ZBrushCentral

Frustration over generating a animation model

ZBrush Usage Questions
#1

I could really use some help.
I am somewhat new to the integration of zbrush with other 3d apps. I am relatively new to xsi, so there is a learning curve more in xsi than in zbrush for me.

I have worked through the digital tutors video re: texturing, displacement maps, normal maps in zbrush then to 3d app( I am using xsi).

So here I am with just enough knowledge to be dangerous:)

I have a model started in zsphere, got the basic full body and face proportions and a moderately detailed sculpt.

I have a few questions regarding a animation model only.

  1. WHEN IS THE BEST TIME TO RETROPOLOGIZE? ( if computer resources are not a concern) when you are preparing a ANIMATION READY CHARACTER?. Before creation of subtools? Before High level detail? What about getting the edge loops in the right place? Or should I just do it all including create clothing etc… and then retropologize at the end? Seems like models that end in a sculpt do this at a different time potentially than a animation model would.

  2. HOW DOES SUBDIVISION IN XSI WORK WITH A MODEL FROM ZBRUSH.
    Sometimes it seems as though some people export a low res model from zbrush, apply normal and displacement map, color map, and then animate. But if this was done, and you needed to subdivide some of the geometry in XSI does it all still work with the model? Seems as if you do not add any vertices, you are okay? I am really confused as to what people do here for the animated figure.

I would love to see zbrush sponsor a WIP competition for a 1 min animation of the superhero’s generated in the last competition or just with some zbrush models, maybe just a a 4-5 sentence diologue form the character with some small static standing poses. This would be extremely helpful, or if someone knows of a tutorial for me ( for fee or no fee) where they take it full circle with the CURRENT 3.1 version. Most stop at the export to 3d package and don’t take you the whole way through.

Thanks for any help you can offer.

John

#2

That’s material for a book or two or three.

Take it easy. Make a cylinder, detail that, export, import, rig, animate, add displacement maps etc. etc. etc. That will work quickly. if you start with a character, you will not go anywhere fast with your current experience.

The reason for the fact that after export most of the ZBrush tut’s go quiet on you is that the 3D app of choice tutorials are in place then.

Best of luck, take it easy and slice the problems into smaller bit’s you can digest. Seems you got the road map down, now you have to work your way through it. No quick solution there. Specially because different characters require totally different strategies. And experience will tell you what to use when and why.

With XSI you also have chosen an app which requires in depth proficiency to master it. TD level is about right. Mental Ray alone is a beast…

Cheers
Lemo

#3

Haha I agree with lemonnado.

Heres some info and tips I can give based on what youre saying, hopefully this helps.

“1. WHEN IS THE BEST TIME TO RETROPOLOGIZE?” You should only have to retopologize when you want to take a model that doesnt have proper poly flow and animate it. For example, if you started off sculpting a head model from a 3d sphere in zbrush, you would need to repologize the model before you could animate it, or else the deformations could be terrible. For a character model, you want to make sure you have the rule of 3 installed anywhere where theres a joint in the model, such as knees, elbows, hips, finger, etc…
What a lot of people tend to do if they want to animate a character is they usually build the basemesh with all of the edge loops and rule of 3 in a seperate 3d app like maya, xsi, lightwave, etc. then bring it into zbrush for high res sculpting. You can build edge loops such as the rule of 3 into a zsphere model, while laying out the zspheres just by adding additional zspheres in those areas.
Im not familiar with the way XSI handles models and subdivision but from what Ive seen in other 3d apps is that all your detail is stored on your displacement, normal, bump and texture maps that you generated in zbrush. For game res models, you can export the lowest res from zbrush, weightmap and rig the model and setup all the the maps to get prepare your model for animation.
You may want to check out some XSI tutorials on setting that up since that steps outside the range of the zbrush workflow.

…hope that helps some.

#4

I don’t agree the only time is when you are ready to animate. If you make something from zspheres or else you sculpt a ball heavily you will probably need to retopologize at some point just to get a better distribution of polies. Otherwise, sculpting will be much harder on some areas.

As for subdividing, no. Any changes you make that add geometry invalidate your normal maps and color maps and you can’t take it back into zbrush without a lot of work ie projectall and painstaking cleanup. You can still have divided versions in XSI but you have to keep the base the same or redo the maps and create the new base topology as another tool in zbrush and projectall fromt he other one onto it, which can be a lot of work.

#5

I here you on the books. I have intro to zbrush and also Scott Spencers book. Have worked through those items. And many a video tutorial on xsi regarding the same. That is why I am frustrated. I feel like I should be able to take a character from start to ready to animate without feeling palpitations. I have imported many a cylinder, and have a pipline work flow that is cresting 3 pages to get all the kinks out suggested by books and videos.

I do appreciate you saying that it is a subject of 2-3 books really!!! Because I was starting to think I was smoking crack with the disconnect between marketing and my ability to achieve what I wanted. I suppose the disclaimer on all big 3d apps and those that are used in conjunction should read.

" Please consult your physician before trying to integrate applications in the 3d animation environment. It is recommended all artists be technical directos with low blood pressure before implementing pipeline integration. As there is a inverse relation ship between the software pipeline widening and you coronary vessels narrowing ":lol:

#6

AA-ron,
I appreciate your comments. Wasn’t aware of the rule of 3’s had not come across this in all of the stuff I have watched specifically, though I have when sculpting from z-spheres pushed the zsphere back into the head for eyes, nose etc… to add edge loops. If you have a favorite reference to this please point me to it.

Nice to know that many do the base mesh in xsi or whatever. I had heard both a lot ( some start in zbrush and some main 3d app). I really like the zsphere to build a quick mesh, but having done that, and ended up with a mesh with way more points on it than wanted, and taking it back to xsi was a big mess ( unless I missed something).

Am I correct to say generally, if I built the mesh in zbrush ( as a carefree low blood pressured artist) and added subtools, clothing etc… I would need to raise my blood pressure by retopologize all pieces to make animation friendly, then project detail to that mesh, and import that mesh into xsi and apply maps created in zbrush within xsi. Is that correct. ( using subtool master where applicable)?

Now I am in xsi, and I want to change something I did on say the character body that affects a subtool. For all my texturing, normals maps to work, I am going to have to go back to zbrush, edit the mesh topology of character, adjust through subtool master, redo all affected maps, import back into xsi?

Another general question,
Is there a recommended film polycount or " active point count" I feel like with games I heard 2000-3000 and with film ?

#7

Interesting title. Did you come up with a action hero for this name–I dig it.
Regarding the subd surfaces.

If I did build something with polygons in xsi, subdivided through the + plus sign, and exported as a obj. Is it actually adding polygons or points? I thought it was just smoothing based on existing points. But this is what I am trying to clarify. What I am hearing is no. Thanks.

#8

For film, it’s what your renderfarm can support. If you don’t have a renderfarm you are going to have to keep things at game levels most of the time and use displacement or normal maps, maybe going up one subdivision level for final render. If you don’t mind it taking months to render you can go as high as you want once you have everything finalized, though.

Well, in maya you can go back to the original and export that. For XSI it may be the same but I would not take it for granted. You can also just keep two versions. If all you want is a higher division for renders then there’s no reason not to have separate versions. At least (again) in maya you can do that. XSI is pretty full featured so it’s likely you can do that there, as well. Then if you edit it in zbrush just resmooth to make the new high level version and discard the old one.

#9

Oh, and the name is due to what I have been trying to do for a while, grow more brain cells. Turns out it is possible to do that with certain herbs or drugs or nutritional supplements. Kind of new research. I feel much better than ever before, so it must be working to some extent. Maybe I will make it a company name or something some day, I like the word a lot, too.