That’s looking Wonderful and Intuitive. A dream some years ago, that comes true today.
You guys never cease to amaze me and your generosity providing upgrades is unparalleled.
I wish to personally thank each and every member of the Pixologic team!
"Whenever you’re freeform sculpting, it’s often hard to extend your mesh too much without using up the available topology – polys get stretched and it’s difficult to add additional detail. But with Dynamesh you simply make a gesture on the canvas and ZBrush retopologises the object, giving you a nice medium-density mesh that you can continue working on.
This function can also be used in conjunction with the new Booleans, or when using a subtool brush to join meshes together. Again, activate Dynamesh and your objects are glued together into a single mesh and any distorted or stretched polys are removed.
You don’t get a completely perfect mesh – ZBrush adds the occasional tri to maintain density and poly flow – but will be completely fine for most artists, and is great for doodling and adding large appendages such as limbs, with the same sort of freedom you get from voxel sculpting."
Source: 3D World Magazine
More info on how it works. Sounds awesome!!! Good job Pixo
This looks awesome, and it will be nice to have Sculptris functionality inside ZBrush, with all the other tools ZBrush offers (not to mention the same shortcuts). If I can have fewer programs in my workflow, I am all for it.
Looks like we’ll still need to re-topo ourselves, but I use TopoGun anyway.
I really, REALLY like the idea of being able to cut holes in dynamic clay, as well as being able to draw out new geo as shown in the video with ease.
Can’t wait for 9.20!
Read above you.
Hi Pixologic Support!
Quick question, with Dyamesh… Am I correct in understanding it to be a mode similar to how programs like Sculptris work? In that it’s essentially topology-independent, or tessellating the as-yet-determined base mesh until the sculpture you’ve made is at a point where you would like to “lock-in” (so to speak) it’s topology? I ask because often, I just wish I could just freely sculpt something with out having to worry or be mindful of topology at all and then just apply the proper topo without needing to retopo the ENTIRE, highly detailed model… because I’d love to just sculpt something -which I’d love to animate, but to just be free to sculpt it and then just wham, slap-on/very quickly assign the required topo for animation and then (after applying UVs, painting the textures, etc) just import it into an animation package and run with it, rather then get bogged down by having to, essentially, remake (retopo) my whole model again just to animate what I’m already happy with. =(
Who are you talking to, and about what?
This post answers alot of questions about DynaMesh AND it’s topology issues (or lack there of). Thank you cyboy!
The good news is that there is no Tessellation like in Sculptris. The occasional Tri can be dealt with. So the mesh doesn’t consist of ALL Tri’s, Sweet!!! I wonder if there will be improvements to retopology. This, alone is a result of that.
I didn’t see any Smoothing during the video. Makes wonder if the result of smoothing would be rough and puckered sets of poly’s.
The new Dynamesh feature helps to prevent one from running out of polygons, and also helps to prevent stretched polygons. It will also add triangles whenever they are needed. This all points to it being a regular polymesh based feature and thus is probably useable in all the ways a polymesh can be (ie: polypainting). I highly doubt Dynamesh is designed to create a low poly base mesh with good edge flow for animation in another 3D application though, meaning artists will still have no choice but to retopologize their mesh manually by hand when they’ve finished sculpting. I’m hoping the retopo tools will have been improved in Z4r2 to at least be a bit more intuitive, if not a lot less labor intensive, but somehow I doubt that will be the case. In addition to Dynamesh and booleans, a lot of the focus seems to have gone towards supporting HDR rendering (hence the updated mats and lights). Users probably won’t see any of the major improvements being requested (like 64-bit) until Zbrush 5.0.
The new boolean brush is certainly interesting too, and is reminiscent of the movies we’ve seen regarding vector displacement stamps in Mudbox. Same idea, different approach. While it is certainly a step in the right direction as a time saver (mix and match pre-sculpted body parts for example), vector displacement as it pertains to generating high poly looking renders (with undercuts) from extremely low poly meshes is still desirable as a replacement for normal mapping and displacement mapping (no undercuts allowed). It is pretty well supported in the latest versions of most renderers, including Mental Ray, Vray, Renderman, etc, and seems to work pretty well as far as quality and render time goes. Hopefully Zbrush will be able to generate vector displacement maps at some point in the future, even if the plan is to never use them as a brush and/or alpha.
http://www.youtube.com/watch?v=wMx-wBwimp8
For those wondering about booleans, my best guess from watching this video and the previous one is that it is simply a new brush. The options for this brush will probably allow you to select any loaded Ztool, which you’ll then be able to use to add, subtract, intersection, etc. This is likely the reason we see several basic primitives in the brush menu during the last video. They’re just copies of the main boolean brush, pre-configured as examples of how it can be set up and used. I wouldn’t be surprised if Dynamesh came about as a necessary part of making boolean operations possible, especially when it comes to things like gluing Ztools onto other Ztools which is where it would truly be helpful (provided we’re not seeing subtools with that ear).
I could be wrong about all of it though. Whatever the case, it will certainly be interesting to see how all these new goodies work when September 20 rolls around, and I’m really looking forward too seeing what other people create with them. My impression from this video was that it will probably help users who are new to Zbrush. You can just keep going at it until you get what is in your minds eye, all without worrying about topology or making undoable mistakes. Add in a configurable incremental manual/auto save feature, and you end up with a lot less frustration from not only new users, but seasoned ones a well. We don’t always remember to save when we should, especially after a long day of work (that occasionally has to be done all over again). 
There have been three times in 10 years that I have been this excites.
-
- ZSphere in 1.55b
-
- ZSketching in 3.5
-
- NOW!
#3 has been my dream for a decade!
Thank you Pixologic!
To zeddicus
Vector disp and booleans are 2 very different things and in practice
Vector disp is incapable o cutting holes or creating intersecting volumes
Vector disp gives u a gimmick of a Boolean union only when u stamp in clear you couldn’t poke a hole or even intersect a mesh with another operations that
Only some sort of solid or volume aware operation can do.
In a free design scenario u want booleans and vect displacement is simply not good enough…
I wonder if this will be useful in hard surface sculpting.
Agree 100%… + GOZ and UV Master 
cheers, djart
Still have no choice? That is assuming is a complete alternative is possible. Yes 3Dcoat does have an auto-retopology feature and while it is a very nice addition, it is no replacement. It does simple things rather well like basic cylindrical limbs, but the more you want it to reflect the definition, the less ideal it becomes.
Dynamic Mesh is a very massive update and likely took a lot of time and resources. A re-hauled Retopology system is so big a task it could probably be its own revision update. Can’t really fault Pixologic if it is the same, Dynamesh is far more important of a feature. I would far rather they focus on making sure it integrates smoothly than half heartedly address both.
Part of why I started development on RetopoG for Maya was because I knew a feature like Dynamesh was coming soon and I wanted to be ready to take full advantage of it in an ideal way for my needs.
Agree
Also Keep in mind that retopo is a feature that is needed only for a fraction of the overall modeling market (concept modeling,rapid prototyping for example has little use of retopo)
Also retopo tools are in fact extensions of existing poly vertex edge modeling tools already in the market while dynamesh or voxels are much more innovative and game changing
It makes total sense to invest rather in that that in something that is already existing or can be easily worked around (using Nex for Maya for instance is very simple and has being around for many years also poly tools in modo Are flexible and powerful enough to retopo very well not to mention 3d coat or topogun
If I was forced to retopo for an animation pipeline tool like Maya I will rather see retopo tools in this soft as afterAll it is that soft asking for specific mesh constraint and in fact they already have a polymodeling kernel with all the code to retopo
these last two posts actually start to make sense. auto-retop is a holy grail that like auto-unwrapping methods can be approached but will never (in my opinion) be replaced by doing it by hand the hard way. I wonder if those shouting the most about a better retopo process for zbrush are made up of people that don’t animate and don’t use traditional poly-modeling software? If that’s not the case please chime in. Also doesn’t ‘ReMesh All’ and ‘Project All’ solve topological problems from a sculpting standpoint? What’s better about 3dcoat’s auto-retopology or is it just a sexy one button fix?
Its amazing!!! 
It is a sexy one button fix. 
Well it does a good job where you only have to fix minor bits usually. So it cuts short the time it takes to retopo your organic models.
Mike
retopo is a relatively subjective topic…show me 10 riggers and i will show you 10 different topos…same goes with the idea of low poly…how low is low? well depends on what industry and what pipeline…even what level of detail you need is project,pipeline and production dependant and this is why if you use a renderman compliant renderer you might want to produce a disp map and a subdiv friendly mesh where if you use lets say MR you might want to put more detail in the mesh and if you are rapid prototyping you will model everything you need to be printed.
triangulated or not is also an arbitrary choice as it is perfectly possible to render triangulated meshes and has always been possible also some game engines will only digest tris as standard.
As you can see retopo is very arbitrary and comes in many flavors and assuming that you need polys (no one seems to mention retopo for nurbs for example) you can use many different ways and 3dcoat has an auto semi auto and manual tool for this.
the last topology scenario which remesh or dynamesh or voxels or dynamic tesselation solves is the cases where u need to add more volume and the polys stretch…again this is to overcome the limitation of poly cages which are no more and no less than the digital equivalent of sculpting with chicken wire…just like the real world equivalent it will be hard to merge,polish or cut with solutions like dynamesh we get much closer to the organic,resolution independent clay medium that also allows holes, cuts and a soft or sharp finish with unparalel flexibility