Jump to content
3DCoat Forums

ZBrush R2


Recommended Posts

  • Contributor

How much thin? We don't expect miracles, of course we don't. Make them less thin, and thinner after retopology, in a nice surface sculpting room.

There isn't any other way except clothing which may need a different approach.

Thin like a vest collar...And no,I am not gonna make a thick vest collar just to avoid getting holes in 3dCoat,its just not nice.

There is no place for retopology in my workflow until lowpoly,i don't have time to do this 2 times.

Anyway it works great now,I remesh using mergetovoxel/densequads-export when I have multimillion detailed organic surfaces or hardsurfaces

and I use Dynamesh only for remeshing very thin surfaces,mostly clothings parts...

(clothings parts at this stage dont have fine details yet,I add fine details on clothes later baking Cubemap shaders bump maps on lowpoly)

Link to comment
Share on other sites

  • Advanced Member
Thin like a vest collar...

LOL

I know, I know... I always go for blender box modeling on these cases, No need to retopo, just surface sculpting in zbrush.

The easiest way, we like it or not. (I also don't enjoy this, artman, but what else can I do?)

Link to comment
Share on other sites

  • Advanced Member

Oh yeah

What BeatKitano notes is absolutely true.

And yes, 3dioot, implementation is everything.

For instance, the implementation of 'dynamesh like' code in blender looks the same but tends to make holes. It came up one year ago but it's still in pre alpha.

It's an open source, so someone like Andrew or Raul could have a look and explain to us, what is this after all? Voxels?

It's very elastic, smooth and easy to handle solution. Being quads, low poly, try to extremely grab a part. Watch some interesting clean lines by stretching. Remesh it and have some nice drapery like shapes.

Difficult to explain why is this so friendly to use.

What isn't so great in zb dynamesh is to try to cut some material. In 3dc using [E] and paint by surrounding irregular area tool. Hold ctrl and cut. It works in zb after lot of clicks, not so direct tool.

Link to comment
Share on other sites

  • Advanced Member

Hey, I knew this.

Thanks beatkitano.

The 3dc way looks still more simple.

What are we trying to say here?

That zbrush is the leader on this 3d sculpting world.

The leader!

Not technologically speaking, quite the opposite.

But, because it's a well organized app, close to artists needs and that's how these so great tools exist.

Let's have a look on this UI

We can start with a basic sphere, dynamesh it, add topology, merge meshes.

We can retopo this mesh, subdivide it, project it (works great) and have a multi res mesh.

we can work on ultra hi density res on this, we can also paint it (MV like), lot of tools here, lot of methods.

we can divide it to groups easily and ask for UV unwrapping (painting areas for denser mapping)

We can export excellent baked displacement, normal and texture maps.

This is a well organized app. A weird but well done UI. An effective one.

We lost Steven Jobs.

He didn't invented anything really. But he changed this computer's world. We'll remember him whatever OS we use.

Mouse, we all use a mouse...

And a windows system

Steven didn't invented these but he persuaded us to use them.

3dCoat has all this advanced technology but I have this feeling, they don't know how to organize it.

Why do I need LC when I have voxels? Why, every time I work in surface mode and press enter, these artifacts, what are these? Why in surfmode only tris? Why I can export dense quads? What are these logical inconsistencies?

Why I don't have an after retopo full operational sculpting mode? A multi res one. Like in blender!

Why these sculpting tools behave like crazy? This is the most important for me.

Link to comment
Share on other sites

Hey Michalis,

I agree with you. While programming is an art form, I would say UI design, workflow and ease of use (or simply UI Ergonomics) is another art form unto itself. People like us use the interface everyday and have specific requirements in our own personal workflows, and as such, know very well what we want and need out of an interface. 3DC needs a huge overhaul on the UI. I know Andrew doesn't like me to say this, but it is true. I've spent countless hours documenting the needed changes and how these things should work, but I am only one person and one person doesn't have the sway of a whole community for something this large of a change. If we could get more than just a few of us to really throw some ideas for Andrew on what to change and/or remove, and exactly how we want things to look and function, I am almost certain he will draw inspiration from what we offer.

For the documentation that written up, I'd love to put up for people to see so they can put their two cents in and hopefully Andrew will see what folks like/don't like and implement what is good about it. And I urge others to do the same.

Link to comment
Share on other sites

  • Reputable Contributor

I very seldom voice complaints, Working toward solutions is much better but I do have to voice a complaint at least once in a long while...

The Artifacts in surface mode "long standing error" after pressing enter can be frustrating. The black errors areas on displacement maps even with a good laid out UV map, "Long, Long standing problem" makes one growl at times. These and some other errors cause me at times to take a break from using 3DCoat. I will never stop using the program as 3DCoat is very powerfull and a good piece of software but me taking a break from using it is not good for 3DCoat in the business world. You never want your customer to stop using your product even for a short time, they might find something else...

Link to comment
Share on other sites

  • Advanced Member

MIchalis, just as a reminder: zbrush started development around 98 (first beta got out in 99 if I remember correctly, I was involved in early debugging), it has taken about 12 years to get zbrush where it's now.

I was not there at the debut of 3dbrush, but I think total dev time (counting before even voxel introduction !) is at most half that time... You can't seriously expect the degree of polish of zbrush and completeness of worflow to stand at the same level in so little time. I understand your will to push the program further but please try to stay reasonable, right now I see a lot of complains (and I agree with a lot of them), but restating them every other thread is not gonna make things happen faster.

Well said Beat.

Link to comment
Share on other sites

  • Member

For the documentation that written up, I'd love to put up for people to see so they can put their two cents in and hopefully Andrew will see what folks like/don't like and implement what is good about it. And I urge others to do the same.

Hi javis,

I would love to see this and to comment on it. I truly find 3D-Coat to be a really powerful piece of software. Even with ZB4R2 now in my toolbox there are still things that I prefer to do in 3D-Coat. Nonetheless I do agree that 3DC needs a serious overhaul on the UI.

I think it should be a top priority for Andrew to fix the UI issues and, of course, the long standing errors that Digman is mentioning.

What is happening now with LC and the Surface room is really interesting. However these tools are still under development, and nowhere near as usable or stable as equivalent tools available in Zbrush or Mudbox for that matter. This hopefully will come in time. But right now I think that some effort should be invested into giving all the tools that already stand out from the competition that final polish.

Link to comment
Share on other sites

MIchalis, just as a reminder: zbrush started development around 98 (first beta got out in 99 if I remember correctly, I was involved in early debugging), it has taken about 12 years to get zbrush where it's now.

I was not there at the debut of 3dbrush, but I think total dev time (counting before even voxel introduction !) is at most half that time... You can't seriously expect the degree of polish of zbrush and completeness of worflow to stand at the same level in so little time. I understand your will to push the program further but please try to stay reasonable, right now I see a lot of complains (and I agree with a lot of them), but restating them every other thread is not gonna make things happen faster.

Oh most certainly. However I do think that if Andrew spent \a few months on just the UI and workflows alone, it would be at a pretty solid place or at least a great jump off point to build from for future builds.

Link to comment
Share on other sites

  • Advanced Member

What I think also might turn out as a helpful experience was if Andrew and Raul took a week to do nothing but creating a complex model themselves..

This way they got an intensive hands on of how individual features interact and of the applications overall performance. They could check on

Tools accessibility inside the given GUI implementation. Not in isolated debugging scenarios, but after hours of use and after several taxing operations in a row.

My feeling is that it's one of the keys of Zbrushs polished Tools implementation that its original Maker is an avid user of the Software himself.

Link to comment
Share on other sites

  • Contributor

What I think also might turn out as a helpful experience was if Andrew and Raul took a week to do nothing but creating a complex model themselves..

This way they got an intensive hands on of how individual features interact and of the applications overall performance. They could check on

Tools accessibility inside the given GUI implementation. Not in isolated debugging scenarios, but after hours of use and after several taxing operations in a row.

My feeling is that it's one of the keys of Zbrushs polished Tools implementation that its original Maker is an avid user of the Software himself.

I agree 100%.

Model does not need to be "artistic",it just needs to be complex,and use various tools.

I understand its our job as testers to give feedback but Im sure it would help Andrew a lot to see issues in a workflow by himself.

I also understand time spent trying to make a complex character is less time spent working on fixing bugs or adding new feature

but in the end I think it can benefit more,trying to reproduce other people's issues can surely sometimes be boring but when you see issue by yourself in a global series of actions it can give a better understanding of areas to prioritize and streamline for users.(I think it could even be more fun for them to do that just a few weeks)

Link to comment
Share on other sites

  • Advanced Member

3d Coat is digital perfection. The End. :drinks:

Not really.

To be honest I liked more the first betas(when Andrew started the work for voxel sculpting),with more features the program becomed slower and less smooth to work with.

For me it didn't help at all the discussion about better brushes in voxel mode(now we have a lot of brush stuff,but it's really useful? I don't think much)

Sometime less is more.

Link to comment
Share on other sites

  • Advanced Member

These last posts became a little out of topic. I have to apologize.

Just watching all these new tools (LC) all this implementation... I wonder.

It maybe took years to have these excellent well tuned tools in zbrush.

How many years for sculptris? Because it still happens to be my favorite on this.

Sculptris has indeed simple yet great tools.

Just found this, have a look. Maybe not the greatest artist but...

http://www.zbrushcentral.com/showthread.php?92849-DrPetter-s-sketches

Link to comment
Share on other sites

  • Advanced Member

It 'a pleasure for me to follow such debates, looking in detail the best solutions for virtual sculpture and analyze the differences between programs to improve even more.

I hope that efforts in 3DCoat developing are not spent too much in a tool like Dynamesh. Do not misunderstand, it's good that there is something similar to dynamesh, but it is not the perfect solution.

Dynamesh is a GREAT tool, a step forward in virtual sculpture, but has a major inherent limitation: it is not possible to maintain a high level of detail while sculpting, except if the hardware makes a big effort.

Sooner or later you must deal with this frustrating limitation. This is not perfect freedom to reach.

If, to try to keep the detail, you activate 'reproject' the calculation times double or triple...

So Dynamesh is based on an idea that is evolving for the better, but dynamesh method has this concept limiting from the ground up, at least until somebody finds a technological trick to preserve detail but without excessive loss of computing time and effort to the computer (without effort could be=3/4 seconds using dynamesh + project with objects with many millions of polygons)

I do not know where in future this technology can go ( dynamesh or similar in ZB and 3Dcoat) without help very powerfull hardware.

Probably its positive evolution ( the detail preservation in every step of work) will be through the implementation and interaction with the idea which Sculptris based on.

Sculptris, whether a coincidence or the result of the mind of a genius, in my opinion currently offers the best technological paradigm for virtual sculpture: you can immediately detailing some part and and leave the rest in low poly with a very effective overall appearance. Immediately you can change the topology and using the button 'Reduced selected' for example you can press it 3 times and reduce about 3 times the number of global polygons to appear without a visible loss of detail. Because, even with 3 times less polygons, they remain 'printed' in their original shape, with good approximation. But this is also however a limit.

If Sculptris can manage already few million polygons would be already fantastic, but unfortunately more than one million it crashes. I think this is really his biggest current limit, which limits its use in combination with ZB.

Besides the fact that ZB speaks for quad and Sculptris speaks in tris and therefore they can not 'communicate' with maximum efficiency: if ZB/Dynamesh work also with triangles only and if Sculptris manage 3 million or more polygons you have a workflow sculpture (ZB> <GoZ> <Sculptris) which would give a near-perfect freedom. Even better, of course, if Sculptris had 'liveclay' quad.

Link to comment
Share on other sites

  • Advanced Member

It 'a pleasure for me to follow such debates, looking in detail the best solutions for virtual sculpture and analyze the differences between programs to improve even more.

I hope that efforts in 3DCoat developing are not spent too much in a tool like Dynamesh. Do not misunderstand, it's good that there is something similar to dynamesh, but it is not the perfect solution.

Dynamesh is a blocking out feature - it is not intended to be used for final detailing.

Zbrush already had awesome tools for adding crazy Detail at but before R2 it lacked a simple Tool to cut Holes through the entire model or to melt different meshes together.

That problem is (quite awesomely) fixed now.

3DCoat in tendency has the opposite problem: Lots of powerful tools for slicing up and melting back together at low to moderate Polycounts but thus far no mature finishing

Tools inside the Voxel-Workspace.

Link to comment
Share on other sites

  • Member

Zbrush already had awesome tools for adding crazy Detail at but before R2 it lacked a simple Tool to cut Holes through the entire model or to melt different meshes together.

That problem is (quite awesomely) fixed now.

3DCoat in tendency has the opposite problem: Lots of powerful tools for slicing up and melting back together at low to moderate Polycounts but thus far no mature finishing

Tools inside the Voxel-Workspace.

Polyxo I think you are absolutely right. Dynamesh is an awesome addition to ZBrush especially for those that do not have 3D-Coat. however I still think that 3D-Coat beats ZBrush when it comes to adding and subtracting from volumes. In ZBrush it is still not possible to "dig" holes in a volume using only brushes. Just like in Sculptris you are still battling the limitations of working with a mesh. But you are also right that the finishing tools in 3D-Coat have quite some way to go before they are able to compete with those in ZBrush. Then of course there is also Retopologizing and unwrapping. Two areas where I personally find that 3D-Coat is still far ahead of ZBrush. So by the end of the day I'm glad I have them both.

Link to comment
Share on other sites

Voxel sculpting, in 3D-Coat, is gradually being emulated and enhanced, in function, by LiveClay - but won't ever entirely replace voxels in every way.

Once the "standard" set of features are in place, users will be able to start and end their sculptures by means of polygons. It logically follows that both topology tools and "high frequency" sculpting tools exist in that same workspace - the high frequency tools capable of being used on an "organized" quad mesh, suitable for animation and texturing.

In this way, an overdue streamlining of the sculpting, topologizing and refining processes can be organized into one space.

Greg Smith

Link to comment
Share on other sites

  • Advanced Member

Dynamesh is a blocking out feature - it is not intended to be used for final detailing.

Zbrush already had awesome tools for adding crazy Detail at but before R2 it lacked a simple Tool to cut Holes through the entire model or to melt different meshes together.

That problem is (quite awesomely) fixed now.

3DCoat in tendency has the opposite problem: Lots of powerful tools for slicing up and melting back together at low to moderate Polycounts but thus far no mature finishing

Tools inside the Voxel-Workspace.

Yes, unfortunately dynamesh is not intended to fine detail, at least for now. And it is true that, as you say, already from previous versions has tools to add details in an impressive way.

But in my opinion, with the arrival of dynamesh, things have not changed much from the point of view- of only workflow modeling/sculpture, for example, a hole in the mesh did before in an external program (ex: Maya) now, with dynemesh, do internally in ZB, and that is good. But you must do before adding a lot of detail.

This means that if I wanted to append to a very detail and finished monster, for example, a second head, also that full of details, so you create a unique mesh (and not just a subtool added) I can not do that with dynamesh method because they would lose the details of the first and the second head and body as a whole.

In practice, Zb, and despite gettin holes and meshes connection (always under medium resolution) is still based on the old system procedural work: 1-Sketch 2-medium detail 3-high detail. So conceptually and technically not from the maximum of freedom to the artist. Although of course you can achieve amazing and beautiful model since the previous versions.

In marble sculpture and partly in modeling clay and plasticine you are forced to proceed with the same methodically consequential 1-2-3. But I think that in the virtual world, the artist may have much more freedom to go back and forth between the work phases without, really and completely, worrying about the number of polygons or lose detail in at every stage of sculpting work. In the near future, maybe.

(sorry for my English, I hope that I explained what I meant.)

Link to comment
Share on other sites

  • Advanced Member
This means that if I wanted to append to a very detail and finished monster, for example, a second head, also that full of details, so you create a unique mesh (and not just a subtool added) I can not do that with dynamesh method because they would lose the details of the first and the second head and body as a whole.

a.

1. Just merge down the two detailed meshes.

2. Duplicate the 'tool' under layer like tools palette.

3. Remesh the first one, subdivide it and reproject all. Same principles as in retopo methods. :drinks:

b.

Dynamesh can work up to 2-3 M in full resolution and having powerful cpu. Just enlarge the mesh. The first trick I found, (reminded me the enlarging like tool when merging in voxels lol)

c.

Dynamesh is more powerful than 3dcoat voxels when inserting objs from shared libraries.

Just use SketchUp free, the obj free exporter for this and connect to google 3dwarehouse. Most of these fine quality models are more or less problematic. They contain lot of reversed normal faces, lot of non manifolds, lot of holes (non waterproof). It will be a disaster if try to drop them to voxels. Not so for dynamesh ! It will close the holes and won't pay much attention on non manifolds or doubled vertices.

Melt anything on a single mesh and add or cut as much clay you like, fast, using two or three tools only.

Maybe some people don't like the ZB UI interface. I'm not one of these. It's effective and fast.

This took me 15 mins testing all these and having lot of fun. But don't ask me to retopo this... :)

music.jpg

Link to comment
Share on other sites

  • Advanced Member

After using dynamesh on a few things I still prefer voxels as a blockout tool. The flow is still faster in 3dcoat since I'm never really increasing the resolution until I'm happy with silhouette. I found I was babysitting my mesh resolution far too much with dynamesh, whereas in 3dc I never think about it. Zbrush continues to have the more refined toolset in several areass so I'll probably not really change my workflow between these apps with R2.

One thing I much prefer with dynamesh over voxels is how the transpose tool can be used to duplicate meshes on the fly, and depending on the mode (add or subtract) you get a boolean without breaking your flow. This applies to mesh inserts as well. The workflow in 3dcoat could be streamlined in this regard.

Link to comment
Share on other sites

  • Advanced Member

@ jedwards

I have a quite opposite opinion LOL

I find dynamesh much faster and more spontaneous.

I prefer the 3dcoat merger, booleans are very fast, subtracting holding ctrl and I much prefer the gizmo. I also prefer the cuts using [E] modes.

There's a similar method for cuts in zbrush but again, instructions are for the birds. ctrl+shift for hiding, delete hidden, turn off dynamesh (use the button only) turn it on again. LOL it works fine, like in 3dc. But some more clicks.

Link to comment
Share on other sites

  • Advanced Member

a.

1. Just merge down the two detailed meshes.

2. Duplicate the 'tool' under layer like tools palette.

3. Remesh the first one, subdivide it and reproject all. Same principles as in retopo methods. :drinks:

I do not understand what the difference between your method and use classically dynamesh, like this:

1 - Merge down the two detailed meshes subtools

2 - choose the resolution you want and turn on dynamesh

The result is not the same?

b.

Dynamesh can work up to 2-3 M in full resolution and having powerful cpu. Just enlarge the mesh. The first trick I found, (reminded me the enlarging like tool when merging in voxels lol)

I used dynamesh with seven64, I7 cpu and 6 giga ram ... but even with a 1 million polygons mesh dynamesh becomes slow. Then if I use the option dynamesh 'PROJECT' calculation time increases a lot....even if the detail is much more accurate.

Example: with about 1 million polygons and project dynamesh activated, resolution 300k, it takes 20 seconds, cpu usage 100%.

c.

Dynamesh is more powerful than 3dcoat voxels when inserting objs from shared libraries.

Just use SketchUp free, the obj free exporter for this and connect to google 3dwarehouse. Most of these fine quality models are more or less problematic. They contain lot of reversed normal faces, lot of non manifolds, lot of holes (non waterproof). It will be a disaster if try to drop them to voxels. Not so for dynamesh ! It will close the holes and won't pay much attention on non manifolds or doubled vertices.

Melt anything on a single mesh and add or cut as much clay you like, fast, using two or three tools only.

Maybe some people don't like the ZB UI interface. I'm not one of these. It's effective and fast.

This took me 15 mins testing all these and having lot of fun. But don't ask me to retopo this... :)

Thanks for the tricks :drinks: I did not know 3dwarehouse

Even for me duplicate meshes on the fly and get booleans without breaking flow is convenience option. And dynamesh is powerful to combine topologies issues.

Pixologic is trying to make the polygons more and more like voxels, move closer to the freedom they offer, and it is happening, though slowly.

The next step for ZBrush (if he want to amaze the world again) will be to preserve almost all the detail since rough sketch, in one click for the user, and fast and easy for the computer.

Perhaps in some way, they will do this mixing sculptris and dynamesh technology?

I hope the next versions does not focus on animation or other departments, because, IMHO, the sculpting methods is not yet complete or perfect (like real clay, like they said...It's real only to low medium resolution) I also hope that 3DCoat concentrates its efforts mainly on sculpture, with voxels and liveclay could get to that result that I dream.

Link to comment
Share on other sites

  • Advanced Member

@ jedwards

I have a quite opposite opinion LOL

I find dynamesh much faster and more spontaneous.

I prefer the 3dcoat merger, booleans are very fast, subtracting holding ctrl and I much prefer the gizmo. I also prefer the cuts using [E] modes.

There's a similar method for cuts in zbrush but again, instructions are for the birds. ctrl+shift for hiding, delete hidden, turn off dynamesh (use the button only) turn it on again. LOL it works fine, like in 3dc. But some more clicks.

As you may know, you can decrease the number of clicks, reducing they to a minimum, costumize the delete key and others button in zbrush. After zb is really fast: ctrl + shift for hiding; (costumize CTRL + X or another combination) for delete hidden polygons.

Also customize the button dynamesh 'group' is also convenient.

It would be very important, in my opinion, also that 3Dcoat will have a masking system on the fly (it is a feature request I have already suggested) This would make some works much faster. I think that not all commands need to be on the fly, but the masking and hiding certainly.

Link to comment
Share on other sites

  • Member

It would be very important, in my opinion, also that 3Dcoat will have a masking system on the fly (it is a feature request I have already suggested) This would make some works much faster. I think that not all commands need to be on the fly, but the masking and hiding certainly.

Absolutely so. An on the fly masking (Freeze) system is a must. Also the Freeze tool and the Pose tool should be using the same masking. This is something that have already been requested in the Feature requests section.

http://3d-coat.com/forum/index.php?showtopic=4869&st=0&p=65452&hl

Fixing these masking/freeze issues in 3D-Coat will definitely speed up workflow.

Link to comment
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

 Share

×
×
  • Create New...