Synthesizer

Hi everyone,

I'm working in XNA (C#), & after successfully making my character slide along walls (character represented as a BoundingSphere & wall as a BoundingBox) I've been strugling to solve the well known corner problem.

I've been decomposing the BoundingBox of the wall (AABB) into Planes in order to check which one was hit, descard the further ones, try to work with normals & push the player somehow away from the corner direction (in case it hits the outer part of a corner).
The problem is that there are plenty of different possible situations...

I manage my walls as BoundingBoxes, & each wall is one of them. So in a corner there are 2 BoundingBoxes which come very close one to the other. So there are many planes (outer & inner) that may be hit, from both BoundingBoxes. It's actually very difficult to manage many planes of many BoundingBoxes & get an algorith (which works for inner corners: you are indoors, & outer corners: you out outdoors).

I'll try to show my issue in a pic (seen from above), & I'd like you to help me in thinking about a solution. Is it a mistake to decompose the BB into Planes & work with them individually? (maybe I'm complicating things too much...) May I cast rays...? May I treat the corners with a fixed resolution?
I hope you can guide me somehow, I've really spent a lot of time thinking about solutions & they always match a new case in which they won't work.

Thank you so much,
Synth

P.S.: Oh, I also have an annoying problem in which when I slide along a wall I can see the interior of the "room" (I see across the wall) & I'd like to get rid of this in an elegant way...!

Hi everyone!

I'm having a little issue when trying to decompose a view matrix:

I basically have a spot light which works fine.
I added an alpha blended cone model to create a light volume effect. It looks allright & everything is fine except its rotation.
I thought I could probably create a view matrix for the light (with position & target) & decompose it to get an automatically made rotation matrix (to apply to the cone & get the same rotation as the light).
But when I decompose it, the values I get don't seem to be correct. The translation matrix is not working for me & the rotation doesn't seem to be right.

I wonder why is this happening &, in case I cannot achieve it like this, how may I get a rotation from a source & a target point in 3D (this http://www.chriskugl...en-two-vectors/ didn't work for me, maybe I'm doing smth wrong, but I don't know...)

Here's how I create the view matrix & then I decompose it:
[source lang="csharp"]light.view = Matrix.CreateLookAt(light.position, myPlayer.animatedModel.position, Vector3.Up);bool deco=light.view.Decompose(out scaleView, out rotationView, out translationView);[/source]


I also attach a pic showing the light (following the player) & the cone (moving in a way that sometimes matches the light movement but it's not right). The light diameter is exagerated to illustrate the case, the idea is that the cone is scaled depending on the light diameter, I just have to tweak it a bit somehow to match). The cone doesn't seem to move in X, it moves in Z following the light movement but not matching it. It moves in a very restricted range & X movement seems to be rendered in Z.

Thanks in advance,

Synth

P.S.: I'm not sure I'm posting this in the right forum because it mixes math, graphics & xna elements, so forgive me if I'm wrong. Thanks.

Hi,

I've been programming in C# for a long time & one thing I'd love to have there (optionally) are pointers.

I've been developing a game project for several years, & I usually declare attributes in my classes in order to store other handy objects.
For example, I've created a custom ModelMesh in which I store it's parent (container) Model:

class ExtendedModelMesh()
{
  -Model theModelWhoContainsMe;
}

I do this for example to be able to access easy & quickly to the model's position, rotation, scale attributes & some other things.


My question is: Is it a performance killer? (in this particular case I store the same model along all it's meshes (imagine 150 meshes per model).
Is it not only eating a lot of memory but also slowing down things when fetching up the model?

Thanks a lot,

Synth

Hello,
I think I may be having a bottleneck when passing parameters to my HLSL effects.
I'm using Deferred Rendering & each model knows how to draw itself (has a draw method).
I tell each model from outside which technique to apply or if to use a specified technique within the model (previously set in loading time).

They always use the effect set by my custom model processor.

I arranged everything so the shadowing technique & the normal drawing technique are inside of the same effect file, so i always use the same effect, the same draw method & the same parameters: depending on the technique I send to the unused params "Matrix.Identity" for example (just to send something & let it work).

But I think that may not be very good...

I'm thinking on splitting the effect files, & specifying them during runtime (just like I do with the Techniques). But the problem is the parameters in the draw method of the model. Some time ago I was having a desing in which the models didn't know how to draw themselves & were drawn by a "Deferred Renderer", a "Shadow Renderer", etc. I keep some of this design, but the final "draw" thing is inside of each model (even I think it was a better design, I changed it because of compatibility with octrees, where you don't know what u put inside, but they just "draw" themselves).

So... how do u advice me to face this problem?
May the bottleneck actually be there or may I be completely off aim? (is it possible to profile this kind of thins in PIX? Do u advice me any other, maybe better, software?)

Thanks a lot
Synth

I'm trying to optimize (once again) my XNA game project.
I've been using an octree for a long time ONLY to check collisions but not to draw, & now I'd like to use it to draw as well.
Nowadays in order to draw I'm looping through all the objects & drawing them if they are in the view frustum.

I'm currently building my models by myself (in 3Ds Max), & a Model usually contains many parts (Meshes). It may be a telephone (with buttons, wire, etc), or a train station (with stairs, lamp posts, rails, house, bridge, etc, etc, etc). Because of this, in order to check for collisions, I store the Meshes in each octree leaf (not the models).
It works good because I can identify if I'm standing over a stairstep or if i collided onto a solid wall or so.

The main issue is that the "draw" method is in my Model & not in my Mesh (& I cannot build a draw method per mesh because the model applies the corresponding transformations to its meshes when drawing them, inside "draw" method).
I thought of storing a reference to the model on each mesh, so when looping through the meshes in the octree, I call the draw method of the mesh owner's model. But the problem is that I will find several meshes which belong to the same model...

It's a mess...
I hope you can help me!

Thanks a lot for reading & helping!
Synth