Modo 501 poly sculpting

[geo_n]

Well I think we really need a poly modelling and sculpting room in 3dc to stay ahead.

[gbball]

I agree, or at least a way to mimic working with poly's in the voxel sculpt room.

[michalis]

I know what you mean. I mostly agree. A surface sculpting tool.

We have already a sculpting room waiting for more development. But the whole workflow confuses me. Its obvious that we have to leave vox room and no way back. Of course we can re-drop the poly mesh to voxels but its another story. This 'room' is called zbrush in my workflow. But I paid a lot more for this. LOL

As for modo, blender has this 'sculpting room' years now, in v2.5x this is optimized to handle millions of faces. What makes a good sculpting app is the tools though. And performance. Modo (from these videos) doesn't seem to have this, I'm not so sure. If it does, 3dc + modo is a great combination.

[Psmith]

Many people don't realize it, but 3D-Coat also has a surface sculpting mode, (Surface tools), and, in this mode, the voxel model is truly being sculpted on its surface only - complete with polygon stretching as found in "Brand X".

This functionality is the equivalent of "going back and forth" between voxels and polygons. To be convinced of this fact, just take any voxel sculpture into "Surface" mode and start sculpting with a Brush set at an extreme depth - you'll soon begin to see the individual polygon faced "skin" start to stretch - and you will hit the limits of such stretching.

Then there is this other Room that also has polygonal sculpting tools - for those rare occasions when you realize you forgot something while working in voxel space. (I'd like to see the functions and tools found in the "Sculpt Room" transferred into the "Surface" area, myself).

Now that the brush engine is nearing its final iterations, in its quest toward completion, it should be quite easy for Andrew to add more surface sculpting tools, (and maybe even polygon functions beyond this)- those that mimic all the tools found in "Brands X & Y, and Z".

Greg Smith

[michalis]

Many people don't realize it, but 3D-Coat also has a surface sculpting mode, (Surface tools), and, in this mode, the voxel model is truly being sculpted on its surface only - complete with polygon stretching as found in "Brand X".

Psmith, there's a misunderstanding here. When we say 'surface modeler' we mean, After retopology. We mean with subdivision levels support. I mean zbrush or similar. Modo looks like it has a new sculpting engine, using a famous subsurf algorithm (pixar's subsurf). I haven't try it, so not much to say. I will though as I'm searching for a nice app to replace blender.

[Psmith]

Michalis:

No misunderstanding, (on my part), at all. By combining the "post topology" functions found in the Sculpt Room with the "Surface" mode, found in the Voxel Room, everyone is made happy - for, not only can one switch back and forth between voxel sculpting and polygon sculpting, (in a temporary way), one can use the same tools, (because all polygonal tools exist in one Room) - all over again - after polygonal topology has been permanently created, via the Retopo Room.

Greg Smith

[AbnRanger]

Michalis:

No misunderstanding, (on my part), at all. By combining the "post topology" functions found in the Sculpt Room with the "Surface" mode, found in the Voxel Room, everyone is made happy - for, not only can one switch back and forth between voxel sculpting and polygon sculpting, (in a temporary way), one can use the same tools, (because all polygonal tools exist in one Room) - all over again - after polygonal topology has been permanently created, via the Retopo Room.

Greg Smith

Yes, that is what I have been asking about for some time now...and I actually think it may prove to be the fastest sculpting model in 3DC. It would also be great for people who want to bring in models and just detail it without having to go through the retopology process. You can do that to a degree with Voxels (by importing the original mesh into the Retopo room and use that as your Retopo mesh), but staying in Polygonal mode when you import a model, has it's own advantages. I think the best thing about it though, is that it would make 3DC the most versatile sculpting toolset (Voxel, Polygonal, and Image-Based) in the business, at that point.

[haikalle]

There is quite nice new features but what I have heard the main goal of this update was to clean the code and makes everything work faster. And reading feedback from

cgtalk forum they have reached that goal for sure. Good job

[geo_n]

Theyre doing an excellent job of leading the 3d content creation market. Sooner or later their sculpting tools will improve and will negate the need for zbrush, 3dcoat,etc for some when you can sculpt details in one package and also does traditional polymodelling so well.

http://www.luxology.com/modo/tour/multiressculpting.aspx

[haikalle]

Theyre doing an excellent job of leading the 3d content creation market. Sooner or later their sculpting tools will improve and will negate the need for zbrush, 3dcoat,etc for some when you can sculpt details in one package and also does traditional polymodelling so well.

I doubt this will happen. Modo is improving but so is Zbrush, Mudbox and 3D-Coat. These programs can focus only for one specific area in 3d. But Modo have to focus more overall, to

maintain all aspects of 3d. And I think that Modo coders are quite busy when they start to add some real animation tools.

[michalis]

Yes, that is what I have been asking about for some time now...and I actually think it may prove to be the fastest sculpting model in 3DC. It would also be great for people who want to bring in models and just detail it without having to go through the retopology process. You can do that to a degree with Voxels (by importing the original mesh into the Retopo room and use that as your Retopo mesh), but staying in Polygonal mode when you import a model, has it's own advantages. I think the best thing about it though, is that it would make 3DC the most versatile sculpting toolset (Voxel, Polygonal, and Image-Based) in the business, at that point.

I really dont understand. I may missed something but I can't merge an obj when in surface mode. I have to drop it in voxels mode. Then I'm loosing all topology. I have to import this as retopo mesh in retopo room. But why to do this? Except of editing topology but better to just import reference mesh instead. I mean that good sculpting needs topology first! And a multi-res sculpting room, a zb room. This is what I'm trying to say here. IMO of course.

To import an obj as retopo mesh is something I'm using in some cases only. BTW

Sooner or later their sculpting tools will improve and will negate the need for zbrush, 3dcoat,etc for some when you can sculpt details in one package and also does traditional polymodelling so well

This will never happen. But its always handy to have a nice sculpt room on these apps. I always do some edits when in blender, after first render tests.

[AbnRanger]

I really dont understand. I may missed something but I can't merge an obj when in surface mode. I have to drop it in voxels mode. Then I'm loosing all topology. I have to import this as retopo mesh in retopo room. But why to do this? Except of editing topology but better to just import reference mesh instead. I mean that good sculpting needs topology first! And a multi-res sculpting room, a zb room. This is what I'm trying to say here. IMO of course.

To import an obj as retopo mesh is something I'm using in some cases only.

I think there is a misunderstanding here....not sculpt room tools in Surface mode (within the Voxel Room), but the reverse. Mirroring the Surface Mode tools/engine into the Sculpt Room, so you are working with geometry.

And as for Modo 501 replacing the need for programs like ZBrush, 3DC and such....nah. It's just like the Viewport Canvas tools in Max. You can do a LOT of your 3D texture painting in Max, with some very Photoshop-esque tools/layers/blending modes, etc....but I really don't even bother using it, when I have 3DC to use instead. These specialized toolsets will never be outdone by built-in tools.

For example, even if Newtek licensed 3D Coat to be integrated with Lightwave, it would soon be left behind in development as these types of arrangements always tend to one time deals. Imagine using LW with an integrated version of 3DC v 3.1....now you can see why Modo's sculpting tools will neither catch up, nor keep pace with these applications.

[Psmith]

Truly, up until the present, good topology has been the starting point for any and all sculpting operations. Which, if you think about it, is exactly the opposite of what should be.

To thoroughly mimic "real" media, one should be able to start with a block of material and add to it or take away from it, at will - without needing to think about what kind of "skin" might need to be placed "over" it, upon completion of the sculpt.

Also, too much emphasis has been placed by 3D traditionalists on "perfect" topology for all applications. Unless an object will be subject to extreme close-up, static rendering or closely rendered animation, "perfect" topology is not necessary - and often, producing it is just a waste of time.

Now, with algorithms like AUTOPO for organic shape meshing and algorithms such as are found in Groboto 3, for hard surface meshing - the sculptor can just sculpt and the modeler can just model. The responsibility for producing good topology always should have fallen to the computer and its algorithms, rather than to the artist and his required technical prowess - and now - this is becoming a reality.

I would suggest that Andrew's AUTOPO produces very good organic topology with very little assistance - and that Groboto 3's new hard surface meshing algorithm produces more than adequate topology - even for close-up rendered subjects.

So, some of us will have to adapt, (or not), to this new "right side up" approach to creating models. You can bet some folks and their technical skills will suffer from the introduction of these kinds of "automatic" tools.

Greg Smith

[michalis]

These specialized toolsets will never be outdone by built-in tools.

+1

As for the sculpt room, the most important for a surface post-topo app is the multi-res functionality.

Have you noticed this subsurf pixar-like tool in modo 501? Does anyone knows something more about it? Subdivisions only where we need them, they claim so. What do they mean?

[geo_n]

+1

As for the sculpt room, the most important for a surface post-topo app is the multi-res functionality.

Have you noticed this subsurf pixar-like tool in modo 501? Does anyone knows something more about it? Subdivisions only where we need them, they claim so. What do they mean?

Sorry I meant that sculpting will improve in modo that there will be a point when some users would not need specialized software like zbrush and 3dcoat. Modo sculpting would suffice. It has happened before when some users gave up zbrush and switched to 3dcoat because they say they don't really need super detailed sculpting and already satisfied with painting details in 3dcoat.

Afaik the new pixar sds makes use of weights to elimate the need for extra edge loops to make hard edge shape. You can model higher detailed hard edged models with very few poly count. Just don't know how they keep track of weight info. Is it a list of weights,etc. I wonder because in animation the more weights the slower the viewport and also the long list of weights is another thing to manage.

Uv mapping will show less distortion using pixar sds. Looks amazing from the pics.

Another good thing about it is transferring that model to maya thru fbx will retain all the weight info from modo. It will be a perfect copy. No more collapsing the subd to export to other appz. Subdiv model is identical in modo and maya.

[AbnRanger]

Truly, up until the present, good topology has been the starting point for any and all sculpting operations. Which, if you think about it, is exactly the opposite of what should be.

To thoroughly mimic "real" media, one should be able to start with a block of material and add to it or take away from it, at will - without needing to think about what kind of "skin" might need to be placed "over" it, upon completion of the sculpt.

Also, too much emphasis has been placed by 3D traditionalists on "perfect" topology for all applications. Unless an object will be subject to extreme close-up, static rendering or closely rendered animation, "perfect" topology is not necessary - and often, producing it is just a waste of time.

Now, with algorithms like AUTOPO for organic shape meshing and algorithms such as are found in Groboto 3, for hard surface meshing - the sculptor can just sculpt and the modeler can just model. The responsibility for producing good topology always should have fallen to the computer and its algorithms, rather than to the artist and his required technical prowess - and now - this is becoming a reality.

I would suggest that Andrew's AUTOPO produces very good organic topology with very little assistance - and that Groboto 3's new hard surface meshing algorithm produces more than adequate topology - even for close-up rendered subjects.

So, some of us will have to adapt, (or not), to this new "right side up" approach to creating models. You can bet some folks and their technical skills will suffer from the introduction of these kinds of "automatic" tools.

Greg Smith

I agree with this, when you are talking about non-animated objects in the scene. Auto Retopo makes a lot more sense when you have a bunch of objects to have retopo, in order to populate a scene or environment, for example. And I think it even makes sense for clothing, as I know 3ds Max's Cloth (garment maker) applies a kind of random triangular mesh anyway. But when we're talking about animating a character, I just don't think it makes sense to even attempt to use an automated process, no matter how close it gets you. It's better to just get an efficient routine down, doing it manually than to make multiple attempts with Auto Retopo and then end up having to do quite a bit of clean up.

For one thing it tends to often spiral edgeloops (instead of making each one a closed loop)...and that makes it a nightmare trying to clean up. In all my attempts, it's been hit or miss. Sometimes it's so good, I have to giggle a bit (feeling like I cheated on a test or something), other times it just doesn't seem worth it.

Going back to Characters....sure, for ones that will not be in the foreground, near the camera, you might could do fine with Autopo....but it's almost like going out to a fine restaurant and ordering a Hungry Man TV Dinner. Sure it's food, but....

[michalis]

For one thing it tends to often spiral edgeloops (instead of making each one a closed loop)...and that makes it a nightmare trying to clean up. In all my attempts, it's been hit or miss. Sometimes it's so good, I have to giggle a bit (feeling like I cheated on a test or something), other times it just doesn't seem worth it.

+1 again.

I use autopo a lot, cleaning for half ah hour... but works in most cases. I'm not an animator but imagine a figure, starting in voxels, then go to another app (zbrush here) to set up the pose and more sculpture. Setting the pose has similar to animation problems. Stretching! I have to have loops! I also need a comfortable topology for surface sculpting. To go for a 20M single mesh everytime, just for not seeing quad artifacts... this isn't helpful for the artist.

@Psmith

I agree with all these, I'm also thinking that this is the vision of Andrew. I can imagine him writing codes for a traditional surface sculpting room, far away from this vision LOL. I think I understand.

Lets not forget how much freedom exists in traditional sculpturing or painting. Not much and the same time a lot. Just start to play with real clay, start with wires-structure etc etc. Keep clay wet (not much) etc etc. Wait for oil painting to dry, or not dry. Watercolor techniques are even more difficult. But all these made me an open minded when talking about techniques. Now we just have to construct a simple wireframe before going for more sculpture. Don't place it out of art because (thinking otherwise) this could be the essence of art, the structure. On my last wip post, I just did as you described, just voxels and a shader. A very nice way for conceptual - illustrations images. But if you have to setup poses... find me another way. See my finished work (aphrodite). How to sculpt this pose in voxels room? How to retopo this pose? A figure with open hands? Just try then to pose it as it is. And sculpt more. I had to sculpt the head only in voxels, retopo fast and finish base body mesh in blender (old box modeling). The fastest way to go for surface modeling in this case. Everything else could give me more trouble than fun. All parts (hands, legs) were groups, I could hide them and work as artist. A multi resolution mesh at the same time. The fastest and more secure way to bake nice displacements.

Most people likes 3DCoat because of its retopo room only. Interesting isn't it? Always I'm trying to explain them what voxels are and what this means.

edit: I just remind you that the real title of this thread is:

Well I think we really need a poly modelling and sculpting room in 3dc to stay ahead.

[Psmith]

Michalis:

True. Not all tools exist, in 3D-Coat, yet. Especially really efficient posing tools. The dilemma, I think, for Andrew is that good posing tools imply "skeletal", armature-based tools. If he is going to work that hard, (it is very hard to develop) - then he might as well go all the way and include some time-based posing - animation tools.

And, once you get into animation - all the "experts" come climbing out of the woodwork, demanding that advanced rigging tools and modifiers get developed too. It becomes a development nightmare.

All you have to do is look at Zbrush development: first Ofer implemented very basic posing tools, then people screamed for better ones - he develops those - and people screamed even more for animation tools, (since the posing tools implied such usage). Now, I think Ofer always wanted animation in his program, but because of the potential development nightmare, he has postponed these tools and functions until "the end".

I imagine Andrew might follow the same course, (by popular demand).

Greg Smith

[michalis]

The first Andrew should do IMO, is to develop a clever way to hide and unhide not only parts of voxels but parts of topology too. I haven't found any way so far.

The difficult parts between thighs and hip... where any shrinkwrap engine has hard time. You see Psmith, I don't really expect easy modern solutions. The old box modeling still remains a good way, the best sometimes.

BTW pixologic did the half of the animation tools. The very clever zspheres rigging way doesn't work for animation. LOL. They never managed to have a decent or almost decent retopo tool yet. The one they have is a buggy useless tool. The only way I found to crash zbrush4. A nice subject here but as I own the two apps, I feel almost happy, a great workflow. If modo can provide a nice surface sculpting tool then pixologic may have a problem. 3DC never as Andrew is looking at the subject from a completely different angle of view. This is the 'secret'.

[AbnRanger]

Michalis:

True. Not all tools exist, in 3D-Coat, yet. Especially really efficient posing tools. The dilemma, I think, for Andrew is that good posing tools imply "skeletal", armature-based tools. If he is going to work that hard, (it is very hard to develop) - then he might as well go all the way and include some time-based posing - animation tools.

And, once you get into animation - all the "experts" come climbing out of the woodwork, demanding that advanced rigging tools and modifiers get developed too. It becomes a development nightmare.

All you have to do is look at Zbrush development: first Ofer implemented very basic posing tools, then people screamed for better ones - he develops those - and people screamed even more for animation tools, (since the posing tools implied such usage). Now, I think Ofer always wanted animation in his program, but because of the potential development nightmare, he has postponed these tools and functions until "the end".

I imagine Andrew might follow the same course, (by popular demand).

Greg Smith

Actually, I think all that can be avoided. If Andrew could merge a variant of the FFD lattices from the Primitives Tool with the Pose Tool. The would give Andrew the Lattice control many of us have asked for. The Pose tool would have a lot more juice, and avoid complicated bone structures.

Instead of bones, you'd have FFD lattice options to turn to in the Pose tool, for more control. The way it would work is to have the FFD primitives tools load when you check FFD, and instead of a green primitive you see a semi-transparent transparent one. Where this primitive intersects with the object/layer(you have selected), that is the volume your lattice will control...and you can use the selection tools in the pose tool to affect the falloff/weighting. I think adding the ability to save the deformer, perhaps as a child layer would make it even better.

Much like Andrew was able to utilize existing tools (strokes and freeze) in 3DC to use in Auto Retopo, he could possibly do the same here.

[michalis]

All these are very very interesting. 3DC will never replace older methods. But this is what we want. A different approach. The box modeling, the surface after retopo method, the voxels method. All are useful. And if modo provides a sculpting modeler (as blender does for some years) this is useful too. To have all in one, could be a choice of an individual artist but not for everyone.

To have users ask for more and more isn't so bad. Development never ends.

So I wish: when I hide voxels, when I hide topology, when snap-to is disabled, then why I can't draw some vertices?