News

following the last post about what is going on around Ogre, here is another update.

Ogre 1.12.6 point release

The 1.12.6 point release kept its focus on integration. Notably, it ships the Qt OgreBites implementation, that was discussed in the previous post. Also there are the following notable changes:

• The OSX & iOS support was vastly improved. If you had any issues with using Ogre in these enviroments, please try again with Ogre 1.12.6. The improvements include scaling the input events along the window content scaling and corrected library resolution from inside an .app bundle. I also started backporting the Metal RenderSystem to the 1.x series. See the metal-wip branch. Help is welcome.
• Also Ogre 1.12 is now available in Debian sid & unstable. Notably, this also covers the python bindings and imgui. So prototyping a 3D application is only a apt install python3-ogre-1.12 away. Also this is a major step forward from the previous Ogre 1.9 package and allows 3rd party apps like ROS relying on it. These packges are also availble in the latest Ubuntu 20.04 release, which is based off Debian unstable.
• Next, the MSVC SDK was refined: it now properly packages and compiles the C# ibindings. You no longer have to compile any Ogre C# sources yourself. As for C++, the pdb files are finally included, which allows you using the SDK for debugging (note that the build type still is RelWithDebInfo and not Debug).
• The refined “Fixed Pipeline Enabled” RenderSystem option allows you to bring the legacy GL & D3D9 RenderSystems into a programmable pipeline only mode. Setting it to false is the first step, if you consider porting to GL3+ & D3D11 as it will enable the RTSS shader generation, while keeping everything else the same. It also allows you toggling forth and back between shader based and fixed pipeline to compare rendering, instead of having to pull the plug once and for all.
• For the full list of changes and bugfixes as always consult the github release

OGRECave ecosystem

Probably the biggest strength of Ogre 1.x is the legacy of content creation tools and addons. Here, I resurrected the following tools beneath the OGRECave umbrella:

• The meshmagick CLI .mesh optimization and manipulation tool. It allows you to conveniently change the coordinate system or merge all sub-meshes into one buffer.
• The shiny shader meta-compiler & management library. In the mid-term we want to extend the RTSS by similar features and generally recommend to prefer the RTSS. However, there are some existing projects relying on this library, and this allows them to upgrade to recent Ogre releases.
• The ogre-gpgpu toolkit. This is a collection of many Computer Vision and Augmented Reality related tools built around Ogre. So far, I only brought back the Ogre-CUDA bridge.
  When it is ready, this probably will be moved into the core repository. Therefore, this project can be considered as a staging area.

OGRE scripts offer a way to define materials at an abstraction level similar to D3D Effects and CgFX where you can define alternative techniques, each consisting of one or multiple passes.
Here, each pass defines a render pipeline state by defining blend modes and referencing shaders. The main difference in OGRE is that you cannot write inline shaders in the script file as we support different render systems with different shading languages.

Traditionally, OGRE allows to use fixed function pipeline (FFP) functionality where you do not have to write any shaders, as long as Phong shading and a fixed set of texture operations is enough for your use case.

However, modern Render Systems like D3D11 or GL3 no longer include FFP parts to reflect that modern hardware does not either and is rather based on unified programmable SIMT pipelines.

To abstract form this difference, OGRE therefore offers the Real Time Shader System (RTSS) component, that generates shaders that seamlessly replace the absent FFP functionality. In most cases OGRE is able to produce pixel-perfect results.

However, as the RTSS generates shaders internally, you can customize the rendering in much more detail then was possible with the FFP. Here, you do not have to write your own shaders but can keep the high abstraction that OGRE scripts offer and just use the rtshader_system section to declare the features you want. Still this, gives you a large amount of control how things are rendered.

The most simple thing to do is enabling per pixel lighting (which is default in 1.12 anyway) or make the shading respect the physical energy conservation rule as described here.

However, the RTSS also enables you to create complex custom render pipelines via OGRE scripts as it offers the following features (the emphasized parts require OGRE 1.12.5)

• Hardware skinning
• Instancing
• GBuffers
• Depth texture shadows
• Lighting models
• Triplanar Texturing
• Offset mapping

Below are some examples how this might look like:

The first screenshot shows the instancing sample, where the RTSS extended the vertex shader to read from the instance buffer as well as the fragment shader to apply depth based shadows. If you switch to the PF_DEPTH format for the depth texture, it will automatically use hardware PCF as it does not incur any performance penalty.

The second screenshot shows integrated offset mapping with multiple lights. As this is handled by the RTSS as well, it can be combined with hardware skinning and instancing – all you need is to add a single line in your material. No need to touch any shader code, while being compatible to all supported render systems.

See the respective Samples on how to integrate this in your own projects.

Deprecation of the HLMS backport

If you are familiar with OGRE, you probably also know that there is the High Level Material System Component in OGRE1. Actually this Component is a backport of the respective core element of OGRE-next (2.1+), where it handles shader variations and thus has a similar goal of the RTSS.

However, it got only little love in OGRE1 after the initial backport, so even as of today there is no way to use it form OGRE scripts. Also I am not aware of any users, as there was not a single bug-report regarding the HLMS.
To reflect that the RTSS is in all cases the preferred alternative, the HLMS is therefore deprecated in OGRE1 and will be removed with the next release, if nobody steps up to object.

Everybody is starting into a new year with good resolutions, so you can now take an advantage of modern OpenGL3+ concepts with OGRE .

Shader storage blocks

The first one is “Uniform Buffer Objects” (UBO) and “Shader Storage Buffer Objects” (SSBO) or simply “shared GPU Parameters” in OGRE speak.
The shared GPU parameters have gained a backing HardwareBuffer which is used for communicating with the GPU.
With OpenGL this can be either a UBO or SSBO which is detected from your shader code and automatically bound if possible.
In case of a SSBO it is possible to read-back the data from the GPU, which is triggered by the new GpuSharedParameters::download() method.
This gives you an easy way to retrieve results from the GPU without rendering to a Texture or Vertex Buffer.

Separate Shader Objects and SPIRV

Traditionally OGRE internally uses monolithic programs for GLSL that explicitly glue vertex and fragment shaders together. Notably, this means the GpuProgramParameters are only valid per combination and not per individual shader.

However, from the API perspective Ogre always exposed the DirectX model without explicit grouping – e.g. in material scripts. This approach is commonly referred to as “mix and match”.

Ogre tries hard to hide this difference for you. For instance one can only retrieve the active uniforms from GLSL once it is linked. This happens on the first render call in OpenGL – instead of at material parsing time when OGRE would need it.
Therefore OGRE goes ahead and parses the GLSL source code itself to figure out the available uniforms. Needless to say, this is quite error prone and does not support more advanced constructs e.g. struct uniforms.

Fortunately, OpenGL provides the DirectX like behaviour via “Separate Shader Objects” (SSO) that allow linking individual shaders and bundling them to pipelines later.
Now we finally take advantage of them and at this also uses ARB_program_interface_query for parsing the uniforms in a standard way and cover all corner cases.
Notably, this allows us to reference uniforms by location only – like in the good old assembly days:

vertex_program Ogre/Compositor/StdQuad_Tex2a_vp spirv
{
    source StdQuad_Tex2a_vp.vert.spv
    default_params
    {
        param_indexed_auto 0 worldviewproj_matrix
    }
}

The corresponding GLSL code for that uniform would be

layout(location = 0) uniform mat4 worldViewProj;

You might wonder why you should care; if you look closely the material snippet above is using SPIRV binary shaders, where the uniform names are stripped away and only the locations are available.

Therefore this is necessary for support of pre-compiled SPIRV shaders, which is now complete.

And while we are at it – why are SPIRV shaders cool? Well, you can compile HLSL to SPIRV and then use it with OpenGL 😉
Also, this is a pre-requisite for the Vulkan back-end and this way you can prepare you shader authoring accordingly.

If you would like to learn more about using SPIRV in OGRE, see here.

Currently (in master) you have to manually enable SSO support via the “Separate Shader Objects” RenderSystem option.
Using OpenGL for shader parsing has some side effects; notably you will now get errors when trying to set an unused (and thus optimized away) uniform. Typically this means your shader is broken – but we traditionally keep your code working during a release support period.