Sneak Peek: 3D Coat's new Auto-Retopo


While Andrew was recently working to develop a new Auto-Retopology routine, based on this recent article:
…there were a number of requests on the forums to “meet in the middle” and allow users to create some guides in the process, that dictate where they wanted the edgeflow and let the program take it from there. It’s looking very promising thus far.

Retopo’ing a Stink Dragon

…and an Orc:

He is working to optimize a few remaining areas, but a new build with this feature should be eminent. They also have a booth at Siggraph, so you should be able to see it demonstrated there, if you’re in the vicinity.



I think it would be useful if he added an extra option where you can paint where you want more mesh density…

Regardless this is badass.


OMG there is few flows in the edge flow but it looks amazing netherless!! it’s magic really, such a progress from what we had to do before! can’t beleive it.




Amazing feature! 3DCoat is getting more and more attractive :slight_smile:


Omg - Wtf - Etc…!!!


The tech itself is cool but I’d never consider this kind of topology for animation. It’s OK for static environmental elements like sculptures, rock formations, buildings etc but not for characters.

No replacement for a human brain yet.


That may be the case in general terms, but if you noticed, the more information you feed it with guides, the closer it is to what you want…plus there will always be some degree of tweaking. That’s the case when we think we are done with a manual job and end up spending more time fine tuning things.

You can remove or add edgeloops at a single click, as well as slide edgeloops, so tweaking can be done pretty fast. On the whole, this still cuts ANY retopo job down to a fraction of what it was.


I totally agree. This is going to make my job SO much easier.




looks fantastic! how long we need to wait for that baby? :wink:

3DCoat FTW! :slight_smile:


Looks fantastic! I agree with Laa-Yosh that it’s not really there in terms of functional geometry, but for an (almost) entirely automated procedure it’s hard to argue with those results.


incredible!! I will buy 3Dcoat just for that feature.


thats perfectly ok though. game work usually has a LOT more prop/ environment work than characters. Characters are such a small portion of cg that its not that big a deal. It’s just what people pay attention to most. I would totally use that for characters. You can always scrap sections of that cage and redo sections all you want but overall I can see this speeding up one of the most mundane modelling processes big time. I mean, think of that dragon video. that is 4 minutes of work…That’s nuts.


great stuff, lets hope the tech is in zb4 as well

looks amazing for props and non-deforing objects, as well as an amazing starting piont for characters!


Is there a youtube link? I cant view avi’s :frowning:


VLC, it works on osx, linux and windows and plays pretty much anything.


It’s curious how its that this has been researched for years, but there were not many, or no attempts to apply these papers for cg. I had very similar ideas long before I realized that this was in fact very popular extended subject.

But current remeshing methods are yet too new, compared to what could be possible.


Don’t have VLC on my work comp. Tried it in everything else including RV but no go :frowning:


wow… if the guy in the video would have took a moment to guide the goggles he’d have been left with a quick tweak job. That’s amazing and undeniably useful.


Try this:

Install as default and it should work. Probably some codec is broken in your machine or the browser plugin is defect.