Need help with Character Workflow

kunglao

I have built a character in max 8 (see Female firefighter in p+p for wip).She will be taken into zbrush to create normal map.
I downloaded a maxscript Skeleton Rig v1.73 tnat aids in posing model.
I want to do a test rig(first time with rigging)so i can make sure it deforms ok.
Is it ok to rig character before taking her over.
My workflow plan is as follows-
1.Build lopoly character in max 8
2.Rig character using maxscript and skin modifier to sort envelopes
3.Export model to zbrush subdivide model to add detail
4.Collapse tool to level 1 save under diff name (the reason for this is the mesh shape will have change slightly where i have added detail to model on higher subdivisions)
5. export back to max Uvw map/unwrap to get meshbitmap
6.export back to zbrush and add to previous tool(can anyone offer a little detail on how i add level 1 model back to tool?)
7.Use zmapper(1st time) to create normal map.

Please,if anyone can guide me i would be thankful

Toomas

For me atleast it seems like a poor workflow.
BTW doesnt ZB have issues with triangles?

I would:
1. build hipoly basemesh in max
2. export to ZB
3. detail
4. export detailed mesh back to max
5. build lowpoly useing the hipoly as a guide
6. UV
7. bake normalmap
8. rig

MoP

Yeah, I think the workflow Toomas described makes more sense - although while it is possible to work the way you're doing it, I don't think it allows as much creative freedom in zbrush or other highpoly areas.

Toomas is also right that ZBrush has problems with triangles (they cause pinching and make it hard to sculpt accurately), so if you're starting out with your game-res mesh and taking it into zbrush, you might run into problems unless it's all quad geometry.

kunglao

thanx for feedback.especially mop who has helped me with his advice in the past.
please allow me to eloborate.
1.I build my lopoly model consisting of mostly(if not all) quads.
2.once i am happy with mesh i export it to zbrush.Here i subdivide working on the detail
3.when i am happy i save tool then collapse to level 1 and save under diff name.
4. i then import in max and uvw map/unwrap level 1 model.My reason for mapping after zbrush not before,is that the lopoly mesh will have changed slightly so that i can then get the meshbitmap ready.(i know you can click crease button which will stop shape changing in zbrush)
5.take level 1 model back to zbrush and add to tool that has all levels and replace old level 1 with this one(if someone can expand on how to do this in zbrush)
6. use zmapper to create normal map(1st time with zmapper)
the only reason i added rigging at step 2 was that i lot of people who post their works do a quick test rig to make sure it deforms ok.I thought most people do this in case model does not deform properly
When i built my last model which used zbrush i followed much the same as above but uvw/unwrapped before taking into zbrush did the hi poly and brought back to max both level 1 and 5 models back to max to create normal map.
I experienced extreme slowdown when importing level 5 model to max(which was 750,000 polys),some recommended zmapper this way i do not have to export hi poly into max as zbrush/zmapper is ok with polys upto(1-2 million.
I am forever learnig and take on board all comments.
As said this is first time using this workflow and some new features(peltmapper,maxscript skeleton rig,zmapper).All my work is aimed at creating videogame characters and i aim to emulate a work flow that is simple,fast and effective.

kunglao

MOP can i pick your brain with one more question please.
My Firefighter model is ready to go into zbrush.
At moment i have attached head to body Welding verts around neck
and i have attached Backpack,axe,gun and belt to together.
My plan is to firstly take head/body as one into zbrush
then when complete do the accessories seperately.If i want 3 meshbitmaps(body,head,accessories) should i detach head making it a seperate object and take that to zbrush,repeat for other parts then when complete take back to max uvw/unwrap in max export back to zbrush replace level 1 and get normal map using zmapper.Finally attaching all together in max.
Sorry for endless questions but aside from the web i have no way of learning
thanx

Zergxes

You'll probably get fewer ray errors if you do things in seperate pieces, I think Daz had a good pic up of it.

I've had good luck with a program called xNormal by Santiago Orgaz. It has fewer limitations than the zmapper plugin.

I need to improve my workflow, too. Your goal is inspiring.

Good Luck!

kunglao

Let me break down my proposed workflow,i need your advice so i can get started.Please note this is my 2nd go using zbrush having created an earlier character called Mangler which was basically a simple human form with a fish head(kinda like creature from black lagoon).However,this character is a bit more ambitious.
Proposed workflow-
1.make lo poly in max (mostly quads)
2.export to zbrush
3.Make subdivisions,adding detail,the highest sub division will be used for the normal map created with zmapper(1st time with zmapper)


Q1.Now before i export to zbrush should i unwrap/uvw map my model in max.Surely if i do this the model will change shape(once detailing in zbrush)
albeit a little thus affecting the mesh bitmap,i thought i could save tool with all sub divisions.Collapse down to level 1 save under diff name and export back to max to uvw/unwrap then bring back to zbrush and replace level 1 of the sub divivded model.If this is possible can someone explain how i would do this?
Q2.When i sub divided the model once, the hair also changed shape,I do not want the hair to change as it is just going to have an opacity map applied to give the detail.I tried using crease but the hair in level 2 would not revert back to the original shape.Should i detach the hair and leave it in max or does it matter that the hair changes as it reverts back to original shape in level 1?The same also applies to the eyelashes.However i did notice that although i could add detail to hair i could not get behind it and add detail to head.

Have attached pic of lopoly model from max.I have detached the eyes and the planes for the teeth(which again will use opacity)The accessories will be taken into zbrush seperately and will have their own normal map
Really appreciate all advice

kunglao

I have MAJOR problem-Below is screenshot of same model.
Image to left is the model imported to zbrush ready for some sculpting.
The image on the right is the same model but it has been subdivided to level 2 when i drop the level back to 1(shown) i am left with an altered model(yet no sculpting has taken place).
Question-
How is it the model has changed and how can i get it back to orig.
Let me explain further.
The front of the hat will have a number and other detail added to it but i cant do this as the shape has changed.
The glove is another example.There are no faces connecting the arm to inside of glove and as screenshot shows it appears to have pulled away leaving a gap.
Please can anyone explain and offer their help,as i am eager to get on

Husch

The helmet model changes its shape because the smoothing algorithm in zbrush (and any other 3d App) tries to change a linear shape into a curve by adding one or more edges between two edges. To keep hard edges you have to add extra edges parallel near to the original edge. A good tutorial which explains this principle of subdivision modelling you can find here:

http://www.secondreality.ch/tutorials/modelling/head/head4.html

at the bottom of the page the autor explains how to set close edge loops to keep the shape of around a human eye.

quote:
A close row of edges help sharpen the edge around the eye when subdivided.
note: the narrower two edges are the sharper a crease you will get when smoothed. This is very helpful to create bevels or corners and get away from that typical subdivision "molten wax look"

kunglao

thanx for help
got this link from good people at zbrush central
http://206.145.80.239/zbc/showthread.php?p=213838#post213838
a plugin called smartdiv
This tiny script retains the proportions of a model when subdividing

Toomas

[ QUOTE ]
a plugin called smartdiv
This tiny script retains the proportions of a model when subdividing

[/ QUOTE ]

But doesnt it make subdividing kinda pointless when you are left with essentially same mesh only with 4 times the polycount. Now you have to make the round parts round by hand, which imho is much more work than modeling so that the model looks like you want it to look after subd.

But if you need to do the stuff your way then more power to you

Husch

Cool, thank you for the link to the Zbrush plugin.

But I also recommend to practice close edge loops to learn subdivision modelling.
Sometimes it is faster to do mechanical non organic high poly models in another 3d app
than Zbrush.