ZBrushCentral

Zbrush2 and Silo

ZBrush Artworks
#1

Just created this ugly guy to test Silo and Zbrush together. Basicly Im
using Silo for blocking my form here and Zbrush for all the sculpting work.
Its nice to have a seperation between the two task. It allows you to focus
on one then the other for beter results overall.
In my opinion edge flows should not drive the sculpt at all! And the concept of
edgeflows is just a temporary shortcomming in computer graphics. Personaly
I dont like it in my way when Im thinking about character and form. Silo and
Zbrush will let you work this way.

Dave

![faceRenderInter.jpg|1000x800](upload://iDLwW6LrO06Nmrv3fiXQaQ0R1y.jpeg)![faceMeshInter.jpg|960x623](upload://76u14KATH3NdipP9RiMulY1s4J0.jpeg)
#2

Cool model! Am guessing you get to test out that new topology brush and am wondering what you think of it? Is it as good as idea of it seems to be?

Also I notice the new topology has triangles…do you have to do anything extra to import it to zbrush in order to subdivide well or do they hold up well on their own?

#3

Oh my gosh! That guy should be in a movie. Great stuff. 50.

#4

Yep, another good looking model from Dave…i saw on your resculpted mesh
that you made a hole for the nose/nostril fold, do you do that to have a sort
of landmark when you sculp in Zbrush ? If i understand your theory you should
no longer have to do things like that in a polygon-based program,…just model
it in Zbrush !

jantim

#5

I really hope this one gets textured.
Amazing to see that an experiment/playign around can produce models of such quality and potential.

good luck :slight_smile:

p.s., what are you using to render the final model?

#6

wow, and so this model would be considered animation ready because of the redrawn good topology?

I remember you saying not too long ago, that your last 2 production models were zbrush models that you created (from zspheres i believe), then used the zbrush mesh as a base to trace over with new topology in another program so that it could actually be used for animation and further sculpting, and you said that it could sometimes can take months to do that retracing.

With this new workflow, Silo lets you trace out a new mesh and it takes an hour instead of a month now, so could you in theory use silo (for it’s topology drawing feature) and zbrush alone to produce animation ready production models? that is to say, use no other software aside from those 2 to make a film quality model?

last thing…about how many polys is that final mesh?

thanks

#7

Well done friend! :wink:

#8

Good model . seems to me that the final silo mesh is a little too dense. I don’t think you really need to end up with a so dense mesh, since normal displacement maps work very well even with lower polygons. and animation will be really easier with less polys. however, the final shape u built is really great and the Silo/Zbrush orkflow works really good.:smiley: Silo has great tools and an even better price. good modelling solution!!

#9

Having some Tris in your mesh isn’t a problem - N-Gons are. ZBrush will tesselate the N-Gons to Tris or Quads when you try importing geometry - And that can give you some unexpected results.

Just try to only use quads and maybe some tris (if you can’t get rid of them or whatever) and you won’t have any problems.

@Dave: Great model! Can’t wait to get my hands on the topology brush, seems to be a great feature! :slight_smile:

#10

Excelent work! I’m so looking forward to Silo 1.3 it’s not funny…I love seeing all the new and cool pics from you “secret beta testers” :wink:

#11

if u don’t like Silo, try and catch the incoming Modo, modeler from Luxology (www.luxology.com) it seems to have great modelling tools, inspired by Lightwave 3D’s modeler IMO Modo will be a great tool, and i guess it will work well with Zbrush too.:wink:

#12

OT (Sorry!): Modo does look pretty sweet but for the price I feel it will be out of the reach of hobbyists. Silo and Wings3D can do pretty much what modo can do and Silo’s dirt cheap and Wings is free. I’m looking forward to testing out a demo of Modo though. :slight_smile: I can wait to see what people will do with it since it’s always exciting to see new work from a new program.

Side note, Dave you’re the one who suggested the Topology brush tool to Nevercenter right? If so then…Thank you thank you thank you. :slight_smile: I havn’t got to try it yet but from what I’ve seen from you guys it looks so sweet.

pnoland

#13

another question, how much time did the whole process, from blocking out the form to sculpting the completed mesh, take?

#14

This process still takes time. I took about 10 hours total time.
And that was after I had done several sketchs to design the guy.
It still took me about 2 hours to draw the new mesh ( you have to think as you go ). The rest of the time I got to have fun sculpting.

There is a time savings plus the real bonus was that you get to base your topology off your blocked in form. And the re-draw of the mesh was not such a
pain in the you know what…

  1. Design and general forms, proportions blocked in.

  2. Construct/Draw new useable topology considering animation needs.

  3. Create UVs then continue sculpting in medFreq detail. The hiFreq detail to be painted in later.

In the future topology will be highly automated and or completely obsolete.
Fixed numbered “meshes” are a pain for modeling ( like blend shapes, and general changes ) textureing ( UV-ing and UVs after modeling changes ) and rigging ( updating the skinned model after design/topology changes ).

I see our Weta Workshop guys do amazing things in clay all the time that you
would never consider doing while working all in CG. Thinking about topology and your character at the same time constrains your creativity. Everyones work suffers because of it. I welcome any digital tools that alow me to seperate the
two task of sculpting and topology.

Dave

#15

Hey Dave!

Well said, but personally, I can’t imagine that topology won’t bother us in the future.

As CG Artists we are not constrained to static meshes (Since ZBrush I always have to think about what DaVinci, Michaelangelo would have done with todays possibilities…) - And therefore, we have to think about the ways that skin moves, about “the underlying structure”, we have to know a lot more than a traditional sculptor. Sure, we can do our work a lot faster, but we also have to think about 10 things at a time - and I’m absolutely with you about the “thinking about topology and your character at the same time constrains your creativity” thing. Absolutely, 100% agree.

How often did you trash models just because afterwards they wouldn’t perform well? How often did you start a model from scratch because the loops didn’t work out the way you wanted them to? How often have you canned models because of all the technical things to think about while working? It’s really frustrating that we have to think very technical while attempting to be creative.

I think the Topology Brush is already a neat way to sepereate both things - Shape and Topology. I think that’s one of the solutions that will help us getting all these technical things out of our minds - I really hope that Topology will become an automated process, but I personally don’t see that happen in the forseeable future. Sculpting has gone a long, a long, long way… and with Digital Sculpting there are a lot of other things to take account of - If we want to animate our work, if we want to give live to it, then we also have to do the “extra work”: taking care of the topology.

But that’s just what I think, if someone could revolutionize the industry, then you Weta guys! (Oh, and of course Pixolator ;).

It’d be great not having to think about “the stuff underneath” though!

#16

Yeah, the topology tool looks to be a great asset for modeling/sketching figures. I’m anxious to get it and try it out. Especially since you can theoritically draw over various objects, and create a new model from that. Looks like fun.

But, for all these improvements in modeling, I still feel like the “other” modelers friend, i.e. UV’s, are still not at the level they need to be. For me, I find dealing with UV’s a lot more of a problem than dealing with topology. Zbrush has made some incredible strides as far as helping folks out with UV’s, but their implementation is not always desirable. AUVtiles, and GUVtiles, aren’t necessarily friendly to game engines, and it also makes file sizes incredibly large because of the number of UV’s. Plus, it’s almost impossible to edit the AUVtile images in another program. Leaving you with little options for “tweaking” your final textures.

What I would like to see is something along the lines of Blender’s new UV tools. From what I’ve seen, it looks pretty powerful, and it seems to be pretty effective. I just haven’t had the time to learn how to use Blender effectively.

#17

Well, Modo will probably be a part of a more complete app, despite being sold as a standalone module too, just because Lux announced they’re working on a sort of platform, called Nexus, with also animation tools and a rendering engine too so probably that app will be part of a more complete app. I too am waiting for a demo of it though.

BTW topology tool seems to be great as an idea, cause it makes the work easier. after all we are not all like Taron wich is able to model a character like this one in less than an our for sure.

I agree also on the topology discussion. getting rid of topology would be very good. but i don’t see this to be so near in time. In fact, differently from ZBrush, current apps can’t handle millions of polygons in animation and so normal maps have to be used to detail your low/mid poly mesh coupled with subpixel displacement for rendering.

Consider also that in the animation process , animators use very low poly cages of characters to speed up the animation process.

If the future will get us rid of these limitations then it will be a bright future for sure. :wink:

#18

Congrats Dave.
Amazing model! :+1:

#19

hey dave great work , so let me ask you how was the performance when you ( if you have done already) import a 2-3 million poly mesh from say zbrush into silo? i mean is this going to become a cheaper approach than say paraform or cyslice?

#20

I couldn’t agree more with skycastle on this. Recently I’ve discovered a similar method for working out in Zbrush the actual design and ‘base-mesh’, and worrying about topology at a different stage in the process.

Using this method doesn’t eliminate the process of creating good topology but it does remove it from the actual sculpting part of the pipeline.

The way this works is by creating a very simple mesh in whatever 3d package which will serve only two purposes:

-having something to work with in zbrush which has the rough proportions that you need, but not necessarily clean or perfect topology

-having the right ammount of points in the places where the smallest details will be created… so in this base mesh you would put more edgeloops around areas like ears, eyes, etc…

After this you spend most of the time in ZB, not really thinking about anything but two things: design and form. This actually lets you use Zbrush as a conceptual tool where character and design can be explored, taking the technical aspect almost completely out of this step. Doing this you are not limited by edges or subdivision smoothness, etc…

When this step is complete and you have a final sculpt with all your detail, you can then export a mesh which has around 70% percent of the density of that final model (something that has the structure of the model all the way up to medium freq detail, but not small things like veins or pores, etc.) into XSI.

At this point the model can be re-parametrized with proper edgeloops and clean topology that will help in deforming and texturing the model correctly using xsi’s shrinkwrap deformer by simply extruding edges with proper flow that will ‘stick’ (hence the term shrinkwrap) to the high res model. This process can take a few hours but the real advantage is that your sole purpose at this stage isn’t form or design but topology and deformation. The actual form is there for you as a template. Then after your ‘good mesh’ is completed you can go back to your fully dense zbrush mesh and extract a displacement map from it (using a 3rd party tool like Nvidia’s Melody or ATI’s Normalmapper, in this case, since ZB2 currently can’t compare two unrelated meshes).

If I get a chance, maybe I’ll post a step by step of how this works.