QuaziGNRLnose's Forum Posts

  • Hasuak

    I have no plans for a demo version.

    You can achieve flat shading like in that picture, but the ambient occlusion isn't something that can be easily achieved (i'm not sure what you're asking, if flat shading is possible, or if AO is possible). You could get a look close to that though with shadows and lighting and perhaps fog.

    Some 3D is visible in the editor, but it's limited since C2 doesn't allow plugin developers do to things that modify the editor. Right now you can mock up scenes with boxes from a top/side view, there is no "movable" camera even though i'd like to add one. The default camera setup matches the view in the editor. hopefully C3 will allow me to do more. In practice it's not a big chore to set things up even with the limited editor view.

    The light mapping isn't done in C2 (again editor limitations prevent me from adding this kind of thing, and hopefully it'll be possible with C3), it'd be up to your tool, but theres support for secondary UV's for a lightmap texture if you want to apply a lightmap to objects made in your 3D software independent of the diffuse map.

  • I exported from Modo to Blender with a fbx file, then set up Blender with an addon to export to json.

    When I try to run the project, I am getting this error:

    I also do not see any animation data in the json file when I open it with wordpad as was suggested on the previous page of this thread.

    Something is not exporting properly i guess. someone explained a bit about the necessary set up here : (https://github.com/mrdoob/three.js/issues/6554)

    Also read the error, and take it's suggestion. You need to add the behavior to the object and add a Q3D bone for a skeletal animation to work.

  • QuaziGNRLnose

    But, what if this object - geometry figure?

    Ive explained this a few times so check back if u want more detail:

    You can load named geometry created with q3d master onto q3d models by writing "mygeometry.geom" as a model filename like you would for .obj/.json. This allows Q3D models to recognize theyre going to use a created geometry instead of download a model. So lets say you create a torus, name it "MyTorus", then with Q3D MODEL write "MyTorus.geom" where a model filename normally goes and it'll load the geometry.

  • How to make obj file work on this?

    Make sure when you export that you triangulate faces and export normals for .obj, otherwise the model may not be visible.

  • You can't do anything like that with objects created with Q3D Master, and you shouldn't really be using those objects very much since Q3D Models are better optimized in all cases for memory management / recycling. The objects created with Q3D Master are a remnant of the original system and are going to be deprecated eventually.

  • Try Construct 3

    Develop games in your browser. Powerful, performant & highly capable.

    Try Now Construct 3 users don't see these ads
  • Why not use the search the function to narrow down the actions?

  • Ban anyone who posts Korean characters on a new account ?

  • LittleStain

    Thats all we have set up sadly. You could try a prepaid credit card from a general store/pharmacy maybe?

  • Me and davio are brothers and work on things together, weve always been a team

    Those actions are deprecated, use the object "q3d light" instead.

  • andreyin

    Here's an example of some various camera controllers set up in one capx using events:

    Camera Controllers

    You'll really need to look around online and learn some math about 3D vectors, trig, and rotation matrices to get a good idea of how to work in 3D and come up with these kinds of things. the plugin has helper functions like "look at" which are difficult to implement for beginners, but orbiting cameras require some knowledge of how vectors work.

  • Everything ralph wrote is correct. Obj format doesnt support animation data. Youll have to install one of the exporters and pass it through blender. Blender can import a variety of formats. Im investigating new exporter options too.

  • The performance complaints are overblown, most users would be hard pressed to get better performance coding directly in javascript. C2 works well if your reasonable and understand the performance on desktops is good but mobiles are still pretty terrible with javascript and youll have to make much simpler/lower resolution games on mobile.

  • If you have debug boxes/collider enabled the performance will be much lower, make sure you turn these off. also, turning on "shared material" can improve performance if all models look the same and don't need individual control over textures/colors/ etc.

    Q3D models will also be slower BECAUSE they support behaviours/collisions/animation/animated textures etc. and need to handle these things. Really theres nothing i can do, you need to work within the performance of the plugins, there isnt anything i can do to make the performance way better without usability sacrifices. Plain and simple you cant create too many objects because this is javascript / a browser, you've got to make the most of the performance you can get like you would in any other case. Q3D is already designed to perform very well, the rest is up to your design choices, and proper application of reasonable practices.

  • Ralph

    You dont need to do anything with shaders for this to work. Also you fuse the geometry in your modeling software. Itd be pretty slow to do it at runtime for a large number of objects. I could try to implement something for static objects but the issue is then that materials and things would cease to be modifiable.

    Its still extra overhead, and the string manipulation would be even slower than an if. Whats the issue with separate actions? Its essentially the same.

  • learningthreejs.com/blog/2011/10/05/performance-merging-geometry

    does this above works same if we add object as child ? will this save some performance if it is positioned as child "for example "3d healthbar" instead of using using set positions every tick ?

    No its not the same, children actually have worse performance than seperate objects in some cases. That article is referring to the practice of lets say making blades of grass all part of the same model instead of seperate objects to make the cpu do less work in calculating only a single transform vs 1000s. This is good practice to delegate work to the gpu.