First, I don't know if this is technically possible. Second: don't know, if this would be legal. It's about a tool that can convert Poser SubD-meshes into real meshes and keep the morphes. Giving me a menue, asking where to store the new Geometrie, the new cr2-file and which morphs to copy from the original mesh.   Can anybody do it?  Never mind ;-) just thinking to much. Have a nice day. :-) ---------...[the reason is, i just like hiRes modells more because I'd like to create more detailed morphs on my character. I don't like to use normal maps instead. Not every program will read normal maps the same way. And it seems easier to me me to have hiRes meshes in Renders like Lux or Octane]...

SubD is a smoothing algorithm where it calculates based on a base cage. So any changes has to happen to the base, not the resulting smoothing; so no, this wouldn't be possible.  This is a shift in how you do character morphs, so you would have to learn how to manipulate the base SubD cage to get more detail or learn how to make displacement or normal maps for your morphs.

Also for luxrender, the less polygons passed to the renderer, the better since the more mesh you have to calculate for the render, the longer the render time.

Well, it would be nice if there would be a program to do that automatically with a few clicks, but it isn't really that hard to do the old fashioned way:

Load the figure and ABSOLUTELY ZERO IT !

Export the geometry keeping the group names.

Load it into your modeller and subdivide it once. In Wings 3D I first have to combine the groups into a single mesh for the subdivision to work across the seams. Then I split the mesh back into groups. In Wings, I then have to redo the group names,but more "modern" modelling programs surely can save those names.

Export the subdivided mesh back into Poser.

Create a new figure using the cr2 of the original low res mesh.

Do not use "Auto-Group", but use "Transfer Morphs"

There might be some cleaning up morphs/smoothing joints left to do afterwards because you went from a low res mesh to a high res mesh, but that is to be expected.

Not a beginners task, but not completely impossible, either.

In the end you have a figure where you can directly manipulate each and any vertice directly in Poser and not just the low resolution cage mesh.

---

BTW, I really don't understand why all of a sudden certain people constantly make such a fuzz about subdivision.

We have props in Poser that are 250.000 polys and more. Some hair is over 100.000 polys. Cars are easily in the 150.000 poly range.

Most elaborate clothing is several times as "heavy" as the base figure.

We waste RAM on multiple 4000x4000 textures.

And even on my old i5 Laptop Poser rarely even bats an eye because of all that polygons, neither in OpenGL preview nor for the Firefly final render.

How much rendertime do those people think they "save" by using a rubbish looking 20.000 poly figure instead of a realistic looking 50.000 poly figure ?

One nanosecond ?

One microsecond ?

Not to mention that subdivision and displacement/normal maps take time to compute, too.

 

Sometimes I really think I'm the only....nah, I rather don't say.

;-)

Use whatever figure makes you happy in Poser.

 

But there's only one thing better than polygons.

MORE POLYGONS !

:-)

 

 

 

Qouote:"I *really don't understand why all of a sudden certain people constantly make such a fuzz about subdivision."   *

I don't know either. These are not the 90* anymore. Lots of People have 64bit programms and 16gByte of ram nowadays, and we worry about 20 000 or 60 000 polygons? Really? We are not creating ego-shooter games in Poser here.

Quote - These are not the 90 anymore. Lots of People have 64bit programms and 16gByte of ram nowadays, and we worry about 20 000 or 60 000 polygons? Really? We are not creating ego-shooter games in Poser here.

Amen! 

Timberwolf, what kind of morph are you attempting to make and in what program? If I have it, maybe I can make a quick timelapse video of how to do what you're looking to do.

Quote - I don't know either. These are not the 90 anymore. Lots of People have 64bit programms and 16gByte of ram nowadays, and we worry about 20 000 or 60 000 polygons? Really? We are not creating ego-shooter games in Poser here.

 

100 % agree...

Polycount? Oh dear, here we go again.

Perhaps the only thing DAZ did right when they reduced polycount from V4 to Genesis 1 & 2.

A WIRE FRAME is just a COATHANGER to hang a texture on.

DAZ and Posers SubD and Posers Smoothing with controllable crease angle take care of the rest.

It is texture you see, it is texture you render.

Be it duffuse texture or displacement or normal map, or a combination of all.

Poser has texture caching, to reduce memory load for textures. (When set to quality or crisp)

Rex and Roxie have perfect polygonflow at 26.000 poly.

Take some of the hi polygon figures. Where are the extra polygons?
In 30.000 poly inner mouths and teeth?
In 10.000 polygon ears?
In 5.000 polygon navels?

Before talking polycount and polygonflow, look at where the polygons are.
Compare the meshes in wireframe.

Hi polygon count is RARELY if EVER justified with muscle topology.

It is a hype and a sales argument, nothing more.

Just look at renders from Reality 3 and Octane for Poser. You can see the low polygons on props and figures. Even on a V4.

So yes. Higher poly counts is the future to get better renders.

If you see the polygons in Reality and Octane?
There is a problem with those renderers. => Use firefly.

Or use another renderer that can handle DAZ & Posers SubD and Posers Smoothing, set at crease angle 180°.

Remember; I have nothing against Hi Polycount.
But it has to be justified for the figures purpose. NOT to please the render engine.

Posers renderer & SubD & Smoothing, are as smooth as they come.
Even on as low as 2.000 poly figures.

I really hope you're joking vilters - I really do!

Attached Link: http://www.renderosity.com/mod/gallery/index.php?image_id=2427984&user_id=280495

***Hi polygon count is RARELY if EVER justified with muscle topology.***

For high quality renders, high poly-count will always give a better render result than low poly - even for FF.  More polys = more detail for the renderer to work with.  Otherwise you render a dept store plastic dummy.  Real people have real shape, and that takes a lot of polys to describe.

The render above was somewhere around the 500k poly count for his head PLUS a normal map (can't find the scene to give an actual polycount).  More polys = more quality.

Paul

C4D,LW,Max,Maya make around 20,000 polycount characters.for UVMaps,better bending Rigs, etc etc

Send to zBrush or app's like zBrush make Vector Maps or Displacment Maps ,polycount in the millions.

This can be done for Poser and DAZ Studio with Displacment Maps.

 

View zBrush Gallery .

http://pixologic.com/zbrush/gallery/

If you can find a better Gallery would be cool with me if you shared the URL.

 

Let me reword that.

If you can find a gallery with more detailed meshes.let me know.

That head face_off - the one you posted a pic of - I have a low res version of that which renders about the same. It comes down to the quality of the underlying forms. If the underlying forms are good then you can get a great render out of the mesh. Doubly so if you are using subdivision alongside high quality maps (high quality maps does not always mean high res by the way). 

 

 Also, for reasons I don't fully understand as it seems many folks here use modeling applications, everyone seems to forget you can subdivide your mesh. So if you plan your low poly morphing with subdivision in mind, you should be good to go. Again, if there's a particular morph that you're having trouble creating I can try and do a video of what my approach to that would be.  I don't understand the...well knee jerk outrage at lower poly figures. I know it may not be familiar to everyone but they're just figures like any other.  As long as you've got good underlying forms and an understanding of the morph you're trying to make you should be able to figure things out after some time playing with them.

 

 I don't know I guess for me, this is just normal - I work lower res  day to day. Every Poser figure I've worked on, Andy, Andrea, the P8 dog, James Johnson, Hanna, Miki3, Miki 4 - they all started out as low res meshes. I had to work them on a low res level. So I guess maybe I'm just used to it and that's why this constant rally against a lower res mesh seems odd to me. People tend to lash out at things they don't understand but if you can point to a form change that you can't do with a low poly mesh that you can do with a high poly mesh, then maybe I could understand where you're coming from a little better. 

I should add, I am not trying to imply that high poly meshes have no place I'm just trying to understand what kinds of morphs you're looking to do that you're unable to on a lower poly model. Because I have a feeling that some may be down to mesh flow as opposed to poly count but I want to hear from you guys first. 

That head face_off - the one you posted a pic of - I have a low res version of that which renders about the same. It comes down to the quality of the underlying forms. If the underlying forms are good then you can get a great render out of the mesh. Doubly so if you are using subdivision alongside high quality maps (high quality maps does not always mean high res by the way).

Hi Teyon.  The render quality of that head model with and without the normal map is huge - so the normal map makes a big difference and that case.  The head at http://www.renderosity.com/mod/gallery/index.php?image_id=2396248&user_id=280495 is 1million polys, and doesn't have the normal map - so it needs all those extras polys.

My general workflow is the pose the figure, merge all the skin materials into one with a python script (requires joining the separate skin texturemaps into one too), export to ZBrush, subd (usually twice), import back to Poser, then use GoZ to push and pull the mesh (now a prop rather than figure).  The biggest issue is that if I want to change the pose, I can't move the base mesh, because then I loose all the work on the subd prop.  Often subd'ing twice is not enough to get the look I'm after.  Say a person leaning to one side and twisting will have skin folds on the lean side, and if the polygons are not going in the same direction as the folds, you need a lot of subd'ing to get the right look.  On the standard Poser models, areas such as ribs, shoulders, neck, knees, etc have no where near the number of polys needed to accurately model the lumps and bumps of a real person.  And those same models split the skin into 3 or 4 zones, which means you cannot use ZB to generate displacement maps (unless you merge all the zones).

IMO, the perfect Poser figure would have ONE material zone for skin, and have way more polys.  Then we could use GoZ to push the model around without subd'ing and reloading as a prop.  Or, ZB could be used to generate a displacement maps from the posed figure.  The way it is at the moment it's a substantial technical exersize to do either of these things.

Paul

HEy Teyon

I used Zbrush to make my own Megan Fox morph- all went well-but V4 did not have enough polygons in the tip of the nose to make the nose crease look right. In her nose tip-and that of many people-the nose tip has two lobes separated by a crease. The topology of V4's mesh is not conducive to that-and I am not good enough at drawing displacement maps to pull it off. A SubD of 1 or two-adds enough polygons to support that detail.

Quote - I should add, I am not trying to imply that high poly meshes have no place I'm just trying to understand what kinds of morphs you're looking to do that you're unable to on a lower poly model. Because I have a feeling that some may be down to mesh flow as opposed to poly count but I want to hear from you guys first. 

Frist, thank you guys for the feedback. It's important to be able to come with specific examples. 

I should point out that Smith Micro's models - Alyson, Ryan, Miki3, Miki4, Rex, Roxie - if loaded from the Runtime folder will subdivide in ZBrush without breaking at group seams so there'd be no having to merge grouping. Doing that would require unziping the runtime folder though and I understand not everyone wants to do that but I thought I'd pass that on.

 

 Now it could be argued that defined rib detail is detail not form but I personally try to throw in morphs that give the impression of a rib cage under the skin so I can understand the desire for such a morph. In fact I made an attempt at it for Roxie (I can't recall if I did it for Rex) but it was fairly subtle.

 

 

UGH! I recorded a video of approaching a rib cage morph on Roxie and effin' Camtasia once again did not capture my audio despite the option to do so being on. I'm annoyed now.  Anyway I just want to thank you again for presenting a specific area that I could in turn be like, "Hey so folks are having problems with this area of the mesh, let's look at the flow of the topology and consider what to do for the next model". That's a lot more useful than just saying low res meshes are crap or that you can't work with them.  Hopefully the video (which I'll post as soon as I can get it up on youtube) will be of some help to change the way you approach morphing. I am so bummed my audio didn't record...I was explaining things as I went so you knew why I was doing things a certain way. F U Camtasia Studio! lol.

I should point out that Smith Micro's models - Alyson, Ryan, Miki3, Miki4, Rex, Roxie - if loaded from the Runtime folder will subdivide in ZBrush without breaking at group seams so there'd be no having to merge grouping.

Yes, but you still can't do normal or displacement maps (to get the details), because there are multiple skin materials on the skin mesh.  Also, if you try to subd just one part of the mesh, ZB merges all the material zones into one (including eyes, etc), so with the current meshes it not possible to just subd the body and not the head.

Paul

Attached Link: Ribs

I flagged the nudity tag even though the genitals/nipples are off just to avoid issue but here's the video sans audio. I tried recording narration but it wouldn't let me save it, claiming another program was using it (that program being Camtasia itself).  Anyway, I hope it helps you to see how you can get details out of the model.  I really appreciate you guys giving a specific example. It's more useful towards our understanding than a blanket statement of dissatisfaction. At least now I can say, hey, people are having issues with this part of the mesh what should/can we do about it on future models?

Quote - I should point out that Smith Micro's models - Alyson, Ryan, Miki3, Miki4, Rex, Roxie - if loaded from the Runtime folder will subdivide in ZBrush without breaking at group seams so there'd be no having to merge grouping.

Yes, but you still can't do normal or displacement maps (to get the details), because there are multiple skin materials on the skin mesh.  Also, if you try to subd just one part of the mesh, ZB merges all the material zones into one (including eyes, etc), so with the current meshes it not possible to just subd the body and not the head.

Paul

Ah. That's a ZBrush thing - it likes single maps - you can auto group by UV and then hide parts before creating displacment and normal maps. That's what I do. Don't need material zones, just the UV's.

Tricks that I explained in the video that you can't hear:

 

1. import the base OBJ from either the runtime or an export into UVMapper. Delete the eyes, teeth and tongue.  Save this back out as a body base for sculpting.

2. You can use ZBrush's polygroup tool to group the mesh by UV's.  Once grouped this way you can do your sculpting as normal then hide parts before creating displacement and or normal maps. This prevents map overlap. - note this works for Mapping functions. For morphs you'll need to load the mesh in UVMapper and import UV's from your base to get the groups back.

3. After importing the OBJ into ZBrush, save a morph target inside ZBrush so that you can always get back to the base.

4. Always work in layers, it's non destructive and allows you to dial in the intensity of what you've sculpted. Great for mixing and matching effects!

 

If I can get this effin' mic to work with Camtasia, I'll do a whole sculpting session so you guys can see/hear the workflow. It'll be more a Zbrush and how it works well with Poser tip section than attempting to achieve any specific morph. However, it sounds like this is an area we haven't supplied enough information on so I think it's needed.  I'll try and do that sometime this week.

One of the things about the new figures that actually is kind of cool is that we don't have all the UV's overlapping in the 0-1 space. Load the figure into UVMapper or another UV tool and you'll see the UV's are spread out. This makes it a lot easier to see what goes with what on a texture map and why using the Group by UV feature in ZBrush or Auto group with UV's is pretty handy. Just run that and then hide the stuff you don't need for the texture you're trying to export.

The head, inner mouth, tongue, eye sockets and tear ducts are all one map. The body,fingernails and toenails are all one map. The teeth and gums are all one map (though texture space could have been used a bit better there). The eyes and cronea are all one map.

So be sculpt away and just remember that you can group by UV and hide what you don't want to create a map for. You can hide groups in ZBrush by holding down the CTRL key and the SHIFT key  while left clicking on a group.  One click hides anything not a part of the group you clicked on and another click into the same space will invert that.

I feel like maybe we haven't done the best job educating the user base on how best to use ZBrush in conjunction with Poser. I imagine for Mudbox users the ideas are pretty much the same also. The only problem that should prevent you from being able to use the two together is if you have a mesh with unwelded groups. That can be an issue. It's resolvable - at least for texturing - but it's an issue. Anything else can be addressed by using the tools in ZBrush or now that we have GoZ, using that in conjunction with Zbrush's toolset. Definitely will try and make time to do a walkthrough.

Oh well...

It gets boring after a while....

Poser users are hobbyists. They hardly ever have pro tools.

My first "ribcage" was made with several dozend magnets, right in Poser 4.

While I do use ZBrush, I still use Version 3.0, as the latest versions with their gazillions of buttons and brushes and shortcuts and menus are too much for my 48 years old brain.

Also don't have a tablet, just use the mouse for everything.

That's why I use Wings3D as my modeller. Everything can be accessed with a right mouse click. No shortcuts to memorize.

No "speed enhancements" to aid "productivity in a professional pipeline".

You select a vertex and you move that vertex. And you cut an edge if you need a new vertex.

Simple and easy.

That's how I want my Poser figures to work: Simple and easy.

Working on a high res mesh is simple and easy. Use a magnet, use the MorphBrush, or even use a simple modeller like Wings.

Working on a SubD mesh isn't easy. Not for a Poser hobbyist.

You NEED ZBrush. And you NEED the latest version. And you NEED to know how to use it.

Another example:

The introduction of split UV-maps almost dried out texture freebies.

During the Posette/V2 split mapping and V3 pelt-mapping days, we had gazillions of home-made textures. Not all excellent, but a lot of variety because everybody could participate as long as he/she could open Gimp.

From V4 on you NEED a 3D painting tool to paint across seams because doing it in Photoshop or Gimp isn't possible any more.

Now we only have "professional" textures and perhaps the occasional merchant resource freebie. In fact even most merchants use resources these days, because making a good map for V4 from scratch has simply gotten too hard for the average hobbyist.

And let's face it, 99% of the Poser merchants still are just "advanced hobbyists" and not "all-out CGI industry professionals".

That's why I'm 100% opposed to any attempt to introduce arbitrary "professional standards" into Poser.

Especially if these "standards" have zero value in everyday use.

Poser is no game engine and your average hobbyist or advanced hobbyist or semi-pro hobbyist users positively do not care about professional pipelines or what "The Industry" does.

Progress is great. But only if it is actually useful for everybody and not just a selected few.

I don't care if a professional merchant saves a few minutes making clothes or rigging them if he can do it on a 20.000 polygon mesh instead of a 50.000 polygon mesh.

The need of the average users is much more important than that.

Content quality has drastically dropped with the introduction of V4, because everything is quick, quick, quick now.

And forcing SubD as a new standard will make things even more "quick, quick, quick."

The more simple, more straightforward solution is always better, because it is more acessible to more people.

And that's what Poser is about.

Giving 3D to as much people as possible.

 

To sum it up:

1. Hardly anyone can create a proper displacement map.

2. There is neither a RAM nor a render speed benefit.

3. Displacement maps are much more shape specific than morphs

4. They can't be manipulated directly in Poser.

5. Low res "cage" meshes can't be "cross flow" morphed.

6. Subdivided polygons can't be manipulated, neither in Poser nor ZBrush

7. Displacement maps are much harder to animate to simulate muscle          movement than a simple morph is.

Guess I forgot a few but that should do.

Muscle STRUCTURE and VEINS ? WRINKLES in a face and MINOR DETAIL ?

I'm all for displacement maps there.

SubD to smooth props that don't have much detail like transmapped hair ?

Or to give "extra smoothness" for extreme close-ups ?

Great tool to have.

But the MAJOR SHAPES of the human body (And of clothing and other realistic props) has to be build out of actual polygons.

Not virtual ones and not displacement map fakery.

I want the anorexic teen girl as well as the Hulk clone as well as the morbidly obese housewive as well as the average guy with a flick of a dial.

And I want them as REALISTIC as possible.

And that I can only do with lots of polygons.

 

 

 

Thank You for pointing it out. I agree most with point 6. and 7. Creating JCM for excample, I'd just like to have controll over the whole mesh. Saying I can subdivide in a Moddeler like c4d, doesn't hit the point , because that figure is not useble as a figure in Poser anymore. However it would be cool, if someone could do it.

Attached Link: Silo morphing

I guess you're just going to feel how you feel. I'm not going to try and convince you otherwise. You seem fairly set in your ways and that's ok. Though I do wonder what you may think the shapes currently in the figure are built out of if not built from polygons? :-)

 

However, I will post up a video for others who may be interested in how to morph these figures and need help in doing so. This video is a quick example of morphing the figure inside Silo without the use of Silo's sculpting tools. In it I first show a quick example of doing a similar morph to the one I did in Zbrush. I then go on to alter Roxie's body so that it appears a bit more flabby, particularly around the lower midsection.  I could continue on, reshaping her into an obese woman or go in another direction entirely using edge or vertex manipulation with soft selection. All it would take is time, a knowledge of what the morph I want will look like (you can use references in Silo) and a knoweldge of how to use Silo. I would hope anyone attempting a morph would be familiar with the tool they are morphing in before begining. If not, you should spend some time learning your tools first and then hop into morphing. Once you do you'll not want to stop. :)

 

Speaking of vertex manipulation, you can manipulate vertices directly in Zbrush and in Poser using a very small brush size.  In Zbrush you'd do this along with the move brush or the move topological brush. For anyone who wasn't aware.  Works well with the wireframe active so turn on Draw Polyframe in Zbrush.

Also if you find that you've lost your groups - either through export/import or due to subdivision quirks in whatever app you're using - you can get them back by loading your morphed figure into UVMapper (I think this is in both Pro and basic) and then choosing Import UV's from the UVMapper file menu.  Select the original unmorphed base as the file to be imported. This will overwrite the groups in your morphed figure. Then just save it out under a new name. Easy. :)

One more thing, for anyone considering morphing for a subD mesh be it Rex, Roxie, Genesis, or Genesis2 - it's basically the same as morphing for a high res mesh. You're simply working on a broader scale as opposed to a fine one. So it will take less geometry to achieve a form. That's the positive. The negative is that you will have a harder time getting finer details. Depending on the mesh this could be wrinkles or it could be muscle tone, what have you. It can be argued that either of those details could be done via morphs or displacment or both combined. Which you chose depends on your skillset, the models and the tools at hand. Details shouldn't be confused with form. A form is something you can see in silhouette. So if you can look at a body in silhouette and the shapes look right, then you've got good form to start your morphing and or sculpting your details. It's a lot like box modeling that way. Box modeling is a detail-in type of modeling, where you start with broad forms and cut in your detail. So if you're a polygon modeler, think of it that way. If you're a NURBS modeler you should be ok because you're probably used to making the most out of as few points as possible.  

As for Poser's latest figures, Rex and Roxie's basic forms are solid. They're not perfect - there really is no such thing - but they're well within the realm of believability. So they provide a good starting point for a morph. As with anything, take your time with it. When you first start out you want to work with large sweeping changes so if your modeling application uses soft selection or has a brush system, you should take advantage of it.  As you're doing your morph, be aware that edges that are close together may cause a crease once subdivided - this can be used to your advantage to help simulate details like muscle tone. 

Do not allow yourself to think something is impossible. If you were to look at Andy2 and note how low res he is you'd think it unlikely you could morph his head into a human face - especially considering it looks like a blob to start with. However I did it and you can too if you take your time and aren't afraid to make a mistake. Remember this is supposed to be fun so don't put pressure on  yourself to get it right the first time or even the eighth time. 

 

I hope the stuff I've been posting has helped someone. I'm sorry I couldn't help the thread starter but if I've helped anyone at all then I'm glad.

Hello Teyon,

While GOZ seems to work very good, (I hear few complains)

I, as many others do NOT have Zbrush dollars :-(

A "GoZ" thing to Blender would be a welcome enhancement.
PS: its in Mantis; :-)

Currently the only GoZ thing for blender that I'm aware of is OBJ and Collada lol. OBJ is the way to go for most applications. We didn't design GoZ, Pixologic did, we just included access to it.  I'd love for us to get FBX support some day but I don't know if that will happen.

I know, I am dreaming sometimes.
But? Never give up hope :-)

Teyon : Yes you helped me. Thanks.

Cool. :-D

Thanks for the time you took to put together all that info Teyon.

My original comment was not meant to be dissatisfaction with the pre-P10 models (I haven't tried the P10 models yet) - someone said high poly models were not needed, which I disagreed with.  For my purposes, I need very high poly meshes to get the look I'm after (I have a life-drawing background, so love to get all the details in).  For many years I tried to develop a ZB workflow to do this, but the Poser model texturemap splitting, plus some ZB shortcomings meant I could never get the workflow to actually work.  The closest I've got is spliting the V4 skin texturemaps into quadrants and merging all the skin materials into one (so all skin referenced one set of UV's, including the head).  But even that is not without it's issues - if you need to change the pose you had to start again. 

With low poly models, if you've got a large polygon running the wrong direction to a seam you are trying to add, you are stuffed!  The only way out is a) subd that area - then you loose your UV zones, or b) subd the whole mesh, so you are adding 100's of thousands of polys to the model just to add one seam.  I guess c) could be use displacement maps, but I have yet to see displacement mapping work convincingly on a Poser model, because of the join areas.  d) could be convert the figure to a prop, then do a displacement map, but then because of all the multiple UV zones for skin, you have many more problems.

So in summary, the quest to build low-poly models makes it impossible to get renders with the details I am after.

Also if you find that you've lost your groups - either through export/import or due to subdivision quirks in whatever app you're using - you can get them back by loading your morphed figure into UVMapper (I think this is in both Pro and basic) and then choosing Import UV's from the UVMapper file menu.  Select the original unmorphed base as the file to be imported. This will overwrite the groups in your morphed figure. Then just save it out under a new name. Easy. :)

You cannot do that if you have subd'd the mesh in ZB.

Paul

As long as you have the lowest level available this should still be an option. I have done this many times in the creation of the morphs for rex and roxie. If at any point you delete the lower leves or you export a higher level or a localised subdivision you are correct that you won't be able to do that.

 

Higher poly models are great. They have a purpose just like low poly models.  Both are valid and both can be morphed into whatever you want.

Quote - Higher poly models are great. They have a purpose just like low poly models.  Both are valid and both can be morphed into whatever you want.

Yes, but when you can actually see the polygons in the final render I don't think it's "valid" anymore, as there's just not enough polygons to get a smooth surface. The G2 is impossible to look good in 3'rd party render engines as her shoulders has not enough polys.

> Quote - > **Rex and Roxie have perfect polygonflow at 26.000 poly.**

 

Sorry - I haven't read the whole thread but had to comment on this from Vilters.

 

No they don't. I spent most of yesterday working on this FBM for Rex and I can tell you, the mesh is pretty damned stubborn. 

That's the most detail I can get out of his hips lower abs and legs. The geometry just isn't there for certain major features where it should be. Shoulders is another area. Good texture can fill in for some of what the mesh lacks but certain forms need to be sculpted in, in order to get proper bending and muscle movement (with the help of weightmapping) during rigging/animation. Morphs can't change a texture or a displacement map.  

Some areas of the topology make sense, others just don't.

The current poly count would have allowed for it with a bit better distribution. A higher poly count of around 30k would definitely have allowed for much better detailing and room for mistakes. There's always going to be areas where you have more polys than you actually need. They're the sinkholes/landfills of mesh design, where all the excess geometry sort of congregates. you can get rid of most of it once you have all the necessities in place and are going back over the mesh for clean-up/optimizing. Rex's face could definitely use more geometry, especially around the nose and ears.  

A mesh designed for subD takes a bit more pre-planning and design in order for it to work just right and be flexible enough to pull all the details out of it that you're trying to get in the first place. 

 

~Shane

Teyon

I'm just going through all your info, and experiementing with Roxie.

Do you know if there is a way to export Roxie base mesh to ZB, subd it in ZB and import back into Poser without loosing the material (UV) zones?  Previously I've converted the material zones into geometry zones, which sort of works - but far from perfect.

Paul

Any one got a link to venders meshes that actually uses displacement maps in Poser & or DAZ Studio ? V1 thru V6 ,Rox's or any mesh don't care.I just want to see how well meshes bend and morph and all with displacement maps.In Poser & or DAZ Studio.

Thanks

Don't really have anything in my runtime .
What came with Poser Pro 14 & DAZ Studio Pro 4.6.
and a very little else. Guess ya could tell buy looking at my Gallery .

@ Timberwolf would snarly's script for poser work? http://snarlygribbly.org/3d/forum/viewforum.php?f=28&sid=abca9c8783ceb0cc19c6662d222648d3

@Teyon, great video's I really wish the audio worked, might have made it a lot easier to follow. Because you're so quick it's hard to catch what's going on sometimes. Also the use of shortcut keys I suspect doesn't come across. I have silo would love to be able to make morphs for poser in it, but I can't even figure out how you got symmetry to work like that. I have swore at silo, but it didn't help it still refused to do symmetry like that. Also how you managed expand soft zones so easily looked pretty cool. I have tried some online tutorials for silo, but they seem to leave out some simple little steps, that maybe easy for knowledgeable users, but for complete newbs, it's frustration.

I got silo to use for poser, so showing and explaining how to use it with poser is a great help, Thank You.

 

Quote - > Quote - Higher poly models are great. They have a purpose just like low poly models.  Both are valid and both can be morphed into whatever you want.

Yes, but when you can actually see the polygons in the final render I don't think it's "valid" anymore, as there's just not enough polygons to get a smooth surface. The G2 is impossible to look good in 3'rd party render engines as her shoulders has not enough polys.

Well the idea with subdivision is that you do your shape changes and then you smooth at render time. If the renderer you're using doesn't support subdivision than you're probably right, that would not be the way to go. I don't know much of anything about Genesis 2 except that it got released and it's still fairly low/medium res. I can't comment on its bends honestly. That may well be a rigging issue as opposed to an issue with the mesh I don't know. 

> Quote - > Quote - > > **Rex and Roxie have perfect polygonflow at 26.000 poly.** > >   > > Sorry - I haven't read the whole thread but had to comment on this from Vilters. > >   > > No they don't. I spent most of yesterday working on this FBM for Rex and I can tell you, the mesh is pretty damned stubborn.  > > That's the most detail I can get out of his hips lower abs and legs. The geometry just isn't there for certain major features where it should be. Shoulders is another area. Good texture can fill in for some of what the mesh lacks but certain forms need to be sculpted in, in order to get proper bending and muscle movement (with the help of weightmapping) during rigging/animation. Morphs can't change a texture or a displacement map.   > > Some areas of the topology make sense, others just don't. > > The current poly count would have allowed for it with a bit better distribution. A higher poly count of around 30k would definitely have allowed for much better detailing and room for mistakes. There's always going to be areas where you have more polys than you actually need. They're the sinkholes/landfills of mesh design, where all the excess geometry sort of congregates. you can get rid of most of it once you have all the necessities in place and are going back over the mesh for clean-up/optimizing. Rex's face could definitely use more geometry, especially around the nose and ears.   > > A mesh designed for subD takes a bit more pre-planning and design in order for it to work just right and be flexible enough to pull all the details out of it that you're trying to get in the first place.  > >   > > ~Shane

 

Yeah see, the only kind of perfection I'd ever claim about ANYTHING is that it's perfectly IMPERFECT. lol. There's no such thing as perfect. There's perfect for a given situation but not all situations. That's why I try not to use absolutes like that.  

That said, we agree that a good sweet spot for models would be the 25K-30K quad poly range. The mesh flow Darrell implemented in Rex and Roxie reflect choices he planned to make in their rigging to allow for good deformation while at the same time trying to accomodate concerns for morphing. Not an easy tightrope to walk.  He did a good job capturing my sculpts and while I wouldn't have made every choice he made in the topology, I still feel you can get good forms. Details - yes, you're right, some details may be difficult to achieve directly with the way the mesh is currently laid out but the forms those details are placed on are there. Muscle detail can be had in the mesh directly and taken further with a map. I have taken the body builder morph for Rex and loaded it in Zbrush so you guys can see the difference between form and form with secondary details.  The secondary details are where things can go either way (easily made and not easily made).  No additional sculpting was done to these models. I just loaded the OBj and subdivided. I could have given the same example in  Poser but I don't have decent shiny materials for what I wanted to show.

This is form. 

And this is secondary detail.  I think this is the area most of the people being vocal about poly counts are having trouble with, right? So here I think it's mostly mesh flow as opposed to poly count that's really the issue and that's something I can be sure to talk to the team about for future models - we take everything into consideration.  So yes, like I said, some of the secondary detail would be harder to achieve but I think where they fall short, since most people are rendering stills, a displacement map could fill the void. It's a moot point if the model has clothing on. Then it's just wasted detail anyway which is why we used to be taught to model only what was seen.  It wasn't until dynamic cloth became more popular that modeling the underlying primary forms were important.  Again, those forms are present.

Quote - Teyon

I'm just going through all your info, and experiementing with Roxie.

Do you know if there is a way to export Roxie base mesh to ZB, subd it in ZB and import back into Poser without loosing the material (UV) zones?  Previously I've converted the material zones into geometry zones, which sort of works - but far from perfect.

Paul

That depends on one simple thing: Does UVMapper Basic have the option to subdivide a mesh? I know UVMapper Pro does.  

 

If so, in theory you can get back the groups and materials in a similar fashion as described earlier (or you could skip ZBrush entirely). If UVMapper Basic has the option to subdivide or you have UVMapper Pro by chance, you'll be able to subdivide the mesh in either app and retain both material zones and groups. HOWEVER, you will lose access to any morphs that were in the lower res version.  This may not be a concern for what you're doing, as many things on the figures are handled with bones and magnets. Including opening the characters' mouths.

Stephanie I/Michael 2

They are the same mesh and are 34700 polygons.

That is the bare minimum needed to create an accaptable "ripped" muscularity.

The example you give, Teyon, looks way too soft.

Quote - @ Timberwolf would snarly's script for poser work? http://snarlygribbly.org/3d/forum/viewforum.php?f=28&sid=abca9c8783ceb0cc19c6662d222648d3

@Teyon, great video's I really wish the audio worked, might have made it a lot easier to follow. Because you're so quick it's hard to catch what's going on sometimes. Also the use of shortcut keys I suspect doesn't come across. I have silo would love to be able to make morphs for poser in it, but I can't even figure out how you got symmetry to work like that. I have swore at silo, but it didn't help it still refused to do symmetry like that. Also how you managed expand soft zones so easily looked pretty cool. I have tried some online tutorials for silo, but they seem to leave out some simple little steps, that maybe easy for knowledgeable users, but for complete newbs, it's frustration.

I got silo to use for poser, so showing and explaining how to use it with poser is a great help, Thank You.

 

 

Happy the videos are welcomed. :) Let me work in reverse:

 

I have a mouse with a track wheel. Silo uses that wheel to increase and decrease selection size when soft Selection is active. You can also use the sliding method by clicking in the Options window for Soft Selection, holding down your click and dragging. I prefer the wheel option. It's faster and just works.

Symmetry in Silo can sometimes be flakey, that's true. It's not just you. I tend to avoid that by deleting free floating bits if I don't need them for my morph. So like when doing a full body morph, I get rid of the teeth, tongue and eyes in UVMapper first and then save that out for my morphing.  Cuts down on the chances of symmetry getting confused. Then I calculate symmetry via that option in Silo and I work on one side. Sometimes, I forget to turn symmetry on or if I run into situations where symmetry isn't quite right I just make the selection on both sides and keep it moving. Tends to be faster than getting myself worked up over program issues.

And yes, I was so angry at Camtasia Studio. I have version 7.1 and it's been like this off on and for awhile. The manga studio 5 videos I did on our site had to be redone like 3 times before the audio finally worked. Annoying. I think I may go back to using CamStudio instead - it's free and not as frustrating.

An example for extreme skinnyness.

This is the kind of body detail I expect a modern Poser figure to be capeable of without the need for displacement maps.

If we have super detailed props in Poser, there is absolutely no reason we can't have realistically sculpted super detailed humans, too.

And the human shape isn't just your average dude and hot chick.

Sculpting them is piss-easy. Especially if they are "stylized" to death and don't stick to proper reference.

Humans are tall, short, old, young, pretty, ugly, fat, muscular, skinny, whatever. They are probably the most complicated shape in the world.

It's the extremes that separate the weed from the chaff.

Poser is about letting the user decide what he wants. If a Poser figure can do only the averages, it's no use.

Quote - Stephanie I/Michael 2

They are the same mesh and are 34700 polygons.

That is the bare minimum needed to create an accaptable "ripped" muscularity.

The example you give, Teyon, looks way too soft.

 

I wasn't going for ripped. I was thinking more muscle man from like back in the day. I can agree that more polys would help in getting that ripped look you're wanting, Joe. I also think - for a still anyway - you could take what I've done, throw on a displacement and no body would really know the difference assuming the person making the displacement had a clue about anatomy that is. I know that isn't always or even often the case though. It's not a perfect solution and I know it's not the way some people like to work.  This go around, this is what we felt was best for us in terms of making a quality set of figures that bent well. For the most part it has worked - outside of the folks who want more polys, the only real complaint I've seen about the figures is the elbow issue. It's a sign, to me at least, that we're heading in the right direction. I mean, you're not telling us you dislike the figures, you're telling you want more polys to work with the figures the way you're used to working. I see that as a good thing and despite what you may think, I hear you.

Like I said, this go around we centered more on getting somewhat attractive figures that bent well and worked easily with subdivision over the previous way of doing things where we were more concerned with morphing almost at the expense of all else. Ryan and Alyson were two of the most morphable (they came with a massive amount of morphs) figures we've ever released but alot of their forms were off and they weren't terribly pretty. In a case like that you have to strip away everything and rebuild to have a solid foundation for future development. I think we've done that.

When the only real issues people are having are wanting more polys or wanting an elbow fix, I can honestly say I'm happy. You guys have been great in this thread. Really great and I appreciate you expressing your concerns about the polycounts and giving specific examples of areas you're less than thrilled with. This is great information to go back to the team with and say, ok...we've learned some stuff about what we did right and what people want to see from us in the future, what can we do with this new information?