thanks!
played with the applink to slap some rough textures on my test model.
:eek: WOW thats Awsome!!!
Lovely tip for future reference.
next up, baking the ambient occlusion into the map (another app) and then combining it with the zbrush cavity pass. Thanks ofer!
-b
Awesome skull spiraloid.
Very cool scary model, cant wait to see it on its motorcycle riding along an highway…hmm…or not…but still very cool scary model. 
Very cool character !
I like the last picture (I prefer the render on the right I d’on’t really like the close up on the left)
Keep up the good work !
Gopher
yeah its funny how working with an ortho camera means it always looks better in the ortho render. hopefull we get pespective zbrushing in windows soon. 
-b
Yes, maybe even by next weekend…or not…
Hi Bay,
I´m sure you don´t need clothsimulation to knock out a
nicely folded cloth but since you are doing alot of belt and
buckles, clothsculpting and such, how about trying that trick
on highly subdivided (quad) clothmeshes?
E.g. solving a cloth using a somewhat interactive syflex solve,
importing a highrezmesh into Zbrush and reconstructing the
subdivision? Could be a great start as well?
Cheers
tim
P.S: I´m playing around with this currently, that´s why I butt in.
http://www.zbrushcentral.com/zbc/showpost.php?p=271207&postcount=30
Never understand why people don’t like modelling in the ortho-view !
It’s the only way to get your proportions right !..right ?
A proportional-model -technical or organic - is measured by its relation of the components-arms and legs or pistons and rods- and these spatial relations are best depicted in ortho-view !..a model that looks good modelled in perspective with a 28 wide-angle lens looks just as funny if you see the model
with a, lets say ,45 lens ! …and you can’t have perspective without a camera/eye of a certain size !
jantim
Tim, like your cloth example. I’m moving onto clothing next. I’ll show it when I get something.
Jantim as for ortho. we see in perspective. even the most beutifull person looks weird in ortho. if you were trying to sculpt her by eye, not by measurement, then you’de be “flying blind” so to speak if you sculpted only in ortho. I’m definitely not a paint by numbers kinda guy.
Hi Bay,
here´s a 160kB *.obj samplefile, the solved clothpiece for the arm,
I tried the reconstruction in Zbrush, allows stepping back one
subdivision but isn´t perfect. Had insufficient memory warnings
thrown and also crashes. I guess it´s because I didn´t use an
evenly laid out quad to create the cloth from but subdivided
a somewhat messy mesh to start the clothsolve with…
http://www.hafenlola.com/downloads/Leydecker_highrezclotharm.zip
I guess Pixolators workflow will produce better results if you
start with a regular quadmesh, maybe even do the subdivisions
in ZBrush, export, solve on that and then later swap out the
highersubdivisionsteps (affected through the solve) instead of
trying to reconstruct?
Cheers
tim
P.S: I´ll mirror this post in my thread and´ll add a link to your
previous posts with Pixolator, for the great info provided.
We see in perspective, true…but the best portrait-photo’s are made with a tele-lens to eliminate most of the perspective distortion… and great looking nazi-skull you made, and it’s a bonus that Ofer reacted in this thread …i must be a good thing for us ZBrushers ,waiting for a certain update, if he has more free time
jantim
Ortho’s work great for front/side/top etc… view, because you’re not worrying about perspective in those views.
However, as soon as you start getting into 3/4 views, you loose all sense of appropriate proportion due to the orthographic view completely killing any “natural” perspective we see in a 3/4 view. It’s hard to judge proportions and how they relate to each other in an orthographic 3/4 view.
That’s probably my biggest pet peeve of Zbrush as of right now. The Mac version has perspective sculpting, although not exactly what I was expecting, but hopefully Pixologic will give us the option to adjust the perspective view in 2.5 like you would in other 3D apps.
Ofer’s tip is something I’ve been doing for awhile. Although I use Silo to crease edges so it works a bit different than what other apps might do, i.e. it creates extra edges to simulate creasing. So I generally can’t reconstruct all the way to the base cage because of the extra edges it adds. What app are you using to crease your edges in Bay? Mirai?
Must be one of those things then, cause my favorite portrait photos are done close with a 50 lens, I only ever use teles in stadiums to shoot music concerts and football games.
I think THE thing to use for 3d models will be this, when its ready that is:
http://www.lightspacetech.com/
in a somewhat near future…and you can use it both in ortho and perspective. Now that would be an update heh? 
tim, nice. yeah I had some of the same memory errors and crashes. I’m trying to learn the rules about what can or cannot be reconstructed… hmm looks like you can have quads or triangles on the base mesh. no 5+ faces, which makes sense, since they subdivide into a triangle- which is the failure for reconstruction.
Giantsun, that volumetric display is rad!
chad, used wings for that one, but I also use maya and bit of xsi.
as for perspective, exactly right about telephoto and shooting supermodels. still, there’s a difference between shooting from a satallite vs, shooting across a warehouse lightstage right?
Word of advice about reconstructing: I’ve found that the Zbrush displacement generation tool will produce weird results on thin and sharp features, when you are creating the map by comparing the level 1 mesh and the level 5-6-n mesh. We’ve had a lot of problems with this on Warhammer initially, as we’ve tried to allow the Zbrush sculptors to modify proportions too, instead of only detailing the models. Sword blades, armor edges etc. got inflated most of the time, characters had bulging teeth, and so on.
However, if I’ve stored the original unsubdivided control cage as a morph target before sculpting, and then re-activated it for the map exctraction, then the maps were correct.
So here goes the problem: you import a highres, creased model and use the reconstruct lower level feature repeatedly to generate the level 1 mesh. I haven’t tried it yet but I’m quite sure it’d mess up the displacement map generation. Point order is also messed up, which means you can’t just import the original unsubdivided cage. So you’d be kinda stuck again - now you have a good Zbrush mesh, but how to get good maps?
Suggestion: keep your old, unsubdivided mesh in your modeling ap, and import the level 1 Zbrush mesh on top of it. You can then activate vertex snap and move the points back to the original place, export and reimport to Zbrush and store it as a morph target for map extraction.
But of course this is a good old waste of quality time… so we have a nice little Maya plugin that can snap meshes with identical topology but different vertex numbers. I’m sorry that I can’t give it out, but it’s probably not that complicated for a Maya API programmer, and this is such a good tool for many cases (fixing messed up blendshapes was another good use).
got exactly the same problem you experienced LY…lucky you got that plug :)…I am doing exactly the same thing and the mesh explode once you bring yur morph mesh into the scene and try to go back to L2