Request: an hello world example? about wickedengine HOT 22 CLOSED

Hi, if you are looking for how to create lights, or meshes, you can check the following functions in wiScene.cpp:

• Scene::Entity_CreateCube : this will create a simple cube mesh in the scene with vertices, indices, materials from code
• Scene::Entity_CreateLight : this will create light inside the scene, with transformation, name, etc.
• Scene_Entity_Create... there are more creation functions like this for other kinds of things

Or are you looking to use low level graphics API example to draw triangle to render target directly?

from wickedengine.

In fact I see nothing after I have created a triangle at (0, 0, 0), (1, 0, 0), (0, 1, 0), and a point light

The code is like above

	auto rootEntity = wi::ecs::CreateEntity();
	scene.transforms.Create(rootEntity);
	scene.names.Create(rootEntity) = "default";

	auto materialEntity = scene.Entity_CreateMaterial("default");
	scene.Component_Attach(materialEntity, rootEntity);

	auto& material = *scene.materials.GetComponent(materialEntity);
	material.baseColor = XMFLOAT4(0.5, 0.5, 0.5, 1);
	material.roughness = 1.0f;
	material.metalness = 1.0f;
	material.reflectance = 0.04f;

	material.SetDoubleSided(true);

	auto meshEnity = scene.Entity_CreateMesh("test");
	scene.Component_Attach(meshEnity, rootEntity);
	auto& mesh = *scene.meshes.GetComponent(meshEnity);


	mesh.subsets.push_back(wi::scene::MeshComponent::MeshSubset());

	mesh.subsets.back().materialID = 0;
	//auto material = scene.materials.GetComponent(mesh.subsets.back().materialID);

	uint32_t vertexOffset = (uint32_t)mesh.vertex_positions.size();

	mesh.indices.resize(3);

	mesh.subsets.back().indexOffset = (uint32_t)0;
	mesh.subsets.back().indexCount = (uint32_t)3;

	mesh.indices[0] = 0;
	mesh.indices[1] = 1;
	mesh.indices[2] = 2;

	mesh.vertex_positions.resize(3);
	mesh.vertex_positions[0] = XMFLOAT3(0, 0, 0);
	mesh.vertex_positions[1] = XMFLOAT3(1, 0, 0);
	mesh.vertex_positions[2] = XMFLOAT3(0, 1, 0);

	mesh.vertex_normals.resize(3);
	mesh.vertex_normals[0] = XMFLOAT3(0, 0, 1);
	mesh.vertex_normals[1] = XMFLOAT3(0, 0, 1);
	mesh.vertex_normals[2] = XMFLOAT3(0, 0, 1);

	mesh.vertex_tangents.resize(3);
	mesh.vertex_tangents[0] = XMFLOAT4(-1, 0, 0, 1);
	mesh.vertex_tangents[1] = XMFLOAT4(0, 1, 0, 1);
	mesh.vertex_tangents[2] = XMFLOAT4(1, 0, 0, 1);

	mesh.vertex_uvset_0.resize(3);
	mesh.vertex_uvset_0[0] = XMFLOAT2(0, 0);
	mesh.vertex_uvset_0[1] = XMFLOAT2(0, 0);
	mesh.vertex_uvset_0[2] = XMFLOAT2(0, 0);

	mesh.CreateRenderData();


	auto entity = scene.Entity_CreateLight("light");
	scene.Component_Attach(entity, rootEntity);
	auto& light = scene.lights[0];

	light.type = wi::scene::LightComponent::LightType::POINT;
	light.color = XMFLOAT3(0.5, 0.5, 0.5);
	light.intensity = 1000;
	light.range = std::numeric_limits<float>::max();
	light.SetCastShadow(true);

	scene.Update(0);

from wickedengine.

I didn't see that you created an ObjectComponent. Without an ObjectComponent, a mesh can't be rendered. After an object is created, you also need to tell it which mesh to use:

		ObjectComponent& object = objects.Create(entity);

		MeshComponent& mesh = meshes.Create(entity);

		// object references the mesh entity (there can be multiple objects referencing one mesh):
		object.meshID = entity;

(Example from the CreateCube() function)

from wickedengine.

Note that you don't need to create the mesh on the same entity as the object, but for simplicity this is also possible. Usually, there would be a separate mesh, and multiple objects referencing it (like lots of tree instances using one tree mesh)

from wickedengine.

After merged the scene I created with sponza, I can see the light, but still found mesh nowhere, even with object component added.

Update: after added transform component to the mesh entity it works.

The code is like below

	auto materialEntity = scene.Entity_CreateMaterial("default");
	auto& material = *scene.materials.GetComponent(materialEntity);
	material.baseColor = XMFLOAT4(1.0, 1.0, 1.0, 1);
	material.SetDoubleSided(true);
	material.CreateRenderData();

	auto meshEnity = scene.Entity_CreateMesh("test");
	auto& object = scene.objects.Create(meshEnity);
	auto& mesh = *scene.meshes.GetComponent(meshEnity);
	object.meshID = meshEnity;
	scene.Component_Attach(materialEntity, meshEnity);

	mesh.subsets.push_back(wi::scene::MeshComponent::MeshSubset());

	mesh.subsets.back().materialID = materialEntity;

	uint32_t vertexOffset = (uint32_t)mesh.vertex_positions.size();

	mesh.indices.resize(3);

	mesh.subsets.back().indexOffset = (uint32_t)0;
	mesh.subsets.back().indexCount = (uint32_t)3;

	mesh.indices[0] = 0;
	mesh.indices[1] = 2;
	mesh.indices[2] = 1;

	mesh.vertex_positions.resize(3);
	mesh.vertex_positions[0] = XMFLOAT3(0, 0, 2);
	mesh.vertex_positions[1] = XMFLOAT3(1, 0, 2);
	mesh.vertex_positions[2] = XMFLOAT3(0, 1, 2);

	mesh.vertex_normals.resize(3);
	mesh.vertex_normals[0] = XMFLOAT3(0, 0, 1);
	mesh.vertex_normals[1] = XMFLOAT3(0, 0, 1);
	mesh.vertex_normals[2] = XMFLOAT3(0, 0, 1);

	mesh.vertex_tangents.resize(3);
	mesh.vertex_tangents[0] = XMFLOAT4(-1, 0, 0, 1);
	mesh.vertex_tangents[1] = XMFLOAT4(0, 1, 0, 1);
	mesh.vertex_tangents[2] = XMFLOAT4(1, 0, 0, 1);

	mesh.vertex_uvset_0.resize(3);
	mesh.vertex_uvset_0[0] = XMFLOAT2(0, 0);
	mesh.vertex_uvset_0[1] = XMFLOAT2(0, 0);
	mesh.vertex_uvset_0[2] = XMFLOAT2(0, 0);

	mesh.CreateRenderData();

	scene.transforms.Create(meshEnity);
	auto meshtrans = scene.transforms.GetComponent(meshEnity);
	meshtrans->Translate(XMFLOAT3(0, 0, 0));
	meshtrans->Scale(XMFLOAT3(1, 1, 1));


	auto lightEntity = scene.Entity_CreateLight("light");
	auto& light = scene.lights[0];

	light.type = wi::scene::LightComponent::LightType::POINT;
	light.color = XMFLOAT3(0.5, 0.5, 0.5);
	light.intensity = 1000;
	light.range = std::numeric_limits<float>::max();
	light.SetCastShadow(true);

	auto& transform = *scene.transforms.GetComponent(lightEntity);
	transform.Translate(XMFLOAT3(0.0, 0.0, 1.0));
	transform.UpdateTransform();

	scene.Update(0);

it gives a result like this.

from wickedengine.

Note that you don't need to create the mesh on the same entity as the object, but for simplicity this is also possible. Usually, there would be a separate mesh, and multiple objects referencing it (like lots of tree instances using one tree mesh)

New question:
How many meshes an object can have? It seems be only one? What if I have an object with meshes organized in an tree like structure? Can I have objects as children of an object?

from wickedengine.

An object can have one mesh only. You can create a parenting hierarchy of objects with Scene::Component_Attach function.

from wickedengine.

That works.

Another question, what's the unit of FOV of a camera?

from wickedengine.

Radians in the code but degrees in the editor slider. Every angle is like that, the code expects radians in every math function, but users can input degrees. Can convert with wi::math::DegreesToRadians finction.

from wickedengine.

Radians in the code but degrees in the editor slider. Every angle is like that, the code expects radians in every math function, but users can input degrees. Can convert with wi::math::DegreesToRadians finction.

Ok. And is there any way that I can know the width of view of a camera? I want to calculate the target point of camera, so that when I roll the camera it rolls around the target point (3rd person view) not roll itself (1st person view)

from wickedengine.

For width of the camera, there is CameraComponent::width. For third person camera, there is an example third person controller script in Content/scripts/character_controller. There is also an orbiting camera mode in Editor that you can take as example.

from wickedengine.

much thanks

from wickedengine.

Is there any simple clear exemple about the terrain generator? I walk acrossed the code in terrainwindow.cpp but still got lost in it.

I can so far have only this, seems a lot of parameter there but don't know how to fill them

from wickedengine.

This is what you should see by default if you run the editor from the correct location and generate a terrain:

The correct location is the Editor folder in the root directory of the engine. Inside the Editor folder, there is a terrain folder that contains sample terrain configuration files and assets.

from wickedengine.

Thank you, it works, now the next questions:

1. What are terrain are, are they meshes or just rendered directly by the shader? (it seems they are made of chunks and each chunk holds its mesh?)
2. For a coordinate (x,z) is it possible to know the y? (y-up)
3. How to get the intersection (x, y, z) of terrain with a ray? or maybe with a (x, y) of the windows?

from wickedengine.

The terrain ground chunks are regular meshes. You can use the regular intersection functions with them, for example with a ray, and those will return a position among other things. The Editor has example of picking with the mouse, in short you generate a pick ray from the mouse position and give it to scene.Intersects()

XMFLOAT4 currentMouse = wi::input::GetPointer();
Ray pickRay = wi::renderer::GetPickRay((long)currentMouse.x, (long)currentMouse.y, *this, camera); // "*this" is a RenderPath/canvas that contains the width/height
RayIntersectionResult result = scene.Intersects(pickRay);
float height = result.position.y;

(The Editor is using wi::scene::Pick() function instead of Intersects, just because it's from older code, but it does the same)

from wickedengine.

The terrain ground chunks are regular meshes. You can use the regular intersection functions with them, for example with a ray, and those will return a position among other things. The Editor has example of picking with the mouse, in short you generate a pick ray from the mouse position and give it to scene.Intersects()

XMFLOAT4 currentMouse = wi::input::GetPointer();
Ray pickRay = wi::renderer::GetPickRay((long)currentMouse.x, (long)currentMouse.y, *this, camera); // "*this" is a RenderPath/canvas that contains the width/height
RayIntersectionResult result = scene.Intersects(pickRay);
float height = result.position.y;

(The Editor is using wi::scene::Pick() function instead of Intersects, just because it's from older code, but it does the same)

Seems I don't get it right, with a terrain in the scene,

	auto ca = wi::scene::GetCamera();
	wi::primitive::Ray ray1(XMFLOAT3(ca.At.x, FLT_MAX, ca.At.z), XMFLOAT3(0.0, -1.0, 0.0));
	wi::primitive::Ray ray2(XMFLOAT3(ca.At.x, -FLT_MAX, ca.At.z), XMFLOAT3(0.0, 1.0, 0.0));
	auto result1 = wi::scene::GetScene().Intersects(ray1);
	auto result2 = wi::scene::GetScene().Intersects(ray2);

With code above, I get always (0, 0, 0) for both results
If I replace FLT_MAX with 300, I got some result but it doesn't change if I move the cam.

from wickedengine.

ca.At is the camera lookAt direction, I don't think you meant to use that for ray origin? ca.Eye is the position.

from wickedengine.

Also, I wouldn't use FLT_MAX, the ray tracing will not work beyond that distance. I notice you trace one from above and one from below, but if the terrain is at 0 height, then neither trace will work for it.

from wickedengine.

What I am looking for is the y height of terrain at the location of my camera, so that I am able to avoid placing the camera under the terrain

from wickedengine.

Get it working, so I set the camera 10 m above the terrain.

But I noticed two problems:

1. what's the white thing?
2. I see clearly the terrain far away is not visible, is it caused by the its LOD setting?

from wickedengine.

the white thing seems to be grass but without texture. The terrain generation distance can be modified with wi::terrain::Terrain::generation

from wickedengine.