Ah that explains it.
I was confused about if I was missing it.
Well we learned modeling in Maya - and that zBrush texturing, and than back into Maya for animation. I looked into the zSpheres and it is an interesting concept - I can see how this might work for clay sculptors.
I just fired up Blender just for an test and used the Susanne head and a multires of 6 and it was stop motion. 
I noticed with some other displacement modelings I did that Blender is getting very slow with the poly count.
Here is the normal workflow using Zbrush. Create a base model outside of Zbrush, import it, detail it to the best of abilities. Then either retopo it inside Zbrush, or outside of it and import it back into Zbrush. You should retopo it so that it has good flowing edge loops, more polys in more detailed places, and also that you create a certain amount of polygons that when you subdivide you can hit your max poly limit right on the head, allowing the most detail. You can do this by finding your poly limit, then dividing by 4 repeatedly until hitting the desired base model polycount. Then start working on the character again, and fully detail it. Once this is done, retopo it once more, bake the displacement maps, and map it on to the lower poly creature.
TADA! That is how the pros do it, I know for a fact that that is how the guys at ILM and Blur do it.
that seems to be the also only logical approach to the memory issue.
I wish Blender would have better displ baker.
If for some reason you dont want to bake your displacement or normal map in Zbrush/Mudbox there are a few other apps, such as Xnormal, that can do the job without crashing. I know that the Xnormal route is pretty popular in the game industry.
Im not sure why they dont like using Zbrush for that, maybe it doesnt generate the best normal maps? Either that, or Xnormal allows more manual control of how the map is generated
i have Zbrush 3 , and i think it makes everything from sculpting to cool texturing and normal maps etc .
I wonder what mudbox can do that Zbrush canāt ? << a tame question because i didnāt see mudbox up to the moment . >> .
why would you do that extra retopo step in the end?
Why not either build a model with zspheres, then retopo once (with that subdivide by 4 rule, and good edgeloops in mind), and if needed (for uvās or something) detail that up again, and make a displacement/normal map. And just apply that to the first retopoād model?
I would guess the workflow in mudbox could be similar?
That would work as well, but I think that is too much restriction. They reason they would do it the other way is because they are allowed more artistic freedom in the beginning, to just sketch out their character; the first run through is just a planning phase to really get the basic character down, then retopo for the places where you know you want more details. It might be a bit more time consuming, but in the end I think it is better because you get to have that freedom in the beginning. The art shouldnāt be hindered by technology.
Well,
we do same in ID as well. sculpting a model in clay
laser scanning it and than retracing sections for surface
creation.
It makes sense when you are good with sculpting to use
zBrush and when done retopo.
Just curious about if Blender could at one point handle
the same amount of polygon. How is Maya actually in that
regards? I funny me never loaded such a big model into Maya.
claas
Well Iām a little late, but lordy, that vid was pretty darn amazing. Mudbox has definitely come a long way(I think its almost safe to say itās passed up zbrush).
Maya canāt really deal with many more polys than Blender can. Blender and Maya will eventually be able to handle that many polygons, but that is only when our hardware becomes super-computers 
Zbrush and Mudbox use a highly optimized polygon processing algorithm or something akin to that. It only really works for sculpting applications and the like, otherwise Autodesk would have implemented it into Maya and 3dsMax already. This is all guesswork by the way, so donāt take me up on it as the gospel.
sebbo
I think the same way in some respect.
I actually find Mudbox more interesting than zBrush. It seems more cut down to what you need.
Some of my friends mentioned that they were pretty impressed with the retopo tool in Blender and that it makes for them more sense than zBrush retopo.
However the texture and displacement map baking is in deed a key factor in zBrush.
To bad that the micro displacement rendering for Blender died.
Yes, many people are beginning to find Mudbox as a great alternative. Zbrush has a lot going on with the 2.5d concept, and sometimes that hinders the toolset and workflow. I tried playing around with the Zbrush demo, but I quickly became lost; the drop down menus and toolbars essentially make up the entire GUI, and until you find your way around all that, you canāt get any work done.
Mudbox on the other hand, is very very straightforward, because it only has 3d sculpting in mind. The worflow tends to be much faster from what I have seen.
Another thing that will really hurt Zbrush is that Autodesk is currently porting Mudbox to Mac and Linux. Having a native sculpting app in Linux will be enough for some studios to switch over, considering the amount of major studios using Red Hat on their workstations.
Also I havenāt played around with the retopo tool in Zbrush, but Blenderās retopo is really quick to use. I really enjoy that feature in Blender.
I looked into zBrush retopo and I have to say it is interactive but complicated.
Blenders is easier to use honestly.
One thing I wish blender would have is actually a smooth proxy feature similar to what Maya has because SDS shrinks the object. Smooth proxy keeps the it!
Having Subdivide Smooth in the SDS modifier but be good
For those wondering the existing retopo tool you may not even need anymore in 2.50, because Iāve seen some nice tests using the new shrinkwrap modifier as a sort of retopo tool itself.
Zbrush and Mudbox use a highly optimized polygon processing algorithm or something akin to that. It only really works for sculpting applications and the like, otherwise Autodesk would have implemented it into Maya and 3dsMax already.
I bet if the Blender devs. find a way to do that kind of processing you would change your mind about it being mainly for sculpting programs;)
I doubt it. Donāt you think that Autodesk already would have implemented it, considering that they own the code for Mudbox, and created most of the new code? It is not a matter of how smart you are to make the algorithm, there are some limitations in the way these high poly processing programs work.
Using the shrinkwrap modifier for retopo is definitely an interesting idea, but you should reread what I wrote about the 3d Suite-Zbrush-3d Suite pipeline. Although the shrinkwrap modifier is cool, I doubt it knows where to make edge loops, and it is certainly limited in the fact that you cannot put more detail into certain spots. Doing it yourself is always better, much more control.
Shrink wrap modifier is fantastic for retopologize.
mh
I disagree that I can replace a retopo tool.
but it can very useful to project a surface onto another surface.
but by its nature it is only good for convex, and not strongly concave
tubical objects.
he shows some cool applications here like cloth and curve projection !
http://wiki.blender.org/index.php/User:Jaguarandi/SummerOfCode2008/ShrinkwrapModifier
any his deformers are nice as well
i hope it will make it as well into the official release
http://wiki.blender.org/index.php/User:Jaguarandi/SummerOfCode2008/SimpleDeform
Heās looking for user testers for his deformer btw, if anyone is interest, please contact him.
Martin
theeth thx for the tip
I tried to google his email but I did not find any email for
AndrƩ Susano Pinto
claas
I read Tonās siggraph report and watched the slide show http://download.blender.org/institute/Apricot-siggraph-slides.zip Looks like Blender has some nice plans in the works.:o