We all are just aware this is a mere BETA version. We accept all common criticism as we post our work here and are openly available to all your thoughts on the work that has been done here. We do however ask to keep it with a nice tone and make adjustment ideas for our engine without the comparison of other third party tools.

All bugs have been worked out and we are preparing additional screens with optimized features to display the advantages our shader engine has compared to others. Screens and beta gameplay will be available for the community again shortly so please keep an eye out

Thanks.

honestley I don't see how this is similar to shade-c at all. In fact, you can use both at the same time. This is an engine extension that improves the basic functionality of the A7 engine. You can add lights and objects in a more intuitive way, and there are no limits (like 8 lights only). Also, I'm not aware of anyplace where you can get code like this on the forums. If this already existed, I wouldn't waste my time.

-Mike-

-comment removed-

Just for my interest: what features do you use that doesn't exist on a Geforce6?

"Just for my interest: what features do you use that doesn't exist on a Geforce6?"

vertex texture fetch (vtf), CUDA ray-tracing, stencil buffer extensions. But, even with these features disabled, it will run slowly just because of the power of the 6/7000 cards.

oh, and mrts of course. If your card doesn't support mrts, then you can have deferred shading. All new cards from up to a few years ago support this.

-Mike-

Thx for your respond, although Geforce6 supports vertex textures and mrts. I don't know what you mean with 'stencil buffer extensions', can you explain it please?

If you use CUDA doesn't it mean that it won't run on AMD and Intel cards atall? Doesn't sound good for me.

CUDA runs fine on AMD and Intel cards (I'm using an Intel chipset). The stencil buffer extension is just an assembley routine I wrote for the GPU directley and interfaces with the normal directX stencil buffer (it allows me to cast very crude shadows for GI). I didn't know that Geforce 6 cards supported mrts, but I guess its obvious that they support vtf because they can run shader model 3.0. Also, how much memory does your card have, this might be the problem.

-Mike-

So why shouldn't you be able to publish your work with shade-c? And what does it have to do with nvidia, i'm not even using any shaders from nvidia? Of course i am using all those nifty effects and techniques nvidia is also using in their shaders (as does any other game nowadays). I have no idea what you are talking about here (maybe i'm just dumb, who knows).

Also, just to inform you, if i want to use nvidias shaders, i am able to do so. From their website:

http://developer.download.nvidia.com/SDK/9.5/Samples/effects.html

To keep this short: You shouldn't make false accusations just to "sell" your product. I will take this personally if you ever do it again.


And to stop shade-c <-> iridium arguments: As you already said, you can't compare those two. Shade-C uses Lite-C only and is open source, Iridium is a closed source .dll (and might have it's own renderer, i couldn't really tell by the demo).

Good luck with optimizing Iridium, as far as i can tell, you have a long road to go.

"Of course i am using all those nifty effects and techniques nvidia is also using in their shaders "

I was flat out wrong. None of the code in your project is non-publishable. I was thinking about a LONG time ago when you first put in the shaders. Since their free, there are no problems PLUS, I noticed that in the recent versions, all of the code has been re-written so it really is yours.

apologies, Mike

(also, your one of the guys who inspired me, sorry for being a jerk)

I had no idea CUDA runs on AMD cards as well -- I have an nVidia card, so it wasn't a personal issue, but more that I don't want to limit anything I make to computers with nVidia chips. As I'm very interested in game-physics I was considering waiting for OpenCL before delving into GPGPU programming.

That's very cool if it's indeed true, however Wikipedia lists it as supported only on nVidia GPUs.

If you don't mind me asking, what do you use CUDA for?

Jibb