oFusion + ShaderFX + Collada?

Hello,

We're currently using oFusion and ShaderFX together, and as I think I've said here before, it works really well. We're now considering exporting from this pipeline to the Collada format. I've used a Collada once, but that was exporting to the C4 engine. I have no experience using Collada with Ogre or oFusion.

Can anyone tell me if there advantages or drawbacks to exporting to Collada from the combined oFusion/ShaderFX pipeline?

Any information would be greatly appreciated!

the authors of shaderFX said not so long ago that the next version would support colladas shader format directly.

http://lumonixsoftware.com/yabb/YaBB.pl ... 07330018/3

I doubt you will see collada support in ofusion since the point of ofusion is to support Ogres native formats through the sceneloader.

The main issue is that you might obtain a different media from what you created in max, as the collada exporter will handle the information differently (and several of the specific features wont be supported by collada, as scene data, materials, mesh specific data, lods, texture alias)

How do you intend to produce the materials and shaders? in this collada based pipeline that you want to use

Thanks for the feedback.

Lioric, I should clarify: I'm not saying I want to use Collada. A programmer on the team suggested we look into Collada, and asked my opinion. I have no idea whether it is a good idea or not, so I'm seeking more informed opinions here.

It took me a while to get to where the oFusion/ShaderFX combination works for me, and I've come to like that pipeline. I'm honestly not all that eager to throw a new element into the mix, but if Collada offers some advantage, I'm open to it.

It sounds like you're saying that Collada would make it hard to create materials and shaders - is that correct? If so, can you elaborate?

If - as seems possible - Collada will bring complications to this pipeline, it would really help to know what they are so they can pass them on in a form that is more convincing than, "I've got a bad feeling about this."

Thanks again, I always appreciate the quick turnaround here.

If you let me know the why of your programmer to suggest you to look into Collada, then i can provide my opinion on if it might be situable for your project

The specific details on your project are not know, in this case i can only provide a broad opinion about collada

The Collada file formats, is just that, a file format, that its target is on being an application exchange format, that is, a way to generalize the method on which data is tranfered between applications, but on the same layer (as from a modeling application to another, or to a different editor)

But this file format was not created to be the solution for all cases or to be the ultimate file format, is just another tool that is suited (and was created for) a very specific case, to exchange data between applications, off-line, where the performance is not important but the integrity of the information is (in the sense that the exact data that was exported is loaded in a different, same layer application without notable differences)

This file format is not suited to be the final, user application (your project) specific file format. In your application is not important if the scenes can be opened in the X modeler, or loaded in the Y paint editor, but you only are interested in loading them in your application (and in your editor tools) with the specific data you need, the full support of your application features and where the loading speed is important

The Collada file format is created to contain all of the scene description in a single xml file, including meshes, skeletons (if suported at all), animations of any kind, probably it can be customized to create a file per resource, but in the general case if you want to update a mesh, you need to export the complete scene again to replace the scene file that contains this mesh

If you are creating a modeler or a different content creation tool, where the produced data must be used with different tools to be of any use, then you might consider Collada as it is supported by many external applications (all of them with different levels of implementation, collada is not the 'real' ultimate solution, not even in this area)

If you are creating a specific application, where the data is used only by your application (and by your own custom tools or editors), then a specialized file format must be used

Hi Lioric,

Thanks very much for the detailed reply!

In this case, I think the programmer thought Collada might make somehow make object placement within a game scene easier. That, and I think there was some hope of being able to open up our app to a variety of artist contributions, and also be able to re-use our current assets in other engines.

However, a little bit of back-and-forth among our team revealed that there was no real will to throw this big of a wrench in our project. Some of what you said in your previous post helped us come to that conclusion. So again, thanks a ton for your feedback.

I think the biggest plus with collada is that you can load and save it in most 3d apps. Plus a scene can be stored in a single file.

At the end of the day though, I think Ofusion is better since if supports Ogres features directly. And if you have pro you can always use the new zip archive feature. Would be nice if Ogre could save a file with all meshes and heirarchy intact though. Thats about the only thing I miss coming when I used to use Blitz3D.

The oSceneSerializer libraries (Binary and Xml versions) will support producing a single self-contained file (as currently supported when exporting from max, the complete scene in a single file) in the next update

Sounds awesome Lioric that would really make things much more friendly for game dev