We are currently trying to lay mesh colliders onto our edges as shown in the pictures. The problem is that the meshes sometimes seem to be 2D instead of 3D (shown in Picture 2 and Picture 3), which makes them unselectable from certain camera-angles. Sometimes the meshes even seem to disappear through some parts of the Edge(Picture 1).
Turning convex on for the colliders makes them way easier to select, but we dont really want to do that because that makes it realy unclear which edge you are selecting right now.
We are creating our meshes through bakeMesh from our previously created Edges as shown below:
LineRenderer lineRenderer = gameEdge.GetComponent<LineRenderer>();
MeshCollider meshCollider = gameEdge.AddComponent<MeshCollider>();
Mesh mesh = new Mesh();
lineRenderer.BakeMesh(mesh, Camera.main, false);
meshCollider.sharedMesh = mesh;
meshCollider.convex = false;
Edit:
We used this https://github.com/mattatz/unity-tubular to generate tube meshes around our edges, working pretty well now!
The mesh generated by the line renderer is actually 2D. I think that your best chance is to update the mesh orientation to face to the camera so that the mesh will always be facing to the camera wherever you are looking from. That way you'll allways be able to click on it.
You have 2 options:
1.- Passing the new camera (with its new position) all the time. This would be in the Update. Be carefull to only use the code that bakes the mesh and not the Mesh mesh = new Mesh() in the update. Because if you create a new mesh in the Update passed some time you'll have an stack overflow error. I recomend if you do this, to make the Mesh mesh; a class variable and start it only in the class initialization (mesh = new Mesh();) so that you use only the created instance to update the mesh all the time, not creating a new one for each update.
2.- If you are concerned about efficiency, you can handle when your camera is moving, so that when it stops moving, you pass in the new camera along with its position. For this you would need to handle the camera OnStopMoving event yourself to pass the new camera in, so whenever the Main camera stops moving, the mesh will always be facing it.
This makes sense because its easier for the user to click on things while stopped, so it can be presumed that the user will try to click on the lines while not moving
There is one third option I did not mention due to feasability, that is wrapping your line renderer with primitive colliders with the points of your line as an starting point for the procedural collider wrapping logic. However you would need to code all that out, which might take a while.
On the other hand making it a convex collider as far as I checked is not feasable, as the behaviour and shape of the collider itself changes on the mesh cooking by the MeshRenderer component. Check if this might be of help.
Related
I have a mesh that updates it's shape according to the user's input, how can I get the collider to match the new shape of the mesh ?
As an important note : the mesh is always always convex ( no exceptions ). I saw that it is important in order for the collider to work properly.
I have found the following video that does it here but the collider is simply erased and recalculated for each frame by doing meshCollider.sharedMesh = null and then meshCollider.sharedMesh = updatedMesh.
Since I am already calculating the position of each vertex, edge and triangle of my mesh, is it possible to me to use those values for the mesh collider ?
No way to do it as colliders do not have constructors for this. What you are doing is explicitly frowned upon in unity docs
'You should not modify mesh geometry that is used for colliders because the physics engine has to rebuild an internal mesh collision acceleration structure every time you change the mesh. This causes a substantial performance overhead. For meshes that need to collide and change at runtime, it is often better to approximate the mesh shape with primitive colliders like capsules, spheres and boxes.'
One work around is to use InvokeRepeating on your method and create the new mesh collider every second instead of every frame (you can test and make it 0.1 second , 2 second, whatever can run without frame drops). You can call CancelInvoke when the object is done shifting.
The Problem: Mesh colliders on some rigidbody objects are passing through colliders on other objects.
Things I have tried:
With the assumption A is a GameObject with a RigidBody attached and B is a normal GameObject with a collider.
Give A a convex mesh collider
Give A a non-convex mesh collider
Give B a convex mesh collider
Give B a non-convex mesh collider
give B a box collider
give B a convex mesh and box collider
give B a non-convex mesh and box collider
Adjusting the weight of the rigidbody
I have tried all of these in all combinations of A and B.
In addition,
Colliders are not marked as triggers
All objects are on the default layer (0)
Checking isKinematic; doing this seemed to make gravity stop affecting the object, so I have left it as false.
Constraints: I want A to use a mesh collider since most of the objects involved are moderately complex, and fitting other colliders to them would take a while.
Weird Behaviour: I have some objects with both rigidbody and convex mesh collider where the collision is working fine with a non-convex mesh collider. This is inconsistent with other gameobjects. The objects have all of the same settings.
I am using unity version 2019.3.11f1 if that is relevant.
The object being used are from this package. Specifically, the filing cabinet with rigidbodies on the drawers works fine. The desk, office chair, pen, and open laptop all fall through the "floor" (a cube with all of the above colliders tested on it).
Do you have 'isKinematic' checked on the objects with rigidbodies that are going through other colliders? If so uncheck it so that external forces affect it.
edit you also need to click convex on the mesh colliders if they are colliding with other mesh colliders, Convex Mesh Colliders are limited to 255 triangles, are the objects that are not passing through have more than 255 triangles in geometry?
I am Assuming you have set your Collision Detection to Discrete In your gameObject's RigidBody, if so, please make sure to select Collision Detection to Continuous In your gameObject's RigidBody.
Reason Why this is not working
You are trying to use collision that has (speed > your Computer's frame speed) so, the frame is not catching the collider properly some time catches and some time fails to catch.
You are moving your GameObjects with tranform.translate or position etc. If so then make sure to use Rigidbody Related function for Positioning, rotating.
Solution Image For First Problem
I'm assuming
-you game object which contain the mesh component is very small ( bellow 0.7 size(x,y,z) so the game engine is little difficult to detect it)
and also
for the colliding objects at-least one rigid-body component must be attach to one object
This seems to be a common big problem, and I have been struggling with the same issue for days, and have finally fixed it, so am feeling like I should highlight one important thing:
move Rigidbody via MovePosition, not via just changing its position field directly, nor via changing position of GameObject holding it.
Changing position via "wrong" ways will "teleport" your rigidbody, physics won't fully work on it in that case.
I am building a 3d Billiard game in VR using Unity3d.
I try to display a miniature copy of the billiard "table" (actually a cube), like it is at any given moment.
I tried instantiating the root of the billiard hierarchy, just to find out it will instantiate the original gameObject (runnings its Start() methods) which totally makes sense, just not what i am trying to do.
In practice I have a billiard root, which has all the geometry of the table, and all balls as children, those balls can interact physically.
On button press I try to create a hologram of the table, with all its balls at their position at any given time.
So If a player hits a ball and it moves in the original, it should display the same in the miniature.
I thought it might be possible to translate the ball positions (and table rotation etc) every frame to the miniature. but that seams very un optimal.
var midPoint = (leftHand.transform.position + rightHand.transform.position) / 2;
var miniature = Instantiate(gameObject, midPoint, transform.rotation);
miniature.transform.localScale = new Vector3(0.1f, 0.1f, 0.1f);
this obviously does not work as i described above. It creates a completely new instance of the billiard cube. (this code was run from a script inside the billiard root)
The board itself probably doesn't move, so you can just make holograms and simulate the balls.
Since Unity 2018.2, you can create physics scenes and simulate them manually.
For that problem you would need to:
Stop physics simulations.
Physics.autoSimulation = false;
Create a new scene for independent physics simulation.
var physicsPreviewScene = SceneManager.CreateScene("Physics Preview Scene", sceneParams);
Copy the physics-relevant objects and components to the preview scene.
Many ways to do it, but basically colliders, rigidbodies, and for objects you need visuals for, renderers.
You can use caching to avoid creating these every time.
You can also, in theory, have the simulation on the main scene, but then you need to make kinematic all the stuff you don't want simulated, and there are some gotchas.
Then simulate manually.
Get the PhysicsScene from the preview scene: var physicsScene = physicsPreviewScene.GetPhysicsScene();
Simulate: physicsScene.Simulate();
So yea, short of avoiding the components unrelated to physics in the copies, you can't avoid the duplicates; only mitigate their impact with pooling and/or caching.
Another technique is to use casting (spherecasting, in this case) to make a mockup physics simulation, stepping multiple casts with velocity, acceleration (gravity for example) and Time.fixedDeltaTime, and reacting to intercepts/hits as collisions, with Vector3.Reflector something similar. But this will only give you an approximation of how physics would react, and is not going to be an actual (accurate) physics simulation like having multiple PhysicsScenes will.
I'm currently working on an object that involves bubble-like movement. This object has a rigidbody and a sphere collider.
I use the AddForce method every 1-3 seconds to make it move continuously and slowly.
Now I'd like to know how to make the rigidbody move in the opposite direction (a.k.a bounce off) when they reach the screen edge already. I have already computed the screen edges using ViewportToWorldPoint method.
One ineffective solution that I thought of is to put empty game objects with collider at the edges but that won't work in my case since I'm building for mobile devices which have different screen resolutions/sizes.
Anyone know a good solution for this?
I'm not sure i got the idea. But i think i had the same problem when i was writing an old mobile game.
I had the same idea you did, use empty game objects with box collider on the edges, but then i thought, this isn't responsive, so i wrote this code:
public class Walls_c : MonoBehaviour {
public Transform righttop;
public Transform rightbottom;
public Transform lefttop;
public Transform leftbottom;
// Use this for initialization
void Start () {
righttop.transform.position = Camera.main.ViewportToWorldPoint(new Vector3(1,1,0));
rightbottom.transform.position = Camera.main.ViewportToWorldPoint(new Vector3(1,0,0));
lefttop.transform.position = Camera.main.ViewportToWorldPoint(new Vector3(0,1,0));
leftbottom.transform.position = Camera.main.ViewportToWorldPoint(new Vector3(0,0,0));
}
}
With this, i always get the corners of the screen. It's no fancy... but it works.
Let me now if it works.
In order to get a reflection effect, you need Material. Attach a bouncy material to the edges(gameObject) and let the physics calculate what should be the reaction of collision.
Create a Physics 2D material and set the bounceness to some appropriate value say 0.2.
Regarding your issue:
One ineffective solution that I thought of is to put empty game
objects with collider at the edges but that won't work in my case
since I'm building for mobile devices which have different screen
resolutions/sizes.
If you are working with UI component then dealing with boundaries should not be a problem (since it anchors with different resolution) but if it is not the case you can create a script CameraResizer and an enum Anchor (Left, Right, Top, Bottom) and using the same way ViewportToWorldPoint you can align your empty gameObject (boundary) to any screen size by attaching it the gameObject.
Hope it helps!
Hello Peeps (first post yay).
I have just started to learn about unity and still going through various tutorials and learning the interface along with C#.
I am running Unity 4 and downloaded a free asset from the Unity store:
https://www.assetstore.unity3d.com/#/content/3174
The Problem.
I created a Plane in Unity.
Went to the Character and created colliders and imported it as a humanoid.
Then added a rigid body = The character fell through the plane
I then added a Mesh collider and attached the mesh for the character to the collider. The mesh does NOT align correctly to the character = the character fell through the plane.
I then tried box and capsule colliders and they worked fine. Attached these and the character would not fall through.
I then tried a character controller = Character did not fall through.
I want to have an accurate mesh for this test character so when an enemy hits him, there is an actual hit.
I have tried several other free characters from the store and they all have the same problem. Can someone please tell me how I can fix this? It seems I just either doing something completely wrong, missing a step or just plain lost.
Link below to image of the issue taken:
Using a mesh collider on a skinned, animated character is a wrong approach. Mesh colliders are generally used for environment geometry and rigid objects. This is all due to the amount of work it is needed for the Unity engine to accurately calculate the shape of the mesh collider, based on the provided mesh. More triangles means more work, and for the reasonably complex low poly character mesh of 10k to 30k triangles it is clear that the fps would most definitely drop to an unplayable level.
For our game we are using all primitive colliders, distributed on the adequate bone game objects. So, head bone has a sphere collider, spine has a box collider (for lungs), forearm has a capsule collider etc. That way,with a little bit of scripting, you can detect a hit and know exactly what part of body is damaged.
If your collider doesn't exactly match your mesh, then it's likely that you have the mesh applied to the wrong part of the object.
Try applying the Mesh Collider to the mesh, instead of the complete object.
You have a pretty complex character mesh, so using mesh collider probably isn't the best way to go. In many cases, character controller or capsule collider works fine; otherwise, you can just use compound collider and create a box or a sphere around various body parts of the character.
This once happened to me. The mistake i made was, i applied the mesh collider after i did some rotation and scaling. So i added the mesh collider as soon as i imported the model. And then it should be alright when you scale/edit the model. Also check with right/left handed co-ordinate system with unity and 3D MAX/Maya.