The complete absence of sculpting development is a bit disappointing.
It may also generates conspiracy scenarios, I wouldn’t like to discuss either.
I’m not talking about the “multires” so called issue, which faces a dead end. What about further development of dyntopo? What about better solutions on retopology? What about holes and join tools in dyntopo?
On the other hand, who cares. Blender is a tool for making movies (ha), and better use zbrush for the rest.
I own zbrush and still waiting for more fresh and simpler ideas on sculpting tools. Blender solutions, please.
Grant for Nicholas Bishop, to work six weeks in Blender Institute Amsterdam
on sculpting tools in Blender.
If I remember correctly, he did a few years ago a Ptex build so maybe he hopefully doesn´t have to do too much
and can squeeze some sculpt improvements into his funded time or he gets anohter funding exclusively for the ptex part.
PTex is the one I’ve seen that is related to dyntopo. Which eliminates the need to unwrap and retopo. Even then that still doesn’t solve a long term need like having an all-quad, animatable mesh, which is only possible if it’s retopoed. But it seems no matter how many retopo tools are used, it still takes hours and hours of laborious work. Which makes dyntopo a discouraging addition to the workflow. (You might as well sub-d model it.)
He did have working Ptex in Blender several years ago, but it was ultimately shelved in part because there would’ve been no way for Blender users to render it without getting a commercial engine.
Also, about the concerns over retopo tools, I don’t think we have specifics on what Campbell will be working on during his months of exclusive work on the modeling tools (I would think tools that would come in handy for retopo would fit in that category).
Also regarding the pessimistic view by Jpb06 of the asset browser, I don’t know if you’ve been following the project at all, there’s already been quite a bit of work put into it and so far it’s far more than just a repackaged GIT/SVN (check the mailing lists for the ‘asset experiments’ branch).
In part, if it says > 90 percent, that means that the work has already been ongoing for a while.
I’d like to see improved Dynatopo and a feature that would allow using 3D curves as guides for sculpting/painting on the model. Also some automatic way of generating masks when baking hi-poly to low poly (unless it’s already possible somehow).
Well it’s more a case of someone being willing to develop more tools for it - a lot of the items on the roadmap are pretty significant and deserve to be high priority for where the Blender Development funds go, or are far into development and won’t take much to finish. If a developer has the time to develop the tools, has interest in that subject and can successfully collaborate with artists to create some good feature implementations, then you’ll see features.
From what i’ve seen, BF can’t even afford that many developers (although what they have been able to do is still impressive), so you can’t see people consistently working on specific feature sets or program areas unless someone donates their time
Hey, at least ReptoFlow is adding excellent retopology tools that will hopefully lead to an automated retopology solution for sculptors?
A creative new year to all !!!
BTW the engraving I posted above is one of Goya’s masterpieces. What it says, no need to translate.
IMO, we should expect an improvement in dyntopo performance, join and holes tools. Better retopo and/or autoretopolgy tools.
With the addition of Ptex and possibly openSubd, we will be OK. The recent cycles baking works beautifully. Painting goes very well too.
It wouldn’t surprise me if we see multires trashed.
Does anyone know if Bastien’s Asset Browser work is related to Elubie’s Asset Browser development? Are these two seperate projects? Is Elubie’s on the back burrner or been scrapped?
The Cycles render engine now has more or less the features we want (although baked rendering and ‘shadow catcher’ material is still high on the list).
I have not seen the words " shadow catcher" out of any developer roadmap for years, it’s refreshing to hear at least it hasn’t been completely forgotten. While I don’t hold hope that it will happen in 2015, perhaps some miracle will happen
Thanks to the foundation, and all the developers who work so hard on continuing to improve and develop Blender. 2015 looks to be another exciting year.
PTex doesn’t eliminate the need to unwrap and retopo at all if your outputting to a game engine. If your animating, then retopo is absolutely necessary. The only time you don’t need to unwrap is if your outputing to a render engine that also supports rendering PTex maps. And even then you still have to retopo.
Dynotopo is NOT a discouraging addition at all if you understand a proper sculpting and modeling workflow. Dynotopo along with PTex allows for a extremely freeform and creative process at the start of the modeling phase, allowing you to focus on form and painting without having to worry about uvs or topology. Once you have the model looking the way you want, THEN you start the laborious process of retopo and uv unwrapping. But by then you’ve already worked out the models form so you dont have to mix the technical side with the creative side.
Once you have your retopologized and uv unwrapped model, you can then bake the dynotopo into a displacement (or normal) map and the PTex map gets baked onto the new UVs (or new PTex on the new model).
This process absolutely depends on a good baking toolset and workflow as well, to minimize cleanup of the newly baked maps. I haven’t done a ton of baking in Blender, can anyone chime in on their experiences with baking displacement/normal info from a dynotopo model to a retopologized model? PTex will need to do a good job of baking to a new mesh as well, I hope the devs don’t overlook that portion of the tolset when integrating PTex.
Auto retopology would be a great addition to Dynamic Topology, and Blender in general. But then every 3d package would have it if it’s that easy. Retopology requires human intervention, that’s a given.
But I thought they got close to solving the puzzle when Bsurfaces arrived. Unfortunately, the main devs were unable to see the value of incorporating Grease Pencil strokes in retopology workflows. Ecclectiel introduced it, but no one had the interest in enhancing it, including the Grease Pencil author.
Just think about using pencil strokes in guiding the topology flow. Only Blender has Grease Pencil feature. And Ton should explore this valuable asset further and put some time and funds into it. Grease Pencil strokes capable of being edited via bezier handles, then auto retopo would be a powerful workflow in retopology chores.
I would take a serious look at the Blender Market’s retopology tools. They are some of the best I’ve ever seen in any package. Yeah, it would be great if these tools came with Blender, but it doesn’t change the fact that the tools are here now, they are affordable and they are awesome!
EDIT: But yeah, an auto retopology tool like the one in zbrush would be outstanding nevertheless.
I am really looking forward to better viewport shading!! Custom glsl code insertion , good integration with the node editor too.
It will take quite a while to catch up with Maya’s shaderFX, which can also output shaders to be used in external game engines (or so ADSK claims) - but it should be possible . http://www.3dtutorials.org/video/7051/autodesk-maya-2015-shaderfx/
Shaderfx has some neat features I wish blender could get:
the ability to expose the shader group inputs at the shader’s properties edtior (attribute editor in maya)
The ability to expose file input to the node group (input).
The ability to pin the group input and output nodes on the screen
The option for a preview window at the mix shader node
The ability to use custom code as a node (kind of like what cycles does with OSL, but with glsl/hlsl)
some other useful things I wish blender node groups could do:
Color code specific values, not just the entire node group
Boolean switches (tick box) to switch on/off things- its nicer than the slider
Ability to expose a normal dot input to the node group (embed collapsible normal dot inside the node group node)
ptex is also pretty exciting, asset management and browsing could open up some doors to a better marketplace infrastructure… The biggest one for me is the animation system. better framerate alone would make a huge difference!
One thing I’d like to be corrected in blender is the way the program handle history and test renders, too heavy on the ram and v-ram. This way the program crashes often and you have to remember to save and restart scene file. If I am not wrong other programs (3d or compositing-videoediting) save temporary scene data on hard disk.
It was rather advanced actually, as when you made a stroke, you could later move your cursor at the end of one of your strokes and continue the lines, something the grease pencil does not allow you to.
I’ve heard about Zbrush’s auto retopo, and I assume the result is not always the desired topology. (It’s probably just the equivalent of Blender’s shrinkwrap modifier. I don’t know.)
I think such auto program should require essential information first before it is run, such as telling it how the direction of the topology should flow. Bsurfaces is exactly that but in smaller scales, which still requires massive amounts of time to form a complete topology. Nevertheless, it gave a clue on how autoretopo should be.
How about, as a start, for example, tracing the outline of the eye using pencil strokes, then, perhaps, click an expand button and each click builds the first layer of the mesh, either going outward or inward, or both, detecting limits…with option of entering the number of rows, columns, whatever you call them.
You may also draw concentric strokes where the outer stroke is the limit of the expansion. In sum, pencil strokes should be utilized efficiently in retopo chores… Some of us may have better ideas, that’s my take on this…
Because there’s no point in requesting anything that got removed, when was the last time you asked for the return of a removed feature and the devs got back ? remember radiosity in blender render mode or the stars one in Blender render more recently ?
Blender 2.5, the Retopo tool has been replaced by improved mesh snapping functionality. This page will change as retopology tools are updated in newer versions of Blender
