V4.1 BETA (experimental 4.1.17D)

[BeatKitano]

I think you're wrong in this instance. It's true of the current state of Auto-Retopo, but from everything I saw in the Z-Remesher videos was that it does a MUCH better job and gives a cleaner result, regardless of whether it will be animated or not. And what I meant about Hard surface modeling, is that 3D Coat's Voxel modeling tools provide a very good alternative to modeling in a traditional modeling app....but the downside has always been that you end up having to retopologize the model...often manually. That deters a lot of folks, who don't see the benefit. If Auto Retopo is far more accurate and usable, then it eliminates that issue.

 

In fact, it already does a pretty good job on hard surface objects...but with spiral loops and other issues removed, I can see it being an invaluable toolset for this kind of work. Yes, modeling in a traditional 3D App might make more sense in many cases, but I always hated having to piddle with verts, edges and polys...and have to weld overlapping verts, and just the normal clean up involved, throughout the whole modeling process. So, if you model in the Voxel Room, you eliminate a lot of that fuss, and can just create. An accurate, clean Auto-Retopo routine would allow me to bypass virtually all of that piddling and cleanup involved. I think a lot of studios would stand up and take notice...because Art Directors and Studio owners would like to get rid of the piddling and fussing with geometry too.

 

The way I look at it, is this....Pixologic made a name for themselves because they introduced a strange but effective way to do things that just wasn't possible before. I think Andrew can make a name for himself by continuing to make 3D Coat a very serious modeling alternative. No one ever thought of Voxels as a modeling platform, before now. Andrew is pioneering through new territory in this regard. It makes sense to ensure that Auto-Retopo lives up to it's promise...cause beforehand, it had not.

 

Zremesher is awesome, you don't get it when I'm not convinced of it in 3dc.

It's awesome, but in Zbrush, which is a quad based app.

3dc isn't, it freed itself from a 1970's paradigm ! Let it fly, that's something zbrush can't have at the moment !

 

 

"to ensure that Auto-Retopo lives up to it's promise...cause beforehand, it had not."

 

That's where the line is drawn btw: after zbrush it's a zbrush copycat (even if Andrew did it first)... Privilege of the "standard" I suppose :/

[AbnRanger]

@Digman:

 

:)My point to summarize (because this has been dragging for too long and we're all using the software in different "sectors" thus obviously can't agree on everything): if Andrew is reworking autopo for animation mesh: it will be "meh" at best. 

If he's working on it for sculpting: I don't get it why reinstate quad based/subdiv paradigm while we have freeform. If there was a subdiv room I'd would be all for it, just like AbnRanger, it would open a lot of doors with all the conversion processes in 3dc !

If he's working on it just to do like zbrush: I don't get it, seems pointless to me.

If he's working for other purpose (output mesh for statics more accurate) why not. But I don't see an emergency. That's all.

Beat, he's doing it because practically all external apps and renderers are still more friendly with low poly quad meshes, with good topology. We aren't talking about remeshing to have a cleaner sculpt. We're talking about final output, to a clean low poly mesh.

 

Just as I previously mentioned....even when you would use the Text tool in 3ds Max > Extrude/Bevel/Bend, you still would need to throw an Edit Poly modifier on top or convert the whole thing to Edit Poly and have to do an hour or two's worth of work cleaning it up all the skinny polygons or ngons...even if it's not going to be deformed and animated. You can see the crappiness of the topology in the reflections and so on. I was never able to fully get around this. I always had to do some kind of cleanup. So, again. clean geometry is often needed even if there is no animation or deformation.

[BeatKitano]

Beat, he's doing it because practically all external apps and renderers are still more friendly with low poly quad meshes, with good topology. Just as I previously mentioned....even when you would use the Text tool in 3ds Max > Extrude/Bevel/Bend, you still would need to throw an Edit Poly modifier on top or convert the whole thing to Edit Poly and have to do an hour or two's worth of work cleaning it up all the skinny polygons or ngons...even if it's not going to be deformed and animated. You can see the crappiness of the topology in the reflections and so on. I was never able to fully get around this. I always had to do some kind of cleanup. So, again. clean geometry is often needed even if there is no animation or deformation.

 

No, not friendly with quad based meshes. Some poor renderer with normal rendering owning the show.

That's all, with a little work on normal fixing you can get past that in most of those (some require more work with custom shaders/importers/exporters). But I'm getting assertive so I'll stop here ^^

[digman]

Also, I am very aware of the fact that renderers or games will triangulate your quad mesh for rendering purposes, impossible to have non-planar polygons that way but poor polygon flow can cause issues with shadowing artifacts under GI lighting... 

The closer we can get to good polygon flow the better and less workarounds for getting our maps and rendering as we still do live in the lower polygon range for our models in most renderers using normal maps or displacement maps to create the detail. Of course for a displacement map the model gets subdivided at render time based upon the subdivision level you chose. Now a lot of renderers will let you see the subdivisions in real time as well with GPU rendering.

My last comments on this conversation...

[manticor]

I'm hoping that Andrew manages to improve the autopo routine so it handles sharp edges too for hard surface models ,as at the moment neither zbrush or 3dc handles that very well ,they work better for organic surfaces

[gbball]

What I think would be most helpful is the ability to manually retopologize certain areas (eyes, hands, joints, mouth, etc) then allow the autotopo to retopologize the rest.

 

This would be the best solution in my opinion.  I don't know how difficult it would be to do.  But hopefully you read this Andrew and implement this if it's not too much work.

 

-G

[manticor]

I think what would be even better would be some "live" autopo .Whereby it first builds a mesh based on basic guides and then as you place extra guides down it recalculates the flow in that local area . you could also paint in extra colour information for more or less density ,same as zbrush does it but updates on mouse up.

[AbnRanger]

I think what would be even better would be some "live" autopo .Whereby it first builds a mesh based on basic guides and then as you place extra guides down it recalculates the flow in that local area . you could also paint in extra colour information for more or less density ,same as zbrush does it but updates on mouse up.

You do realize they copied that painting feature from 3D Coat...right? You mean paint select areas for greater density...same as we have had for a few years now....

[AbnRanger]

What I think would be most helpful is the ability to manually retopologize certain areas (eyes, hands, joints, mouth, etc) then allow the autotopo to retopologize the rest.

 

This would be the best solution in my opinion.  I don't know how difficult it would be to do.  But hopefully you read this Andrew and implement this if it's not too much work.

 

-G

Why waste development time on that, when you can do the same thing now, in reverse order? If you want to manually retopo the hands and face, for example, you can simply select those areas > delete, and have finish do the manual part from that point.

[BeatKitano]

What I think would be most helpful is the ability to manually retopologize certain areas (eyes, hands, joints, mouth, etc) then allow the autotopo to retopologize the rest.

 

This would be the best solution in my opinion.  I don't know how difficult it would be to do.  But hopefully you read this Andrew and implement this if it's not too much work.

 

-G

That would be the best cause what can potentialy take long in a retopo is if you lay down every loop instead of singular points.

Singular points are what make a good topology, and that's what an automated process will have the hardest time to create properly. I second this idea as the loops all revolve around those special areas of interest.

Filling the spaces in between is much easier (and that's what takes time if you don't go with the simplest poly cage you can create before subdividing) for an automated process.

Basicaly you would just lay down polystrips to create the "good loops" and let the software handle the spaces inbetween.

 

Doing so would:

 

-Give you perfect control over critical parts

-Speedup autoretopo process as the software only have to "loop around" those areas

-Speedup retopo because those transition areas are what takes the most time and is the most boring.

-Give the best result possible between man made and machine made retopo.

[AbnRanger]

That would be the best cause what can potentialy take long in a retopo is if you lay down every loop instead of singular points.

Singular points are what make a good topology, and that's what an automated process will have the hardest time to create properly. I second this idea as the loops all revolve around those special areas of interest.

Filling the spaces in between is much easier (and that's what takes time if you don't go with the simplest poly cage you can create before subdividing) for an automated process.

Basicaly you would just lay down polystrips to create the "good loops" and let the software handle the spaces inbetween.

 

Doing so would:

 

-Give you perfect control over critical parts

-Speedup autoretopo process as the software only have to "loop around" those areas

-Speedup retopo because those transition areas are what takes the most time and is the most boring.

-Give the best result possible between man made and machine made retopo.

I have no idea what it is you are saying.

[gbball]

Why waste development time on that, when you can do the same thing now, in reverse order? If you want to manually retopo the hands and, for example, you can simply select those areas > delete, and have at it

Mainly because I am envisioning a particular workflow...And this would contribute some functionality towards it.

 

One where you can sculpt your voxel or surface mesh, then

manually retopo a certain area where you want precise structures.

Use the retopo geometry to 'model' the structure you want.

Have the changes to the retopo geometry propagate to the voxel/surface mesh.

Continue to work this way, going back and forth between vox sculpt and retopo 'modeling' until you need a fully retopologized mesh.

 

At which point, you'd already have various sections already retopologized.  Then you'd have the option of doing autopo on the remaining areas that you haven't manually retopologized, or just retopologizing the rest manually.

 

My thinking is that there are some things that poly modeling is really good for, where sculpting falls short.  I would love to have the ability to manipulate my voxel sculpt using a low poly mesh (but only in certain places where I need it).  And if you're going to do that, why not retain all the retopo work you've already done?

 

Hopefully that makes sense.

 

-G

[AbnRanger]

Mainly because I am envisioning a particular workflow...And this would contribute some functionality towards it.

 

One where you can sculpt your voxel or surface mesh, then

manually retopo a certain area where you want precise structures.

Use the retopo geometry to 'model' the structure you want.

Have the changes to the retopo geometry propagate to the voxel/surface mesh.

Continue to work this way, going back and forth between vox sculpt and retopo 'modeling' until you need a fully retopologized mesh.

 

At which point, you'd already have various sections already retopologized.  Then you'd have the option of doing autopo on the remaining areas that you haven't manually retopologized, or just retopologizing the rest manually.

 

My thinking is that there are some things that poly modeling is really good for, where sculpting falls short.  I would love to have the ability to manipulate my voxel sculpt using a low poly mesh (but only in certain places where I need it).  And if you're going to do that, why not retain all the retopo work you've already done?

 

Hopefully that makes sense.

 

-G

You can already do that. You can model something like you are talking about, in the Retopo room > go to the Voxel Room > click on the Merge tool > click the "Pick from Retopo" button (tool option panel) > Apply

 

You can do the same thing when you want to use the Cloth tool in the Voxel Room

[gbball]

You can already do that. You can model something like you are talking about, in the Retopo room > go to the Voxel Room > click on the Merge tool > click the "Pick from Retopo" button (tool option panel) > Apply

 

You can do the same thing when you want to use the Cloth tool in the Voxel Room

I know that I can do it that way, but it's not really ideal.  because if I merge without voxelizing, then I have a mesh that doesn't sit together with the other vox layers.  Also, if I want to tweak it, I'd have to delete the layer, make my edit and merge again.

 

We have the option of moving the retopo geo based on changes to the sculpt via checkbox, I wonder how difficult it would be to go the other way around.  Either way, this is changing the topic.

 

I'm still a fan of manual topo, autopo combo, because as Beatkitano said, you would be able to define the areas you really care about and let 3D-coat connect the dots after that (especially if it's something like the back of the head or some other area that will not been seen much).  We give ourselves more control and leave less important decisions to the computer if we define what areas we want topolologized a certain way first.

Edited July 26, 2013 by gbball

[manticor]

You do realize they copied that painting feature from 3D Coat...right? You mean paint select areas for greater density...same as we have had for a few years now....

Yeah I did know that.  Zbrush just expanded upon the idea,seems to me that most of the cool stuff in zbrush were originally Andrews idea's

[Tony Nemo]

Manually retopology where it is indispensible and then calling the autopo engine to fill in seems a great, time saving idea. Has anyone carried out the strokes tool for edge flow to a greater extent than may at first appear necessary? I mean, closer to a manual retopo without creating polygons.

[AbnRanger]

I know that I can do it that way, but it's not really ideal.  because if I merge without voxelizing, then I have a mesh that doesn't sit together with the other vox layers.  Also, if I want to tweak it, I'd have to delete the layer, make my edit and merge again.

 

We have the option of moving the retopo geo based on changes to the sculpt via checkbox, I wonder how difficult it would be to go the other way around.  Either way, this is changing the topic.

 

I'm still a fan of manual topo, autopo combo, because as Beatkitano said, you would be able to define the areas you really care about and let 3D-coat connect the dots after that (especially if it's something like the back of the head or some other area that will not been seen much).  We give ourselves more control and leave less important decisions to the computer if we define what areas we want topolologized a certain way first.

I understand. Not trying to be a wet blanket, just would rather Andrew spend time add features/functionality that we really don't have. To me, what you and he have described is very close to the "Bridge Selected Edges" request, already on the Mantis:

http://3d-coat.com/mantis/view.php?id=1056

 

I think the SHIFT key option with the Transform tool, to extrude on the fly, will speed up Manual Retopo considerably...allowing you to rotate the angle, scale in/out and move the extruded edges simultaneously.

http://3d-coat.com/mantis/view.php?id=1057

[Gravin]

I think I'm going to agree with abnranger, I too would rather see some muscle added to the manual retopo workflow. As for the Shift key extrude on the fly request, is this something that we can accomplish with a macro now that we have scripting in 3dcoat? I haven't looked at the scripting at all in 3dcoat to be honest but this sounds a lot like the extender+ tool in lightwave and if we can script the switching back and forth between transform and extrude and back again it would be pretty easy to setup, just maybe not with the shift key as the shortcut...

[TimmyZDesign]

Yeah I did know that.  Zbrush just expanded upon the idea,seems to me that most of the cool stuff in zbrush were originally Andrews idea's

I've often wondered if there are any copyright laws that would apply to any of this? Do any of these software companies create patents for their ideas?

I know that Apple recently won a settlement against Samsung for patent infringement. I wonder if Pilgway could do the same with Pixologic?

 

My guess is that it would be a long, stressful, drawn-out, legal thing that probably wouldn't even be worth it in the end...

 

Andrew would probably be happiest just creating awesome new tools instead of worrying about that.

Edited July 26, 2013 by TimmyZDesign

[alvordr]

I already do most of this.  Edge flow matters regardless of animation or not, especially when it comes to normals.  I often will now just start manually retopo work, and if I feel it's quicker and just as easy, if not faster...I'll use the strokes tool to do the autotopo work I prefer not waste time on.  There are times when this doesn't work well, so I have to manually clean that up, but I've gotten better about knowing when this might occur and just doing the retopo work myself in those cases.  It seems that when you autotopo from the voxel room, you get the benefit of painting areas where the mesh would be more dense and dictating the target amount of polys, but that's the big difference.

[digman]

Manually retopology where it is indispensible and then calling the autopo engine to fill in seems a great, time saving idea. Has anyone carried out the strokes tool for edge flow to a greater extent than may at first appear necessary? I mean, closer to a manual retopo without creating polygons.

I'm not sure if this is what you are taking about Tony but I will throw out the information for anyone who does not know the strokes tool will do this when manual retopoing...

 

The strokes tool will do this for you for filling in between polygon strips... Simple example but works on more complex polygon strips for good edge flow. You do have to connect the stroke line with the vertex of the polygons on both strips. The polygon vertex will turn yellow when you are hovering over it. 

 

I held down the shift key to get a straight line for the strokes tool in this example. The above method does save a good amount of time as you are using the stroke tool to guide your polygons creation between polygon strips when manual retopoing...

 

I used the brush tool to even out the polygons a little after the polygons were created.

 

I set my stroke control points density to about 25%

[AbnRanger]

I'm not sure if this is what you are taking about Tony but I will throw out the information for anyone who does not know the strokes tool will do this when manual retopoing...

 

The strokes tool will do this for you for filling in between polygon strips... Simple example but works on more complex polygon strips for good edge flow. You do have to connect the stroke line with the vertex of the polygons on both strips. The polygon vertex will turn yellow when you are hovering over it. 

 

I held down the shift key to get a straight line for the strokes tool in this example. The above method does save a good amount of time as you are using the stroke tool to guide your polygons creation between polygon strips when manual retopoing...

 

I used the brush tool to even out the polygons a little after the polygons were created.

 

I set my stroke control points density to about 25%

In 3ds Max, you can select an edge or edgeloops on the two parts your want to join together (Bridge) and click the Bridge button. You have a caddy or small options dialog that allows you to set how many (parallel) segments you want....and you can slide them together one way or the other, or pinch/spread them apart. This would effectively give you what Beat is asking for

 

http://www.youtube.com/watch?v=LrYIrJsZmS0

[digman]

Yep, that would be better and more effective AbnRanger as I was just letting some users know what we have now... The 3ds Max solution sounds very good and something like that in 3DCoat would be great...

 

Man, I spent a lot of time of the forums today, off to digital clay land now in 3DCoat...

[AbnRanger]

Yep, that would be better and more effective AbnRanger as I was just letting some users know what we have now... The 3ds Max solution sounds very good and something like that in 3DCoat would be great...

 

Man, I spent a lot of time of the forums today, off to digital clay land now in 3DCoat...

Oh, yeah...I agree on that technique. I was just thinking of something that could be done in a few clicks

[alvordr]

I was recently thinking it would be great to have a bridge tool for times like when I just had to do the inner topology of a tube (gun barrel).  This would have saved a good bit of time.

[AbnRanger]

I was recently thinking it would be great to have a bridge tool for times like when I just had to do the inner topology of a tube (gun barrel).  This would have saved a good bit of time.

Here is another example...at least the first half of the video. Pretty nice, interactive way to bridge. And we really, really need a Shell/Thickenss modifier, to make it easier to apply topology both sides of a thin object

http://www.youtube.com/watch?v=7o2CO9wY6os

[manticor]

Here is another example...at least the first half of the video. Pretty nice, interactive way to bridge. And we really, really need a Shell/Thickenss modifier, to make it easier to apply topology both sides of a thin object

http://www.youtube.com/watch?v=7o2CO9wY6os

Is this not already possible ? Import model into topo room as surface with make shell option on ? Not tried it myself but worth a shot .

Thing is people seem to forget that everyone has blender available to them which have these tools at their disposal and with the app link it's a single click operation .i know it's nice to stay in one application but Andrew is just one guy ,I think it would make his life easier so he can focus more of his time on features and improvements that are not available in most other packages.

And you have max which has the most robust modeling tool box of any package ....

Don't mean to cause offence btw

[Tser]

Its really good that Andrew is working on Auto re topology (autopo) as spiralling geometry made this unworkable for me and the lack of edge flow meant many hours of trial and error with placing curves or just simply doing the whole retopoing manually.

zremesher in ZBrush is very good, I can create a sculpt with dynamesh (as triangles), then, when finished with all the finer details, I duplicate it, run zremesher on it, then increase its subdivisions high enough so that it can use the finer details and re-project the original sculpts details over the new retopoed mesh. Then I generate the UV map with UV Master and export the required disp etc maps and low poly mesh with MultiMapExporter.  This work flow is VERY good and easy and works every time and the resulting lo poly mesh with its maps animate very well in Softimage/Maya and Modo. Also, this work flow is stress free so I can be an artist and not keep putting on my technicians hat every step of the way.

[BeatKitano]

Raaaah calm down... it's not him AbnRanger. I do agree I don't understand manticor's reasoning but that doesn't mean we should get angry (Funny coming from me).

[Tony Nemo]

Thanks for the grins, guys!