Jump to content
3DCoat Forums

farsthary

3DCoat Developer
  • Posts

    276
  • Joined

  • Last visited

Posts posted by farsthary

  1. @ farsthary

    My question, once again. How can I add density on specific places without building anything? The opposite of reduce tool (should exist on ctrl+tool)

    Tools with 0 value still affect surfaces. Should I report it as a bug? I did it lol.

    "The build tool? have you tried LiveClay? it can quickly build up layers of fresh mesh...is the most basic and the most important brush to start with, anyway, every artist is free to have an own workflow."

    well, free... this depends on your work as well. The basic digital tools, the build, smooth(with some flatten behavior), grab, pinch and now pinch +- (Crease in our case which looks good). Clay tool as basic, where did you read this? Of course we need some more tools but this doesn't make them basics.

    In fact there're two only real basic tools. The grab tool and the reducer/better remesher +-.

    Hi Michalis :)

    Oh yes, I have being able to replicate the issue, you're right, it should not deform with Depth 0 , don't worry, I will fix it ;) thanks!

  2. I have a suggestion concerning the new snakeclay tool. The brush starts out perpendicular to the surface it starts the stroke on, but then never changes. It would be better if it could adjust it's

    orientation to maintain perpendicularity to the direction of the stroke.

    The way it works now, if you try to pull a snake out and turn a corner, the snake flattens out because the brush doesn't turn with it.

    Thanks,

    Tom

    Oh yes, that's one of the causes of the flattening effect, I'm working on that ;)

  3. @Michalis

    I see your point, to get a remesher effect, simply use LiveClay brush without deform, just set the deform curve to cero and you will get the remesh effect, on the other hand a global subdivision command can be easily added for surface mode, in early alphas it was implemented already....

    @Artman

    I may have some good news upcomming for you, the flatten smooth may be similar to a new TangentSmooth tool that will be soon included, it does quite good from clean topology, to no shrinkage smoothing/relaxation up to full shrinkage relaxation/flattening ;)

    And about a full Crease, yes, I can work on this too ;)

  4. Oh, I saw the first hole here .... I will see what that happens.... thanks ;)

    I think it may happen with bigger brushes and relative low detail subdivision and/or fast move, because faces may fall outside the influence radius before they have time to get correctly subdivided... I will need to impose some restrictions and improve this situation ;)

    The speed of stroking has a great influence in the output quality of snake, slower stroke gives smoother results while too fast you get easily stretching problems... this may happen because of the way it is handle that event currently .... a current workaround may be to use a slow/moderate speed and avoid too fast stroking. definitely I need to work on that.

  5. Hi all :)

    So summarizing.. what are the reported bugs/suggestions?

    1- rename (can be easily done)

    2- some stretching in tube and snake tool?

    3- invert behavior? (I don't see an actual invert behavior for Snake, in that sense is not a symmetrical tool, Tube invert works fine, Wrinkle? I can think something of because the current invert behavior is not the intended ;) )

    please, feel free to explain me the issues.

    Cheers

    Farsthary

  6. Not the OSX build. Snake works terribly.

    BTW Raul.

    Where is the build tool, the move tool, and the remesher (reducer) +and - ? I mean the basic tools to start with?

    You probably already noticed that I'm somehow not a positive guy. This is correct.

    I never regret saying NOW, I regret for lot of YES actually.

    Ops, sorry, I'm not responsible for cross platform builds.... I will pass your complain ;)

    The build tool? have you tried LiveClay? it can quickly build up layers of fresh mesh...is the most basic and the most important brush to start with, anyway, every artist is free to have an own workflow.

    The Move tool is on the way, and the Remesher too, along with other goodies, they can be quickly added... hey, those are Beta tools that takes time to develop and absolutely out of schedule, is systematically underestimated the complexity of creating software and improvements takes place STEP BY STEP, tool by tool ... or do you prefer Autodesk like only annualy updates?

    "Snake works terribly" please send me some screenshots and/or use case that I have missed, remember I'm not a sculptor so I can easily miss testings...that's why is released to get your feedback. (EDIT: I'm aware of a flattening effect caused by non volume conservation, is something I will definately will work on)

    There's a spanish saying: "A drop of honey attracts more flies than a gallon of gall" (in spanish is more poetic ;)) I appreciate your honesty and sharing your behavior philosophy but remember that the "means" is as important as the "message" itself, it does not help none to use harsh words for file a complain and constructive critics are better listened than rants ... now I have shared with you my "behavior philosophy" ;)

    • Like 1
  7. Hi guys :)

    A couple of Good news! new updates in the beta builds include our latest improvements and stabilization to LiveClay tools and also the new brushes: Wrinkle (can do creases like in Sculptris, use large displacements) , Tube (strong normal extrusion) and Snake (yes, what you have being waiting for ;))

    http://farsthary.wordpress.com/2011/08/27/snake-brush-a-much-awaited-liveclay-feature/

    cheers

    Farsthary

  8. Hi all :)

    First of all LiveClay is still a WIP but it has an important milestone implemented recently: Subcells or the posibility to refine any voxel hierarchical to any level ... what this mean? unlimited small details, you can sculpt a small zone and have there 10Million polygons without too much hassle and can jump to a nearby zone of the model and continue sculpting without loss of performance (of course, hardware is the limiting factor but not the software) ;) The current released beta have some issues (mesh output quality) that prevent to fully take advantage of that feature, so users can find a performance cap with very small pen radius and lots of polygons but still in that case the system is quite fast! Once fully aviable and implemented, trust me, performance won't be an issue no matter the amount of detail of the model, this will be a critical feature because 3DCoat will be able to dynamically sculpt incredible detailed models in a fully local way even on modest hardware :)

    Just for testing I was sculpting using that feature and after reaching in a very small zone (the size of a voxel) 1,5 million polygons I quit because I got bored and didn't see any performance slowdown on a slow PC! of course, getting mesh gaps and disconected faces because of the current issues I talk about, that's why this is currently not aviable for public usage.

    Previously 3DC used static voxels so in order to increase resolution you have to add globally more voxels, now you can refine any voxel into subcells and each of them into more and so on! currently is in a raw state and only functional for Surface LiveClay, it basically allows to have a hierarchical dynamic mesh partition so performance of LiveClay is fully local and is independent from the global mesh reosolution :) you can keep adding more detail and detail in smaller and smaller zones on the model as long as you want, I think this is awesome and can be used for the rest of the bruses since the required changes was minimal :)

    Currently we are working toward improving mesh quality and reducing extraordinary vertices, this is very important too since dynamic subdivision by definition introduce lots of them, the work in this area will allow to solve the current performance limitation in order to take advantage of full locality and after that we will tackle dynamic Mesh Simplification/Decimation , that is the opositface of the same coin. LiveClay includes both features: dynamic mesh subdivision and simplification.

    With this beta we aim at user feedback for tool usability and enhancements, we would like to see artist models and real life use cases since we developers tend to focus too much on testing fast usage and not heavy usage of the tool, of course this will not be the final GUI of LiveClay and expect some crashes too ;)

    Hope you enjoy it and feed us with some gorgeous models made with LiveClay :)

    Cheers

    Raul

  9. Hi all :)

    Just another quick update, I have implemented a zero shrinkage mesh relaxation that will solve the current issues with the SmoothAll tool, it is awesome because it preserves the shape in every detail while gives the advantage of the smoothing algorithms and improves topology, consider it as an upgrade of the new smooth, soon I will post in my site new screenshots and will sned the code to Andrew :)

    It also performs very very well with LiveClay.

    Cheers

    Farsthary

    • Like 2
  10. Hi All :)

    Since this forum cover a wide variety of topics and I can't track them all due to my slow internet connection I will try to focus on this thread insetead :)

    Here I will post updates and will answer community questions as well if I can.

    LiveClay is progressing very well, I'm finishing the smoothing part of the algorithm and soon will tackle extraordinary vertices that arises when the detail level between the pen and the base mesh is too different.

    After that, I think there's no big showstopper to make the first public update, Andrew will review the code and commint some design changes, I can't foretell how long will that take but not too much ;)

    Beta testers will be wellcome in this part in order to stress the new feature and discover improvements path (I'm not a modeler so I will need a good feedback to improve it further)

    Further in the roadmap is the mesh simplification feature, dynamic subdivision is not only adding detail but also removing detail.

    All of that in the Surface mode, but eventually all of those tools will be replicated into the Voxel Room to get a very powerfull Voxel LiveClay that will allow interchangeability between both modes without loosing adaptive detail and improving memmory footprint a lot :)

    Voxels are very powerfull but currently in order to sculpt fine detail on relative small zones of an object you need to double the voxel resolution in the whole model, so having LiveClay for voxels will allow to finely refine any portion of the object in a local and adaptive way :)

    More surface tools will come, polygons are inexpensive to handle by current graphics pipelines and many users find them very easy to work on, I see the voxels as the heart and the power of 3DCoat and the Polygons as the skin and the face, that way we can serve every modeling paradigm :)

    Stay tunned.... I'm very ambitious in terms of development and research so with the Lord's help an with Andrew and the Pilgway tem help I think the sky is the limit ;)!

    Cheers

    Farsthary

    • Like 4
  11. Hi all :)

    well, I can't foretell what would be the needed changes but I would like to clarify some things: Dynamic Subdivision can (and will) work with Surface mode as well with voxel mode, in theory both, voxels and polygons can have dynamic adaptive subdivision, though the final implementation may differ because both have radically different algorithm ... currently Unlimited Clay work on polygons so is starigthfoward to first target the Surface mode ;) but at the same time if we allow voxels have sub-voxels (further refined like a tree) then the Voxel mode can easily handle adaptive refinement and the output from Surface dynamic subdivision will be easily tackled, I have high hopes in the voxel mode too because voxel refinement will allow implicit dynamic subdivision surface reconstruction at no aditional cost ;) and combined that with the current arsenal of features that 3DC currently boost ... well, WOW :)

    Also, dynamic subdivision is just the top of the iceberg :) the beauty of UC is that it aims at mixing mesh/voxel operators and sculpting, that way in the future, could be easily extended to allow dynamic reduction (decimation) among other things :good:

    Cheers

    Farsthary

×
×
  • Create New...