I'm trying to render a fairly complex lamp using Three.js: https://sayduck.com/3d/xhcn
The product is split up in multiple meshes similar to this one:
The main issue is that I also need to use transparent PNG textures (in order to achieve the complex shape while keeping polygon counts low) like this:
As you can see from the live demo, this gives really weird results, especially when rotating the camera around the lamp - I believe due to z-ordering of the meshes.
I've been reading answers to similar questions on SO, like https://stackoverflow.com/a/15995475/5974754 or https://stackoverflow.com/a/37651610/5974754 to get an understanding of the underlying mechanism of how transparency is handled in Three.js and WebGL.
I think that in theory, what I need to do is, each frame, explicitly define a renderOrder for each mesh with a transparent texture (because the order based on distance to camera changes when moving around), so that Three.js knows which pixel is currently closest to the camera.
However, even ignoring for the moment that explicitly setting the order each frame seems far from trivial, I am not sure I understand how to set this order theoretically.
My meshes have fairly complex shapes and are quite intertwined, which means that from a given camera angle, some part of mesh A can be closer to the camera than some part of mesh B, while somewhere else, part of mesh B are closer.
In this situation, it seems impossible to define a closer mesh, and thus a proper renderOrder.
Have I understood correctly, and this is basically reaching the limits of what WebGL can handle?
Otherwise, if this is doable, is the approach with two render scenes (one for opaque meshes first, then one for transparent ones ordered back to front) the right one? How should I go about defining the back to front renderOrder the way that Three.js expects?
Thanks a lot for your help!
Related
I compose multiple STLs for 3D printing / milling. For that I also use CSG and need some raytracing for detecting features of the models.
My scene is pretty much static. Just have to move around the models to arrange them. For this use case I'm not really sure which approach for moving / rotating the models is right.
Currently I manipulate the BufferGeometries directly. So everything in the geometry is like in the real world. Each position, each normal. No calculation from / to local or world coordinates.
On the other hand I could do the same thing with changing the meshes, which means to change just a matrix.
For me, working with the mesh is more for animation etc. While working with the geometry to manipulate the real object, which is my intention.
I'm wondering when one would translate / rotate the geometry and when the mesh. I know that manipulating the geometry is not best for CPU, which is not a problem for my use case.
Geometry can be translated so that subsequent transformations (such as scale or rotation) originate from a more preferred vector. Meshes can share a geometry. There are unique use cases for either if you care to memorize the list. Sometimes I integrate preexisting code samples. Sometimes the decision is made for me by some aspect of the process. As for the properties which may be similar, which is more convenient? I like the pattern of modifying an Object3D dummy using those methods and then updating from its matrix. There's a whole book on normals, but I didn't write it, sadly...
Hello I'm trying to archive the effect in the image below (that is like shine light but only on top of the raw image)
Unfortunately I can not figure out how to do it, tried some shaders and assets from the asset store, but so far no one has worked, also I dont know much about shaders.
The raw image is an ui element, and renders a render texture that is being captured by a camera.
I'm totally lost here, any kind of help will be appreciated, how to make that effect?
Fresnel shaders use the difference between the surface normal and the view vector to detect which pixels are facing the viewer and which aren't. A UI plane will always face the user, so no luck there.
Solving this with shaders can be done in two ways - either you bake a normal map of the imagined "curvature" of the outer edge (example), or you create a signed distance field (example), or some similar method which maps the distance to the edge. A normal map would probably allow for the most complex effects, and i am sure that some fresnel shaders could work with that too. It does however require you to make a model of the shape and bake the normals from that.
A signed distance field on the other hand can be generated with script from an image, so if you have a lot of images, it might be the fastest approach. Getting the edge distance in real time inside the shader would not really work since you'd have to sample a very large amount of neighboring pixels, which might make the shader 10-20 times slower depending on how thick you need the edge to be.
If you don't need the image to be that dynamic, then maybe just creating an inner glow black/white texture in Photoshop and overlaying it using an additive shader would work better for you. If you don't know how to write shaders, then maybe the two above approaches are a bit of a tall order.
I am curious about the limits of three.js. The following question is asked mainly as a challenge, not because I actually need the specific knowledge/code right away.
Say you have a game/simulation world model around a sphere geometry representing a planet, like the worlds of the game Populous. The resolution of polygons and textures is sufficient to look smooth when the globe fills the view of an ordinary camera. There are animated macroscopic objects on the surface.
The challenge is to project everything from the model to a global map projection on the screen in real time. The choice of projection is yours, but it must be seamless/continuous, and it must be possible for the user to rotate it, placing any point on the planet surface in the center of the screen. (It is not an option to maintain an alternative model of the world only for visualization.)
There are no limits on the number of cameras etc. allowed, but the performance must be expected to be "realtime", say two-figured FPS or more.
I don't expect ayn proof in the form of a running application (although that would be cool), but some explanation as to how it could be done.
My own initial idea is to place a lot of cameras, in fact one for every pixel in the map projection, around the globe, within a Group object that is attached to some kind of orbit controls (with rotation only), but I expect the number of object culling operations to become a huge performance issue. I am sure there must exist more elegant (and faster) solutions. :-)
why not just use a spherical camera-model (think a 360° camera) and virtually put it in the center of the sphere? So this camera would (if it were physically possible) be wrapped all around the sphere, looking toward the center from all directions.
This camera could be implemented in shaders (instead of the regular projection-matrix) and would produce an equirectangular image of the planet-surface (or in fact any other projection you want, like spherical mercator-projection).
As far as I can tell the vertex-shader can implement any projection you want and it doesn't need to represent a camera that is physically possible. It just needs to produce consistent clip-space coordinates for all vertices. Fragment-Shaders for lighting would still need to operate on the original coordinates, normals etc. but that should be achievable. So the vertex-shader would just need compute (x,y,z) => (phi,theta,r) and go on with that.
Occlusion-culling would need to be disabled, but iirc three.js doesn't do that anyway.
I want to swap textures on the faces of a rotating cube whenever they face away from the camera; detecting these faces is not entirely equivalent to, but very similar to hidden surface removal. Is it possible to hook into the builtin backface culling function/depth buffer to determine this, particularly if I want to extend this to more complex polygons?
There is a simple solution using dot products described here but I am wondering if it's possible to hook into existing functions.
I don't think you can hook into internal WebGL processing but even if it's possible it wouldn't be the best way for you. Doing this would need your GPU to switch current texture on triangle-by-triangle basis, messing up with internal caches, etc. and in general - our GPUs don't like if commands.
You can however render your mesh with first texture and face culling enabled, then set your cull face direction to opposite and render your mesh with 2nd texture. This way you'll get different texture on front- and back-facing triangles.
As I think of it - if you have a correct mesh that uses face culling you shouldn't get any difference, because you'll never see a back-face, so the only ways it could be useful is for transparent meshes or not-so-correctly closed ones, like billboards. If you want to use this approach with transparency then you'll need to carefully pick the correct rendering order.
I'm trying and failing to work out how to achieve a quad-tree of materials (images) on a single plane, much like a Google Maps-style zoomable tile that gets more accurate the closer you get.
In short, I want to be able to have a 1x1 image texture (covering a plane that is 256 units wide and tall) that can then be replaced with a 2x2 texture, that can then be replaced with a 4x4 texture, and so on.
Like the image example below…
Ideally, I want to avoid having to create a different plane for each zoom level / number of segments. A perfect solution would allow me to break a single plane into 8x8 segments (highest zoom) and update the number of textures on the fly. So it would start with a 1x1 texture across all 64 (8x8) segments, then change into a 2x2 texture with each texture covering 4x4 segments, and so on.
Unfortunately, I can't work out how to do this. I explored setting the materialIndex for each face but you aren't able to update those after the first render so that wouldn't work. I've tried looking into UV coordinates but I don't understand how it would work in this situation, nor how to actually implement that in Three.js – there is little in the way of documentation / examples for this specific case.
A vertex shader is another option that came up in research, but again I don't know enough to understand how to construct that.
I'd appreciate any and all help with this, it will be a technique that proves valuable for other Three.js users I'm sure.
Not 100% sure what you are trying to do, whether you are talking about texture atlasing (looking up and different textures based on current setting/zooms) but if you are looking for quad-tree based texturing that increases in detail as you zoom in then this is essentially what mipmaping is and does.
(It can be also be used to do all sorts of weird things because of that, but that's another adventure entirely)
Generally mipmapping is automatic based on the filtering you use - however it sounds like you need more control over it.
I created an example hidden away in the three.js source tree which may help:
http://mrdoob.github.com/three.js/examples/webgl_materials_texture_manualmipmap.html
Which shows you how to load each mipmap level in manually, rather than have it just be automatically generated.
HTH