Autoretopo wish

[michalis]

Long time to see some serious development on autopo.

It seems it works a bit better on newer builds but:

Will be ever possible to have forced loops instead of generic loop guides?

Looking at the awesomeness of booleans surf mode, this reminded me the lack of auto forced loops.

A second wish. What about an anti ngons tool? A shortcut to find them anyway. Not just find/ngons click click all the time.

Newer autopo seems to work better, smoother, but produces 30-40 ngons every time.

[Roger_K]

This would be nice indeed. Also it would be great if the Auto retopo was available in the retopo room. Say if you could select an area of your high poly and engage to tool only on that area. Being able to retopo defined areas would be much more usefull than all or nothing and It removes the need to segment your sculpt

[pavig]

What I'd really like to see is an autopo-fill tool; i.e. manually topologise the parts you need, then select an area and get autopo to fill in the gaps. This would make correcting or rebuilding topology much easier, as you would only need to manually build topology in areas where edge flow is important. Just a thought.

[michalis]

@ roger, pavig

something like this? Expensive tools though... lol

http://www.headus.com.au/products/cyslice/

[Zeddicus]

What I'd really like to see is an autopo-fill tool; i.e. manually topologise the parts you need, then select an area and get autopo to fill in the gaps. This would make correcting or rebuilding topology much easier, as you would only need to manually build topology in areas where edge flow is important. Just a thought.

I really like this idea. I'm looking forward to the multi-core optimizations Andrew mentioned, which should make a feature like this work well on powerful machines. I'm having trouble with a 2m poly mesh from Zbrush right now that autopo keeps freezing up on. Manual retopo still works fine, except there are areas that are going to be really tough to do, areas some automation would be welcome on since edge flow isn't so important in these places. Being able to paint in what's missing would be incredibly helpful.

[manticor]

+1 for local autopo

[AbnRanger]

What I'd really like to see is an autopo-fill tool; i.e. manually topologise the parts you need, then select an area and get autopo to fill in the gaps. This would make correcting or rebuilding topology much easier, as you would only need to manually build topology in areas where edge flow is important. Just a thought.

I don't think they will work well together on the same mesh, because on one hand you are largely letting 3D Coat dictate the topology, and the other, you are. It's like a movie producer having two different directors on set. It might seem like a good idea, but in a pragmatic sense, it would be more headache than an aid. However, I use Autopo and Manual Retopo together often, when there are multiple (sub) components to a model, like belts, belt buckles, rings, braclets, ammo pouches, etc.

You might could test out your idea, though, by going to the Voxel Room and using the Hide Tool on the parts you do not want Autopo to topologize for you. Then unhide once you have the mesh it created.

[michalis]

Nice tip Abn. Not much different to just autopo and delete the areas you don't like. But it's very useful and less messy.

I was talking about forced, really forced loop guides, this was my wish, and if I recall it correctly, it was your wish too.

Something like this, though I haven't test it. here

Notice how many n-gons, lot of edit again.

[cookepuss]

Long time to see some serious development on autopo.

Yeah. I'd much rather have Autotopo get fixed than have new features. The idea of guided retopo is great, but only if it works. The reasoning behind retopology is so that you can establish a new, clean edge flow. 3DC currently produces way too many ngons and triangles. It's practically worthless as-is. Autotopo isn't a time saver if you have to do a lot of cleanup. Nobody likes working in an app with half-implemented features.

Clean, quad only autotopo. That's what I want.

[AbnRanger]

Yeah, and I wish Andrew would implement some kind of rule in the algorithm that forbids spiraling of loops. It can look clean at a distance, but as soon as you go to layout the UV's, You find, like you said, it to be relatively worthless. And it's a nightmare if found on a character that you need to animate. Even still. For many things, it actually is a big time-saver. I find a lot of the sub elements are quickly done with Autopo, like rings, bracelets, belts, beltloops, shoestrings,etc. Those things can eat up lot of time on their own. Using Autopo on them truly feels like cheating

[michalis]

Yeah, and I wish Andrew would implement some kind of rule in the algorithm that forbids spiraling of loops.

You're so right Abn,

You probably noticed a trick though. Cutting the edge loop perpendicular to some kind of axes that we can imagine through the volume of a hand for instance. What am I saying now... lol ?

This indeed creates a real clean loop.

No, we need forced loops. What could be a problem with these. Maybe to create more stretched quads probably. Less evenly subdivided.

But, autopo creates a lot of stretched and dirty n-gons quads and tris already. Takes some time to delete and redo it manually. Especially when a 2x density areas brush is in use.

I have a big problem with the new releases of 3dc, similar with the blender ones. Lot of tools stay unfinished, new implementations are coming. I'm a fun of these new features but the simpler is always the better. A productive tool! A stable one. Please.

For a developer this isn't a dream I guess. Neither for me as an artist. I love doodles most of all. The man kind.

[cookepuss]

I have a big problem with the new releases of 3dc, similar with the blender ones. Lot of tools stay unfinished, new implementations are coming. I'm a fun of these new features but the simpler is always the better. A productive tool! A stable one. Please.

That was always the problem with trueSpace back in the day. Caligari would implement new feature after new feature. The app would sound fantastic on paper. When you got to the meat of it, many of those new features were only ever half implemented. It got so bad that not even UNDO or booleans worked properly. They were intent on adding new features instead of fixing older broken/incomplete ones. That's the big problem I see with 3DC. Andrew is doing the best he can to keep the app competitive, but is losing sight of the fact that most artists would prefer reliable and stable over flashy and edgy. Cool new features are great, but not at the expense of the existing ones that matter most.

Autotopo needs to get fixed and stabilized sooner than later. It's the one unique topo feature that 3DC has over the competition. If Andrew doesn't fix it and make it work as advertised then 3DC will eventually fall behind as an option for artists interested 3D for its retopology tools.

Git er done.

And, yeah. Moving autotopo the retopology room makes sense. There's no reason for it to be isolated on the menu. It should be where we need it most.

[dfp123]

The spiraling of loops in Autopo absolutely needs to be fixed. At first I thought it was something I did wrong (trying to nail down a proper workflow in 3DC is rather obscure). But after filling my character with endless guide loops and still getting spirals, and reading this thread, it appears to be a bug.

I know there has been a lot of talk on the forums about 3DC branching off into new features, which look interesting, before the existing ones are working as they should; optimized and bug free. But 3DC really needs to clean up these areas immediately and get them working as they should, to speed up workflow rather than being a 'rather cool feature but not usable in production'.

I love 3DC! I recently switched music composition software (yet again), this time to a big name, because of purchasing independent titles that couldn't do the things I needed. The bullet point features made these applications look perfect but when getting into actual production with them, they fell short through bugs or features not working entirely as expected. The last application, which seemed to have a rather healthy update and bug fix schedule, had actually halted any work on the main application to produce an ipad version. After about seven months with no appreciable attention or fixes I needed in sight, I was forced to switch applications.

Now 3DC has no problems with attention and updates (the work being done on 3DC is staggering and unprecedented in the 3d graphics community). However, I don't want to see the things that have brought us to 3DC and/or kept us here, neglected for the addition of new features. I am still using 3.5.27 for fear that the application will become unstable in the areas I need to get work out. For instance, the Autopo spiraling loops and the inordinate time it takes for the fill tool to work in the Paint room are examples of tools left by the wayside for quite a while. But they need attention for others to fully embrace 3DC to be on par with or surpass the competion. I know these things have been talked about ad-nausem here in the forums, but I thought this was important enough for me to give my two cents.

Thanks for all the hard work on the application and for hearing me out.

Respectfully yours.