Quote -
Oooh - good catch, Mike on the stacked vertices. In the script I only weld exact matches and I probably forgot to force those points to 0,0,0, resulting in them not being welded. Any tiny deviation from 0 results in those points moving during the "revolve" operation, resulting in duplicated points.

Thanks.
Actually, if it works OK in Poser, I'd say don't change it. I welded them on my own to see how it would look, and as expected, 32 triangles sharing a common center point looks terrible in SubD. There should be other ways of avoiding any render artifacts in other programs.

Quote - MikeJ: I'm almost willing to bet that those missing points on the rendered UV template have u and v at 1 or ever so slightly above 1, respectively.

Yeah, you're probably right about that. The points are very close to the edges, if not actually touching. In theory you should always back off ever so slightly from the edges.

In theory, those converging triangles should not produce artifacts, since, by definition, they are supposed to be completely black - not reacting to light at all. But I'll change it.

There's no need to even have those polygons there. A simple square at the base of the pupil would work just fine.

Well, you seem to be on to something good here, BB, so I'll say it again, nice job. :-)

Thanks.

I'm wondering something that maybe you guys could answer.

I managed, from within Poser, to load and position two eyes, two covers, and an "EyeFocus" prop that the two eyes point at. I saved these 5 props in one Poser prop file and I can load them back just fine, which makes mounting them on Antonia for testing purposes really easy. However, the eyes are not parented to the head as they were before I saved them.

Is there a hack I can do to the prop file that will make them automatically parent to the head like a smart prop?

If you did in fact use the smart prop option... no idea. I've seen that happen in Poser before and wondered about it too.

It didn't ask me if I wanted smart props. Perhaps because I was saving multiple props? Perhaps because the covers were parented to the eyes, which were parented (each) to the head, and both eyes were set to "point at" the fifth prop? I dunno.

I'll try saving just one eye.

Quote - Thanks.

I'm wondering something that maybe you guys could answer.

I managed, from within Poser, to load and position two eyes, two covers, and an "EyeFocus" prop that the two eyes point at. I saved these 5 props in one Poser prop file and I can load them back just fine, which makes mounting them on Antonia for testing purposes really easy. However, the eyes are not parented to the head as they were before I saved them.

Is there a hack I can do to the prop file that will make them automatically parent to the head like a smart prop?

The reason may be, that there is already some parenting going on within your prop file.
I think, Poser ignores the smartparenting part when one part of a multi-prop group has already
been parented to one of the props within the group.

Obviously, I am typing way to slow... :-)

Antonia is watching you. :lol:

With the eyes looking so real now, I think one of these days I'll have to experiment with strand-based lashes and brows for Poser.

Quote -

With the eyes looking so real now, I think one of these days I'll have to experiment with strand-based lashes and brows for Poser.

Nobody answered me before when I asked this. ;-)

But can Poser use image maps to mask out where hair will or won't be on a "growth mesh" object? How about image maps for hair color?
If it can't, it needs to.

Quote - I've refined the shaders a bit more. This looks pretty real to me.

Yeah I'd definitely say you're onto something there.

Quote - > Quote -

With the eyes looking so real now, I think one of these days I'll have to experiment with strand-based lashes and brows for Poser.

Nobody answered me before when I asked this. ;-)

But can Poser use image maps to mask out where hair will or won't be on a "growth mesh" object? How about image maps for hair color?
If it can't, it needs to.

Hair color for sure. Masking growth, I don't think so. There should be lots of masks for hair. There should be one for hair density, and one for hair direction from the root. And they should work with nodes, so they can be procedural.

Thing is, the hairs are nothing more than poly-lines. You can make them in any app. In fact, once I get squared away with this eye and with the few bits of hardware I need for my room generators, I plan to tackle procedural hair. I know I'm strange, but a hair styling GUI, especially what Poser offers, is unusable to me. I can't work the mouse that well and it's fricking impossible to easily grab the hairs I want to move as a group. I gave up learning to use the hair room. I'm going to program my way out of the problem. Once I have algorithms for various common hair styles, then I can generate tons with no work. In fact, I can probably just set up a random number generator, render a few hundred hair styles over night, examine them in the morning, and keep the ones I like.

Yeah, you do that. Might keep you occupied for an hour or two.

Impressing what went on here while I was away from the computer!!!
Too bad that I am away over the weekend and so will have to wait until I can try out the eyes...

But I have a question anyway. You, BB, asked
Does anybody have a definitive answer as to what the angle subtended by the iris typically is?

I tried my best but I am not absolutely sure what you need to know. Maybe odf or someone else who talks english as good as german can translate to me what your question is and I will try my best to find out. I know several opticans and maybe one of them can give me the exact answer.

About making eyes point to an object - IMHO this does not work well. Every time I see it the figure is more or less crosseyed. Of course there are many people that have a slight strabismus and so you can call this realistic. But as 3d is always a bit idealizing as well I am not so fond of figures showing the effect you can see on many photos of Lucy Liu in a pretty extreme way. I tried to figure out what to do for an artist who likes to point the eyes of characters to the camera but didn't like the crosseyed effect. One possible solution is to point only one eye to the camera or another object and adjust the other eye by keeping a likewise space between iris and eye-corner (the opposite eyecorner for the opposite eye, of course). This may not be anatomically correct but it looks correct.

I have a render to do at the moment but will try to come up with examples later.

bagginsbill: According to Wikipedia, the diameter of an adult human eyes is 24mm. Measure the diameter of your iris and you can calculate the angle. I'm sure it can be foundonline, too, but I was too lazy to search.

SaintFox - I already took into account the cross-eyed problem. I adjusted the "point-at" direction for each eye so that they do not cross. Look at my render above - the eyes are "pointing" at the prop only 32 inches away, yet they do not appear to be crossed.

The word "subtended" is not easy for non-English speakers. It means given an arc, or curve, that forms part of a circle, what is the angle, or fraction of the complete circle, occupied by something. For example, the moon subtends an angle of 1/2 degree in the sky. If you were to place many images of the moon side by side around the sky, after 720 of these, you come back to the first moon, so the moon subtends an angle of 360 degrees per 720 moons, so 1/2 degree per moon.

I am asking what is the angle subtended by the iris, so that I can make it correctly proportioned to the eye itself.

I looked at the render (I am talking about the lower one) and although you may slap my hand for it - it shows exactly what I mean. The iris of the left eye almost "touches" the lacrimal while there is a part of the sclera visible in the left outer corner of the eye. As said: This happens to many people - but looks like a strabismus.

Maybe it's just my personal taste, but I think that when looking aside the eyes should look like this
http://root.kevinmillsphoto.com/images/Eyes/original_eyes.jpg

and not like this (and this photo was manipulated, but the photographer took only care of the color...)
http://www.designnation.de/Media/_Generate/Galerie/l.48d8ba494f633,ice-eyes.jpg

About the subtended angle: Am I right that you ask about the space required by the iris in proprtion to the complete eyeball's surface?

In any case, I adjusted the point-at angle. Here she studies her little point-at prop. She quite literally can't take her eyes off of it. Hehehe. Does it look right?

Also, be aware that one does not have to use the point-at feature at all. It is just something I find convenient for testing, since there is no other way to rotate the eyes together up/down and side/to/side until odf integrates the eyes into the figure.

In answer to your question - yes I'm asking about the proportion of the iris to the complete eyeball surface.

Actually, I think the problem is that I scaled the eyes up to make the iris look right. That forced me to move the eyes closer together than they should be.

I think when I re-build the iris so that the eyeball is the correct size, it will all work out right.

Awwww... this is too sweet!! Now that's the perfect use for the point-at feature - and now lift her arm up

By the way: If I use the point-at option with V4 and then have to repose her I always have problems that her jaw stays where it is, if I use propped hair it stays in the previous postion as well...

I just tried the same with Antonia and there is no problem at all!

And before I forget: I just tried out the new version of Antonia and wow... Joint Control Morphs!! And empty morph channels!! The Injection poses by LaurieA do not work yet, they seem to require different channels. So I loaded some of them manually and as the morphs are very useful I vote for keeping them in the finished version.

Just so you know: Antonia's original eyes don't actually point straight forward, but are rotated 5 degrees to the outside. That seems to work quite well with point-at as long as one makes sure that the end points of the eyes have the same x-coordinates as the eye centers.

Quote - It didn't ask me if I wanted smart props. Perhaps because I was saving multiple props?

Yes, that's the reason. Poser does not give you the smart prop option when you use "Select Subset".

Quote - However, the eyes are not parented to the head as they were before I saved them.

Is there a hack I can do to the prop file that will make them automatically parent to the head like a smart prop?

Yes. Open the pp2 in a text editor. The parts that parent to something external to the pp2 (eyes and point target) will say "parent UNIVERSE", change that to "smartparent NameOfExternalParent", eg "smartparent rEye". The parts that are parented to items in the pp2 will be assigned to the correct parent, but with a normal "parent" statement. You need to change every instance of "parent" to "smartparent".

Quote - Just so you know: Antonia's original eyes don't actually point straight forward, but are rotated 5 degrees to the outside. That seems to work quite well with point-at as long as one makes sure that the end points of the eyes have the same x-coordinates as the eye centers.

Just to add: I googled a cross-section of a human head (the actual thing, not a model), and just from eye-balling that (no pun intended), the 5 degree outward rotation seems about accurate. As far as I know, if one tries to have the eyes point straight forward, one ends up with either a strangely-shaped face or a cross-eyed look.

bagginsbill,

Slight correction to my last post. You should only need to change the lines that say "parent UNIVERSE". Items assigned parents that exist within the pp2 will already use "smartparent". My excuse is that it's 5:45 am here, I've been up all night, and my brain has turned to mush (again).

One other point the new eyes in your pp2 should NOT use the internal names "rEye" or "lEye", those names would cause confusion for Poser as there are already items with those names in the figure. "rEye_1" or "rEyeBB" or "rightEye" are OK.
 

Attached Link: Point At

Okay - about the space the iris subtend of the complete eyeball I have an answer as Leo is awake now:

Diameter bulbus = 22-23 mm (0.87-0.91 inches)
Diameter cornea = 11,7 horizontal, 10.6 vertical (0.46 H, 0.42 V inches)
Diameter iris = 12 mm horizontal and vertical (0.47 inches)

I know, if one tries to have the eyes point straight forward, one ends up with either a strangely-shaped face or a cross-eyed look.*

If you want to google some more you may want to search for "horopter".

And off to bed for some sleep - as said I will be away for the weekend and try to catch up with the thread on monday!

Oh my, I forgot: Lesbently, your geometry swapper works like a charm!!! :thumbupboth: I just did a quick tryout but did not run into any problems!

BB.

Why do not you do the eyes of a FigureProp. Easy to setup their assigned room Bone. And then in the cr2, the motion units of the correct Antonia eyes implanted.

Then they should also move as Antoinas eyes.

Quote - Antonia is watching you. :lol:

With the eyes looking so real now, I think one of these days I'll have to experiment with strand-based lashes and brows for Poser.

Strand-based lashes sounds interesting.
As long as they are not the only lashes ... DazStudio users wouldn't be too happy about that.
Please!

Like to make some additional face/head morphs for Antonia. Not sure if it's wanted here.

Elke,

I'm sure I speak for most Antonia users when I say that any additional face morphs, or any other type of content for that matter, would be most welcome.

Antonia is just starting out, and more morphs, poses, clothes, what ever, means that she becomes more useful and versatile.

MikeJ,

As you know, I have been experimenting with injecting different geometry into Antonia, primarily as a way of changing the UV mapping on the fly. It all seems to be working well at the moment.

Of course the simple alternative is just to have a separate cr2 that uses a different base obj file, so I'm not sure how useful people would find my technique for swapping UVs on the fly. One place where it may be handy is with the different eyes that are about at the moment.

Any way, I sent SaintFox the files to swap your UV mapping into the 118 figure, and she seems to like it. Of course the files I sent her use your old 116 UVs for the eyes, and I understand that you have now done new UV mapping for the eyes. That brings me to the point of this post. If you are happy with the new eye UVs, would you send me a copy of the eye obj files, so that I can make a package to inject them into the 118 figure? I will IM you my email address.

Another question is how you want to designate your new eye UVs to distinguish them from your previous 116 version. As there are now multiple UV versions of the eyes, it seems inadequate to just refer to the old and new UV mapping.

Eventually I want to post the geom/UV injections to the developers site, but I feel I should wait until SaintFox has completed a texture map for your new eye UVs.

I would also like to include an injection for bagginsbill's eyes, but; a) I need to find out how he would feel about that, and; b) there are some technical difficulties. Because BB's eyes come in two parts, an eye, and a cover for the eye, so it not sufficient to just inject the eye geom into the existing eye actors. I would also need to inject new actors for the eye covers.

I can inject new actors for the covers. The problem is that I need to use a pp2 or cr2 to inject new actors, where as I need to use a pz2 to inject new geometry into existing actors. Thus at the moment it seems like the end user would need to apply both a pp2 and a pz2. So I am not sure if this is any improvement on just loading BB's eyes as a smart props, then hiding the original eyes.

I am going to do a bit of experimenting to see if I can come up with a solution where the end user only needs to apply one file to inject the BB eyes with covers, but I fear that this may not be possible without the use of Python, and I know nothing about writing py scripts.

The best solution would be if odf were to add empty and hidden eye cover actors to the 118 cr2, but I am not sure if he would want to change the cr2  now that it has been released.

I reduced the geometry of the cover, and added a translucence and some shadowing around the rim. This works very well to remove the hard edge of the iris. It should be very easy to texture as you don't have to pay any attention to the quality of the texture map at the transition between iris and sclera. Just size/position it correctly and all is taken care of.

Of the floating eyes on the left, the second one is separated from its cover, so you can see what the cover does. I have a little more tweaking to do on that geometry as it is a new shape and isn't quite right. Getting the normals to match between the cover and the eye is tricky and involves some serious math.

The good thing is it's a win all around. It is less polygons. It previews pretty well, not black. With the previous full cover, the details of blood vessels were somewhat reduced rendering through the cover - that problem is gone. You can mix different images for sclera and iris for the most part.

(Click for full size)

Looks fantastic, BB. It's always nice to come across another perfectionist. :laugh:

Now that we've got the info from Saintfox, are you going to fix the size of the iris?

Quote -

I would also like to include an injection for bagginsbill's eyes, but; a) I need to find out how he would feel about that, and; b) there are some technical difficulties. Because BB's eyes come in two parts, an eye, and a cover for the eye, so it not sufficient to just inject the eye geom into the existing eye actors. I would also need to inject new actors for the eye covers.

Can the cover geometry not just be made a sub-group of the eyeball geometry and thereby be controlled by the same actor?

Quote - Looks fantastic, BB. It's always nice to come across another perfectionist. :laugh:

Now that we've got the info from Saintfox, are you going to fix the size of the iris?

I did fix it. Based on the metrics I read, if we assume the eyeball is 23.5 mm diameter, this iris is 12 mm diameter, and the cornea is somewhat less than that.

And I made the eye the same size as Antonia's existing eyes, so they turn correctly. I have them parented to Antonia's eyes now, not the head.

Hey there lesbentley,

I'd be happy to send you anything you want. Actually, after I finish typing this I'll just IM you the d/l link for them. Seems easier that way. ;-)
Could you send me the swapping thingy too though, after it's complete with everything, including the eyes? I'll include my email addy in the sitemail.
Of course, the eyes with the sclera UV fix are only the lo res version, since that what's needed to use with remapping, and I'm hesitant to use LW's CC subdivision on them since it's different from Olaf's program's CC, and doesn't work the same. Might not be necessary with a simple transfer though.
I have to admit I've not been following this all too closely, as far as all the geometry switching goes. And I don't even have the AP118 version yet, since it still has the old UVs and my whole original purpose in redoing them was to have a version that would conform more to what I would want from UVs in a figure.
(Actually mine uses multiple regions/tiles which still work with all the new SaintFox textures since all the overlapping parts are just moved around in whole number values to different regions, and has nothing overlapping)
I can transfer all the UVs from 116 to 118 myself though using UVLayout, just haven't done it yet.

Quote -
Another question is how you want to designate your new eye UVs to distinguish them from your previous 116 version. As there are now multiple UV versions of the eyes, it seems inadequate to just refer to the old and new UV mapping.

Designate? Can I use something cool like X-17 like I wanted to earlier? ;-)
I don't know - it doesn't matter to me what they're called. I've been using dates instead of numbers on everything I've been doing for Antonia, but while that works for me it might not be the best way. maybe something like AP116c, since the last version using my UVs before the sclera fix was 116b.

But now we have this relatively new situation of BB making these great eyes, and it seems safe to assume he wants them included in the "official" Antonia, which is a great thing and certainly has my vote, but it's beginning to get a little confusing. ;-)

Anyway, I'm just going with the flow here and am happy to do whatever I'm asked to do when I have time. I have a pretty large project I'm working on though, which is taking away from my ability to closely follow all these new changes, so my attitude with the Antonia development is whatever happens, happens, and y'all have it under control pretty well. ;-)

Reminds me, I still have yet to finish my "One Map" Antonia for ZBrush usage. I need more hours in my day. ;-)

Quote -

Can the cover geometry not just be made a sub-group of the eyeball geometry and thereby be controlled by the same actor?

It can, indeed, and everybody else does that. What you lose is the ability to use refraction on the cover, because then it casts a shadow and the iris is not lit. If we make it transparent, then no problem, but I do not want to use transparency, I want to use refraction.

So, because there is no way to make some polygons in an actor cast a shadow and others not, except through transparency, it is required that the cover be a separate actor. I have it parented to the eye so its no big deal. Plus, it lets you hide the covers so you can work with pupil dilation morphs (which otherwise you cannot see in preview).

Quote - > Quote - Looks fantastic, BB. It's always nice to come across another perfectionist. :laugh:

Now that we've got the info from Saintfox, are you going to fix the size of the iris?

I did fix it. Based on the metrics I read, if we assume the eyeball is 23.5 mm diameter, this iris is 12 mm diameter, and the cornea is somewhat less than that.

And I made the eye the same size as Antonia's existing eyes, so they turn correctly. I have them parented to Antonia's eyes now, not the head.

Ah, excellent! Yeah, realized later that you had already changed the size. I thought they should appear larger with those relative sizes, but I'm glad that's not the case.

**Edit: **Ah, phantom3D's question was already answered by BB.

Yes I see what you mean, about the shadows.  Well A conforming figure should work fine, in any case fantastic idea. Refraction has definitely been lacking in eyes for Poser. No more trans mapping reflections for the eyes, they never looked real anyway. 

Also, Poser 8 IDL goes crazy with these covers. There is a bug that indirect light does not pass through transparency, and refraction is opaque, too, so you need to make the covers invisible to raytracing if used with IDL.

On the subject of dynamic UV re-mapping, please do not bother working out how to do that with the eyes. If you re-map the UV layout, then all the UV-based math in the shaders will break.

For example, I know precisely what the radius is on the cover, based on UV coordinates. I use this information in the shader to form the shadow ring, and the translucent ring, around the rim of the cover. I also use this information to know where the iris ends and the sclera begins, so that I have no specularity/reflection on the iris, but I do have specularity/reflaction on the sclera, and it transitions smoothly. You cannot remap the cover - it will totally change all the radius based math.

Instead, I use shader math to remap the UV coordinates for any texture map. Thusly, I can keep the same actual UV coordinates, while re-mapping other textures for this eye in the shader. For example, I can make a shader that uses iris images from V3 or V4 texture maps. Not a problem at all.

Quote -
On the subject of dynamic UV re-mapping, please do not bother working out how to do that with the eyes. If you re-map the UV layout, then all the UV-based math in the shaders will break.

Well I already figured that out, Ted. ;-)
While I haven't looked at the specifics regarding the coordinates you're using to planar map eye textures and whatnot to them., I do know that it can't be remapped and stay the same. ;-)
No worries, I'm not touching them.

This is just to give you a taste for how it is done. You can't run this script, because you don't have the rest of the geomatic library classes. I'll share those eventually, but right now they are changing to fast for people to find them useful.

> Quote - Elke,

I'm sure I speak for most Antonia users when I say that any additional face morphs, or any other type of content for that matter, would be most welcome.

Antonia is just starting out, and more morphs, poses, clothes, what ever, means that she becomes more useful and versatile.

THX for the answere. :) Will jump on this train and model some basic clothes for her later also.

Elke,

Good to have you on board. Just one point though, please don't embed a link directly to images that are wider than 400 pixels, as it makes the whole web page go wider, and then it becomes necessary to scroll sideways to read all the text. If you do want to post a big image, just use the "Attach a File:" box below the Reply Editor. Or provide an external link to the image.