Jump to content
3DCoat Forums

PolyStrips in Blender


probiner
 Share

Recommended Posts

  • Advanced Member

Like C4D this retopo is antiquated and slow. Blender has an excuse; it's free. C4D is way out of the running. People here unjustly criticize Andrew's autopo for not being perfect for not being as good as Zbrush's remesher but remember Andrew was first and he did it pretty much all by himself. 

 

C4D has a lot more money than 3d Coat and they're still 6 years behind Andrew.

Link to comment
Share on other sites

  • Reputable Contributor

That looks pretty cool. They also have that cool plugin that's similar to the Strokes tool, but has some enhanced features. Andrew mentioned he has plans to visit the Retopo Room soon and make it more of full-featured modeling app...so you can have the option to model in Voxels or polys, all inside the app. I requested 4-5 features as well, that could help a lot. One of those is allow the artist to use all the marquee selection tools in the E-Panel to create strokes with greater precision. For example, there are many times you want a perfect circle or elliptical stroke. If one can use the CIRCLE draw mode to generate a perfectly circular stroke, it can speed up the process greatly.

  • Like 1
Link to comment
Share on other sites

  • Advanced Member

i love 3d coats tools, but i would love the ability to paint poly strips. Its just easier to vizualize what you are doing and often takes less effort.

 

this blender tools looks a bit more hardcore than c4d's... though...

Link to comment
Share on other sites

  • Advanced Member

Hi !

 

Andrew is working in a better retopo solution.

 

And for this task, i prefer Cinema4D newly developed PolyPen.

 

just an opinion.

 

Ty for sharing !

 

That's great as long as nothing that is currently there breaks - we must keep 3DC's current manual retopo in place in addition to anything new.  I really like it and nothing I've played with feels as nice.  I'd rather not have it replaced - but am always happy for other options...

Link to comment
Share on other sites

  • Advanced Member

I mean I like the new autotopo tools in 3DC and all, but who here uses them for hero characters or key animateable assets?  For background or static objects - yes.  For my main characters or riggable assets - manual is the only way.  Am I missing something?  I don't think autotopo can work (I mean be ideal) for something if it has no idea where the bones are going or how you intend to deform the object.

Link to comment
Share on other sites

  • Contributor

The best solution for retopology would be a mix between manual retopology and automatic retopology. It would be nice to just place 3 and/or 5 pointed poles on the mesh (to dictate the desired edge flow) and then let the autopo algorithm intelligently fill in the rest of the polygons for you.

An algorithm which does something very much like this has already been written and presented at Siggraph 2013, but it hasn't been added to any commercial 3D software yet (as far as I know).

The most useful part of this algorithm is how it automatically updates the edgeflow when 3 or 5 pointed poles are manually moved across the mesh. In the video they simply refer to the poles as "irregular vertices". It almost seems as though the authors don't realize how useful and essential such vertices are for creating an animateable mesh! In the video the users are simply sliding these vertices across the mesh and the edgeflow around the vertices is updating in real-time. That functionality alone would be fantastic in 3D-Coat!

As far as I know, Andrew and Raul are already working on something like this. I get this impression because when Zbrush 4r6 was released, Andrew updated the autopo algorithm in 3D-Coat. At that time he mentioned that he had some ideas for a "sketch-based" version of autopo, and was thinking about passing that project on to one of his other staff members. To add to that, recently Raul mentioned that he was working on a new quadrangulation algorithm, so maybe this "sketch-based" autopo is finally on the way! Let's keep our fingers crossed!

In the meantime you can wet your palette with the Siggraph presentation video:

  • Like 1
Link to comment
Share on other sites

  • Contributor

The best solution for retopology would be a mix between manual retopology and automatic retopology. It would be nice to just place 3 and/or 5 pointed poles on the mesh (to dictate the desired edge flow) and then let the autopo algorithm intelligently fill in the rest of the polygons for you.

An algorithm which does something very much like this has already been written and presented at Siggraph 2013, but it hasn't been added to any commercial 3D software yet (as far as I know).

The most useful part of this algorithm is how it automatically updates the edgeflow when 3 or 5 pointed poles are manually moved across the mesh. In the video they simply refer to the poles as "irregular vertices". It almost seems as though the authors don't realize how useful and essential such vertices are for creating an animateable mesh! In the video the users are simply sliding these vertices across the mesh and the edgeflow around the vertices is updating in real-time. That functionality alone would be fantastic in 3D-Coat!

As far as I know, Andrew and Raul are already working on something like this. I get this impression because when Zbrush 4r6 was released, Andrew updated the autopo algorithm in 3D-Coat. At that time he mentioned that he had some ideas for a "sketch-based" version of autopo, and was thinking about passing that project on to one of his other staff members. To add to that, recently Raul mentioned that he was working on a new quadrangulation algorithm, so maybe this "sketch-based" autopo is finally on the way! Let's keep our fingers crossed!

In my opinion Sketch-Based Generation and Editing of Quad Meshes is probably the best retopo solution that currently exists, and It's surprising it wasn't adopted by any of the 3D programs yet.

If Andrew's team is indeed working on something similar, then I think (with PBR just around the corner) we have at least two very interesting 3D Coat builds to wait for. :good2:

Link to comment
Share on other sites

  • Member

I'm just praying that Andrew bites the bullet and creates a "Poly" Tab.

 

3DC is so close to being a great polygon modeler when you look at the topo tools - adding a bit more and putting that into a Polys tab would really make 3DC a one-stop-shop for me.

 

I'd love to see Andrew's spin on poly modeling; if 3DC could ever become like, or surpass say modeling in Silo, that would be perfect!

 

 

-Will

 

Link to comment
Share on other sites

  • Advanced Member

i have to +1 that. if it had some regular poly modelling tools. would, especially with the freedom that andrew has to create something original and not tied to old concepts it could be an amazing toolset.

Link to comment
Share on other sites

  • 8 months later...
  • Contributor

It looks like PolyStrips and Contours were merged into a bundle called RetopoFlow. This is a perfect example of a semi-automated retopo workflow. You first manually draw main poly loops and then automatically create fill geometry between/inside them.

 

Something I would really like to see in 3D-Coat as the only thing it automates now is the Strokes tool - and only to a certain degree, because when working on non-cylindrical objects, you still need to manually draw all of the horizontal and vertical spans (and then fix N-poles and E-poles by merging vertices).

Edited by ajz3d
Link to comment
Share on other sites

could be hard to develop manually draw main poly loops and then automatically create fill geometry between/inside them ?

-----------------------------------------------------------

  AbnRanger, on 09 Aug 2014 - 9:13 PM, said:

    https://www.youtube.com/watch?v=5RR06wjt2rs [^]

It and a few tweaks to the Strokes tool (like an option to "Avoid Selection through object" and the ability to use the Rectangle or Elliptical Marquee Selection draw modes to lay down perfectly round, square or rectangular strokes) would make 3D Coat stand out again as a Retopology toolset

-----------------------------------------------------------

Please feel free to add a +1

Contours functionality to the Strokes Tool

Link to comment
Share on other sites

  • Contributor

I had a chance to try the plug-in out.

One particular thing that I find to be very powerful in it, is that you can always edit any poly strip and quickly increase the number of spans. This change propagates on all connected geometry (along edge loops), be it other poly strips or poly patches between them. Of course sometimes it can invalidate the latter and you need to recreate the patches. This is one of the disadvantages of the plug-in (something to be improved, I guess). And speaking of them:

- It's not possible to create two junctions sharing the same edge (or at least I couldn't figure out how).

- Once, the snapping algorithm got confused and instead of following the stroke, it kept creating a poly strip starting from an unwanted junction (it might have been a user-error).

- If you have two poly strips and a poly patch between them, you can't insert more junctions to either one of the strips without having to delete the patch first.

 

Nevertheless, if only Blender could handle high polycounts without a severe impact on viewport performance, like 3D-Coat does, and had a standardized interface, I'd completely switch to RetopoFlow for all of my retopo jobs. I'm very impressed with this plug-in and I believe it's one of the best semi-automated retopo implementations I've seen. I feel that the time taken for creating retopo mesh for my test object was like hundred of times shorter than if I were to retopo it in 3D-Coat.

To sum up, and please do forgive my unintentional bluntness, I think that instead of focusing their resources and energy on Autopo, Pilgway should shift their attention to semi-automatic solutions. Meaning - let the user to paint main edge or poly loops, then let the algorithms to create the remaining geometry (fill the gaps), edit if necessary and relax fill patches. This would be ideal IMHO. I've been using 3D-Coat on daily basis for the last three years and I never had to use Autopo, no matter how close the deadline was. To let the algorithm fully retopologize the model on its own is just too much of a risk for me and it takes too much of the precious time to calculate. And the outcome is unpredictable. It might be good, but it may as well be bad and you can do nothing when you're awaiting for the algorithm to present you the result.

But this is of course only my own, personal opinion on the subject. :)

Andrew had done so much good job in the Autopo department that I'm sure the algorithms he had developed during the process could be used in the semi-automatic retopo workflow with a great success. And Raul is working on the close holes tool that could also help here a lot. Combine the two with the parts of the Strokes algorithm and the other available retopo tools, and I say we would have The King of The Retopo World.

Edited by ajz3d
  • Like 2
Link to comment
Share on other sites

  • Reputable Contributor

 

I had a chance to try the plug-in out.

One particular thing that I find to be very powerful in it, is that you can always edit any poly strip and quickly increase the number of spans. This change propagates on all connected geometry (along edge loops), be it other poly strips or poly patches between them. Of course sometimes it can invalidate the latter and you need to recreate the patches. This is one of the disadvantages of the plug-in (something to be improved, I guess). And speaking of them:

- It's not possible to create two junctions sharing the same edge (or at least I couldn't figure out how).

- Once, the snapping algorithm got confused and instead of following the stroke, it kept creating a poly strip starting from an unwanted junction (it might have been a user-error).

- If you have two poly strips and a poly patch between them, you can't insert more junctions to either one of the strips without having to delete the patch first.

 

Nevertheless, if only Blender could handle high polycounts without a severe impact on viewport performance, like 3D-Coat does, and had a standardized interface, I'd completely switch to RetopoFlow for all of my retopo jobs. I'm very impressed with this plug-in and I believe it's one of the best semi-automated retopo implementations I've seen. I feel that the time taken for creating retopo mesh for my test object was like hundred of times shorter than if I were to retopo it in 3D-Coat.

To sum up, and please do forgive my unintentional bluntness, I think that instead of focusing their resources and energy on Autopo, Pilgway should shift their attention to semi-automatic solutions. Meaning - let the user to paint main edge or poly loops, then let the algorithms to create the remaining geometry (fill the gaps), edit if necessary and relax fill patches. This would be ideal IMHO. I've been using 3D-Coat daily for the last three years and I never had to use Autopo, no matter how close the deadline was. To let the algorithm fully retopologize the model on its own is just too much of a risk for me and it takes too much of the precious time to calculate. And the outcome is unpredictable. It might be good, but it may as well be bad and you can do nothing when you're awaiting for the algorithm to present you the result.

But this is of course only my own, personal opinion on the subject. :)

Andrew had done so much good job in the Autopo department that I'm sure the algorithms he had developed during the process could be used in the semi-automatic retopo workflow with a great success. And Raul is working on the close holes tool that could also help here a lot. Combine the two with the parts of the Strokes algorithm and the other available retopo tools, and I say we would have The King of The Retopo World.

I concur. That does look like it would be a mind-blowing addition to 3D Coat's Retopo tools, if Andrew implemented those, in addition to the standard poly-modeling tools mentioned elsewhere...that Raul was supposedly working on. The Strokes tool has always been a semi-auto retopo tool, but it needs an update like this...and both major features seem to be STROKE related. So, just trick the heck out of the current Strokes tool

Link to comment
Share on other sites

  • Contributor

That does look like it would be a mind-blowing addition to 3D Coat's Retopo tools, if Andrew implemented those, in addition to the standard poly-modeling tools mentioned elsewhere...

Exactly. I think this would make retopologization once again - a pleasurable activity.

Let's take a closer look what the retopo is. It's recreating the topology from scratch while simplifying the geometry and making it (if it's required) friendly to deformations. One defines main edge loops that are crucial to the topology flow, and then fills the gaps between them with geometry according to some mathematical rules. And it's those gaps that are the most time-consuming and tedious things to create, especially with the currently available toolset. There is also a matter of immediate geometry feedback that the RetopoFlow plug-in provides and the ability to tweak spans of all strips in any given moment.

Currently with the Strokes tool it's difficult to visualize the poly flow (if you lay down too many strokes, especially if they're on both sides of the model), and it's impossible to predict if Strokes will create all of the polygons you intended it to (sometimes gaps can appear which you then need to fill). Not to even mention that it can't handle vertices with a valency of 5 or more.

Then, after you press the Enter key to create the geo, that's it. To increase or decrease the number of horizontal or vertical spans, you need to either start deleting edge loops one after another or add more splits.

And if you want to fill the gaps between the main loops, you need to paint strokes manually for each span - horizontal and vertical.

Time spent for these tedious tasks adds up and results in boredom and frustration. :( (For some time Benny Hill was helping me with those tasks). ;)

Retopo world has moved on, and 3D-Coat's retopo tools have not been touched for a long, long time. Maybe it's time to keep up with the others?

I know it's not an easy task, but I think that the retopo room really needs some refreshment.

Edited by ajz3d
  • Like 1
Link to comment
Share on other sites

you meant something along the lines?

http://farsthary.com/2015/07/19/alpha-quadpaint-tool/

It is nice to paint shapes and fill them for shure.

But for us retopo artists it is much more interesting to paint with strokes like poly paint does.

The point is: An artist want to decide exactly where the main polyflow have to be placed.

For me the very first decision is to create a good polyflow at areas that have to be animated later. A good quad filling of the rest of the areas is just "secondary".

Don´t misunderstand me. It is really cool to safe time and fill an area just by a click, but the very first step is always to decide where the important polyflows have to be placed.

Raul I think you know what we mean. :)

My dream is to paint the polystripes and fill the open areas with quads later. Totally cool imagination.

Link to comment
Share on other sites

  • 6 months later...
  • Contributor

I'm one of the devs of retopoflow and wanted to chime in on sketch based retopo.

https://youtu.be/uDU34OcUjtI

Great! I like how it shows the number of points on each edge and that the user can increase or decrease that number as desired.
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...