particularly making JCMs at the moment, so I'd leave that part for later unless phantom3D feels inclined to do it.

Quote - > Quote - Oh - I'm all for leaving odf alone. Just wanted to get the idea on the table - in the queue, so to speak.

Also, sometimes, what seems hard at first turns out upon some reflection is easy. I get that all the time at work. I tell somebody "that's hard - we're not doing it", then the next morning I wake up and how to do it easily pops into my head. Must have worked on it in my sleep. But I wouldn't have if the request didn't come up.

That's quite true. I've had that happen many times as well.

The thing is that there is not one, but three toe groups, and we have JCMs in place to get the curling right. So it's a little bit more complicated than what you described, but I imagine we could indeed add the toe cap geometry to those groups. For some reason I thought we'd need geometry switching or something like that.

Anyway, I think I can make the geometry fairly easily,  and I'll throw in some masks to drive this and support stocking texturing in general. I don't really have a workflow set up for morphing and particularly making JCMs at the moment, so I'd leave that part for later unless phantom3D feels inclined to do it.

No problem, I expect to likely do some fixing once people get to testing and find problems with the rig & JCM's, so I'm still on board till she's finished. Plus I still need to hide the JCM's once it's all tested and ready to go.

Cool, phantom3D! I'll send you an updated mesh as soon as I find time to make those toe caps.

In other news, I'm so intensely frustrated with every UV mapping program I've seen that I'm really, really tempted to write one of my own. That's obviously not a tiny project, and I'm kind of wondering if it would be possible in the long run to generate some money with it. This stuff is obviously a bit off-topic here, but maybe I can get a quick show of hands: if I started a thread, say, "The UV mapping program you always dreamed of but were afraid to ask for" in a more appropriate forum such as 3D Modeling, who would be inclined to participate in it?

ME!  I would worship you as a god forever if you could make a UV mapping program that could actually create some flat UV maps.  I know it's probably impossible but at least something approaching flat without all the blood sweat and tears that come with all the others I have tried.

odf, you asked who the model reminds people of. Not the body, but once BB got going with the skin shaders, I thought Gwyneth Paltrow pretty much immediately.

Anyway, this is a very exciting project. You and Phantom and BB are all doing an absolutely fantastic job. Really nice to see a female model that deals with shoulders/underarms properly.

Quote - That's obviously not a tiny project, and I'm kind of wondering if it would be possible in the long run to generate some money with it.

Well the best UV mapping tool is probably headus UVLayout, but it's not cheap. There's also free apps like Roadkill and of course the tools in various modellers. If you make something that can compete in terms of features and/or price, i'm sure a lot of people would buy it.

Quote - > Quote - That's obviously not a tiny project, and I'm kind of wondering if it would be possible in the long run to generate some money with it.

Well the best UV mapping tool is probably headus UVLayout, but it's not cheap. There's also free apps like Roadkill and of course the tools in various modellers. If you make something that can compete in terms of features and/or price, i'm sure a lot of people would buy it.

Yep, UVLayout is the one that doesn't suck. Everything else I've seen, even in high-end software that goes for 10k a license, was pretty pathetic. You see, I have some professional experience optimizing mesh geometries, so I think there's a good chance I could do better than UVLayout's algorithm. What I'm not so sure about is whether there's actually a big enough market. You need a decent GUI and lots of different tools and options for a program like that to be useful, which obviously takes some effort. But if I could figure out what people are missing the most in available software, I might be able to start small, maybe with something that would let people map typical clothes really easily, and build up from that.

Anyway, I'll go and start that thread in 3d modeling later tonight. I hope they're not to fussy over there about subjects with a potential commercial aspect.

 odf et al— nice work!

One suggestion— with the "should the lips have a separate material" question. I find it easier to just modify the underlying color of the lips then to change it on the texture or use a lips-only texture.

I like the fact that someone is making a "normal" human bean. Model she is — just not one that Hollywood would consider. I'm actually dialing a V4 morph right now with a similar body shape, based on a real person. Realism is one of the big things with me.

Facial features and expressions are a must. V4 is good. But above good is better! Her expression morphs are generally fakish (did I make that up?). I am interested in "3d portrait photography" and as such accurate facial expressions are a must.  

If I could help in any way, let me know. I can't help a lot, being a college student and all, but I'll do what I can. Maybe a set of initial poses? Search "Rancor Poses" in the freebies to see my first bitty upload…

At this point it sounds like she will be free. Well and good. If however you find that you need to take payment, that's fine too. Just don't price her as high as some of the other figures! I only have V4 Complete because I got her on sale for $10.

Antonia's shoulders are excellent. V4 is lacking glaringly there. If there is one change v4 needs, it is the shoulders.

Initially I thought her shnoz a bit large, but not now.

Sorry for the long post— had to catch up with everyone else.

Again, nice work.

Quote - > Quote - I had always thought the figure was based on Maria Callas...

http://www.divasthesite.com/images/Maria_Callas/Maria_Callas_28.jpg

Well, not intentionally, but the resemblance is stunning.

Keep them coming, guys.

Persis Khambatta from the first Star Trek Movie?

odf: I'd be HIGHLY interested in a better UV mapping tool! Right now, I'm using a combination of 3DS Max integrated UV mapper and UVMapper Pro by Steve Cox, and what I sorely miss in both programs is freeform deformers.

Mapping an organic object in a way that is both stretch--free and 2D texturer friendly is almost impossible, but for the items I make it should be quite possible - with freeform deformers.

svdl: I'm not sure I understand what you mean by freeform deformers. In UVLayout and Wings3D - the programs I've used the most for UV mapping - you can select some elements and move them around in UV space with a falloff radius applied. Is that what you mean, or were you thinking of something more sophisticated?

At any rate, you can find the thread I started here.

I hope they will not be too much of a pain to rig.

Quote - Quick toecap render. Have to find an empty spot for the UVs and tweak some vertices, then they're done.

I hope they will not be too much of a pain to rig.

Very nice matching of the toe contours.

Were they a pain to rig?

Cheers,

Rod

 Having a toe cap is a good idea, much better than having a "sock" morph for the foot like the Daz Mil 3 characters had.

Yeah nice job odf. I wanted to get back to work on this but I haven't had time.

I did some work on procedural eyebrows. I still haven't worked out all the math, but I think it can be done.

How are you UV mapping the toe cap? I think it would be most convenient if it UV area they occupy is identical to the UV area of the actual foot. That way, we can mask both with great precision using the fact that the UV coordinates for the common areas are identical.

What I"m thinkin is that for the rest of the leg, I'll be using the nylon shader from my previous work, that combines skin and nylon in one material. Where the toe cap takes over, the skin shader is pure skin, and the toe cap is pure nylon. If the UV's overlap perfectly, I can just smoothly blend the two so that the transition is invisible.

Or - if somebody is doing this with images, they can make a transmap where the toe cap should become invisible and let the double skin shader of the foot take over.

See you in a week. I'm off to ski.

So, I grouped the toecap, and it follows all the bends nicely. The only problem is that when the big toe moves up or down independently of the others, the fabric doesn't stretch the way it should and there's some nasty poke-through. But that was to be expected and should be fairly easy to fix via JCMs. I'll send the new mesh to Phantom3D once I'm done with the UVs and see what he can do.

Quote -
How are you UV mapping the toe cap? I think it would be most convenient if it UV area they occupy is identical to the UV area of the actual foot. That way, we can mask both with great precision using the fact that the UV coordinates for the common areas are identical.

What I"m thinkin is that for the rest of the leg, I'll be using the nylon shader from my previous work, that combines skin and nylon in one material. Where the toe cap takes over, the skin shader is pure skin, and the toe cap is pure nylon. If the UV's overlap perfectly, I can just smoothly blend the two so that the transition is invisible.

I wasn't originally planning to make the UVs match precisely, but I agree that it would make texturing much easier. I'll see what I can do.

Semi-relatedly, I was thinking about the best way to texture second skin stockings. My thoughts were as follows: since stockings are elastic, texture alignment is more important than avoiding distortion. In fact, some distortion is actually desirable. So one would basically want the legs to map as a UV rectangle and do something clever with the foot and heel so that they map as much as possible like real stockings would. Putting a map like that on my figure's skin and making it match were the legs meet the body would be very tedious and probably not ideal for mapping skin, anyway.

So I thought what if I provided a separate map just for the legs that would allow for easy texturing of stockings and also a script that would convert textures made using that map so that they could be applied to Antonia's actual mapping? This would also allow me to very easily produce masks for use with procedural texturing.

The same strategy could be used for other second skin items such as gloves and maybe even bra caps. I'd try to write the script in Python so it would be easy to install and use, but if it gets too slow and/or tedious I might have to switch to Scala (which would produce a 4MB-ish binary but would basically run on every platform that supports Java).

To quote myself:

Quote - I wasn't originally planning to make the UVs match precisely, but I agree that it would make texturing much easier. I'll see what I can do.

Yep! That's one of those things that sound easy and should be easy, but in practice, with the tools at hand, turn out to be incredibly painful. So I'll leave that for now and see if I can fix it later using my "alternate UVs conversion" idea. I'll put the new mesh with the added toecaps in my file locker once I've done some final cleanups, and if someone with better tools, more experience of more patience would like to try their hands, they are very welcome.

I thought that you were going to make the toe caps starting from existing geometry, including the UV maps already on them. See what I mean? I would have thought you loaded Antonia with her UV map assignments on her vertices, and you cloned part of the foot to start the toe cap. Which would mean that the UV mapping and vertex positions would start out identical.

Quote - I thought that you were going to make the toe caps starting from existing geometry, including the UV maps already on them. See what I mean? I would have thought you loaded Antonia with her UV map assignments on her vertices, and you cloned part of the foot to start the toe cap. Which would mean that the UV mapping and vertex positions would start out identical.

That would be a valid point if UV mapping programs were able to perform the trivial task of reading an .obj file with existing UVs without messing everything up. The ones at my disposable, except Wings3D, can't do that, and Wings3D is horrible at relaxing UVs.

I'll try this once again with UVLayout. I'll have to scale back manually because UVLayout refuses to accept the existing scaling factor on UVs it hasn't produced. But maybe I can produce something reasonable nonetheless.

At any rate, the current version is in my file locker now.

Hmm, I exported only the feet and toecaps this time, and although I have not the faintest idea what UVLayout was thinking when it garbled the UVs, they were recognizable enough that I could cut and paste them back together the correct way. So I finally have toecaps that match the feet both vertex-position- and UV-wise for the first four rows.

Yay! (I guess.)

That was way more painful than it should be. It's close to bedtime now, so I'll puzzle the pieces back together tomorrow and upload it then.

A word of warning:

As you can see on the image, the UVs for the right foot (her right) match precisely, whereas for the left foot they don't. That's because although the UVs on the left and right toecap are mirror images of each other, the ones on the feet aren't. That's what happens when one is too cheap to buy software and tries to do serious work on demo versions. :biggrin:

I'll fix that later when I've augmented my toolbox a bit. I could do it in Wings3d, but it would involve some rather awkward mesh surgery which I'd like to avoid. The "correct" solution is of course to copy the UVs from one chart to another, and I'll do that as soon as I can.

At any rate: where precision is important, use the right foot for testing, not the left one.

Oh, and in regards to my previous post: I think in all fairness to UVLayout I should point out that the messed up UVs I saw were rather the result of a joint effort by Wings3D, UVLayout and myself. When you change the topology of a mesh that already has UVs, Wings tries its best to produce a mapping for the added vertices, but unsurprisingly that can lead to a bit of a mess.

Will upload the new version tomorrow.

since the object can be downloaded i have started some idea of clothes

Quote - i go away two weeks and when i come back i see some nice progress on the Antonia project

since the object can be downloaded i have started some idea of clothes

Cuteness!

That looks like a deceptively simple yet beautifully made model.  As you've made it available - if you don't mind I'd like to download it and take a closer look.  I'm just making my first humanoid figure.

Antonia is already available? Did I miss where? I thought it was only mesh & texture templates so far....

Yes, that's what I meant - in odf's file locker.

Don't think it's a release yet, just the mesh etc.

I just wanted to look at the mesh - as I'm doing some organic modeling myself.

rjjack: That's a pretty dress. I'm glad she doesn't have to go naked anymore.

Fran: Of course I don't mind. Just remember it's only a preview version at this point.

timag: You're correct. So far it's just the mesh and the templates. It shouldn't be too long though until we can put up the complete thing for beta testing.

Quote -
I just wanted to look at the mesh - as I'm doing some organic modeling myself.

Hmm, in that case I should probably upload the original lo-res mesh for you.

ooo!  low res mesh?  the shape's not changing at this point, right?  i just ask because running some cloth sims on a cube i realized how much of my slow down was dealing with the V4 mesh as the colliding object.   i was just thinking  it might make things a lot faster to have a good low res mannequin.

Quote - ooo!  low res mesh?  the shape's not changing at this point, right?  i just ask because running some cloth sims on a cube i realized how much of my slow down was dealing with the V4 mesh as the colliding object.   i was just thinking  it might make things a lot faster to have a good low res mannequin.

As for Antonia, I do all the sculpting on the low-poly version and then subdivide once. So the shape should be pretty much the same, if somewhat smoother.

Speaking of V4: incidentally, I tried to do a reverse subd step on V4 the other day, and to my slight surprise it worked. Which means that apparently V4 was modeled at a lower resolution and then subdivided for Poser just like I did with Antonia. That's probably common knowledge, but it was new to me.

Anyhow, I now have a 15k poly V4 sitting on my hard drive. Obviously I couldn't share that with people without having DAZ go ballistic on my arse. But, you know, if you got your hands on a program that does reverse subd, you could do it yourself.

Quote -
timag: You're correct. So far it's just the mesh and the templates. It shouldn't be too long though until we can put up the complete thing for beta testing.

That would be awesome... I'm really looking forward to seeing her and putting her through her paces, especially given the conversation here regarding rigging.

I read somewhere tht there is already a low poly version of V4 distributed by DAZ!?

Quote -
Anyhow, I now have a 15k poly V4 sitting on my hard drive. Obviously I couldn't share that with people without having DAZ go ballistic on my arse. But, you know, if you got your hands on a program that does reverse subd, you could do it yourself.

Actually, if you RTE-encoded the new lowrez V4 with the normal V4's obj, that would ensure that only people who has V4 could use it. And as far as I know, RTE encoding is kosher with Daz.

IF it is possible (and legal, of course) I would LOVE a low rez V4. I'm already using the M3RR and V3RR models a lot more than their full high rez counterparts. And aV4 that doesn't bog down the computer would be a godsend!

TrekkieGrrrl: The reduced rez V4 is already available with the DAZ|Studio download as a 17k version, as well as a 4k, 2k and 1k version.

Quote -

Anyhow, I now have a 15k poly V4 sitting on my hard drive. Obviously I couldn't share that with people without having DAZ go ballistic on my arse. But, you know, if you got your hands on a program that does reverse subd, you could do it yourself.

You could RTencode it against the original V4 .obj. 

I've never really looked at the V4 version that ships with DAZ Studio-is that a low-poly version?

Cheers.

Rod

But you have to DL DAZ Studio or is there a seperate DL for these lower res figures? Also, do you know if the extra creature pak and character addendums and clothes work on the low poly versions?

Also, for anyone here who has V3 & V2....how does the V3RR compare in size to a full res V2?

Quote - TrekkieGrrrl: The reduced rez V4 is already available with the DAZ|Studio download as a 17k version, as well as a 4k, 2k and 1k version.

Well, there you have it.

To those who suggested RTE: I'm pretty sure extracting the low-poly original of a figure would be considered reverse-engineering. It's a bit as if I offered to give you the source code of a program that you bought as a binary version. Not remotely the same as distributing a remapped or otherwise modified figure based on the original mesh.

Besides, if DAZ already offers a low-res version, what's the point? I just thought it was interesting that V4 was modeled at lower res and then subdivided for Poser as well.

oh, i knew both of those facts (low res V4 in recent D|S, subdivided for use in Poser).  but i don't want to install the new D|S and all its content.  i don't think i can do reverse subdiv.  besides, i'm not that concerned about V4.  i make stuff for V4, and it's automatically unimportant because i'm one of hundreds if not thousands.  it's fun, but if i hit a wall, it's no big deal.  if i make stuff for another figure, and i should figure out how to do it right because there are probably people who could benefit from what i've done.  lower res object means i can up quality on the cloth sim without crippling my machine.

Okay then! As mentioned earlier, I'm trying to keep the low-poly version of Antonia fully compatible. The UV-mapping, grouping and rigging has been done on the high-poly one, so there'll be some backporting necessary. But I hope it will go fairly smoothly.

oh, for me, i wasn't concerned about anything but shape.  i mean, those sound useful and great, but i'm not asking for you to make your life crazy.  in terms of developing clothing, all that isn't necessary.  though if you do the rigging stuff, it will mean that people can do strand hair sims easier, too. 

Oh, so you only want the low-res version for draping? That's easy, then.

Okay! So the version with the completely fixed toecap UVs is up now.

Hopefully, the next upload will be the complete figure.

Yes!  That's how I'm modeling - do it all in low res, then smooth to see how it's looking, and undo the smooth to continue modeling, planning for a final smooth to finish.

I would very much appreciate being able to see your low res version of Antonia.  She looks nicer than V4 to me.

(by smooth in Wings3D I mean subdiv)

Hi ODF

Nice progress on Antonia. Will the toecaps be parts that can be switched off? and perhaps the brow as well.

Quote - Hi ODF

Nice progress on Antonia. Will the toecaps be parts that can be switched off? and perhaps the brow as well.

Actually, I might leave that one as an exercise for the beta testers. :biggrin:

More seriously, at least two ways of hiding these parts come to mind.

The brows and the toecaps each have their own material group, which can be made completely transparent and hence invisible in renders. I imagine the final figure will come with a default material setup file that will make the toecaps invisible and use either a procedural or a hand-painted transmap on the brows.

If that's not enough, there's always the possibility to make some quick morphs to hide the brows within the head and the toecaps within the foot. The way this interacts with custom morphs will not be completely foolproof. If someone comes up with, say, a morph that turns the feet into hooves, there's a chance the hidden toecap geometry will resurface. But I guess that's not too serious a problem.

Speaking of caps: it just occurred to me that I could make a genital cap in much the same way as the toecaps. That could work well with - possibly see-through - second skin underwear and have the additional advantage of providing a slightly more modest default presentation of the base mesh. Anyone interested?