Octane vs Brigade

[mbetke]

For me it would be a dream to have a tool like Lumion3D which I use often for cheap animations combined with the visual fidelity of Brigade. Due the big success of real-time arch-viz tool like Lumion3D it would be the next step to have a real-time renderer like Brigade for fast biased animation rendering.
Most of the clients which I cater don't care about the visuals too much and are very pelased with Lumion3D graphical quality.

If I look at todays game-engines they have a lot of post-effects to make stuff looking good. With PT and proper reflections it goes even better.

[darkline]

I don't think anyone was suggesting it could magically be done, just that it would be a very good idea.

Perhaps it goes against what Octane has set out to do, but IMO you'll win many many new customers with a faster/realtime rendering option for when it's needed. I don't agree that everyone wants perfect accuracy at the expense of speed all the time. Or that Brigade would not be good enough for production renders. You want the options because time is money.

An example of this is Element 3D for After Effects. It has tons of limitations (too many to mention), but because it delivers nice AO 3D objects with metallic surfaces, DOF, motion blur, and fog in about a second per frame, with no noise - it's become an essential plugin for AE. 3DS Max currently doesn't seem to have anything to produce similar results in the same time frame. Element 3D is super feature limited yes, it's using a ton of tricks true, you can't even have light sources with shadows (yet)... but all that hasn't stopped it from becoming widely adopted and used. It's the No1 plugin for AE.

http://www.youtube.com/watch?v=Xe0aADhVjE4

Brigade looks far better and is capable of a lot more than element 3D can be. Conversely, how long would it take for Octane to render 1 frame of an AO cityscape with DOF and fog like in this video?

If Octane had options to either use brute force or something quick but 'faked', it would be lapped up by many. If Brigade is the ticket to that happening then I think it would be a mistake not to try and implement it into a production renderer. Games designers using Brigade can see exactly how the final game is going to look from within MAX if it's a plugin renderer also.

[abstrax]

I really wonder where this idea comes from, that Brigade is much faster than Octane. It's not. The direct lighting kernel of Octane is fairly close to what Brigade does and has similar speeds. And AE Element 3D is not Brigade?!? It all just doesn't make any sense to me.

The only major difference between Octane and Brigade is that there has been some major work in optimizing the imaging pipeline in Brigade to keep latencies low. But there is not much point for that in a normal production rendering environment. If a clean rendering is at least many seconds or minutes, it doesn't matter if the first frame takes 20ms or 200ms. This latency only matters in a gaming environment, because it's the difference between 50 fps and 5 fps.

On the other hand, Octane has a fairly complex node and material system and scales very well in multi-GPU setups, which is not the case for Brigade. And again if you plug in the material system of Octane into Brigade I would expect it's performance to drop considerably. But of course, this would 't make sense either, since Brigade is for realtime gaming graphics. It's two different things.

Cheers,
Marcus

[gabrielefx]

regarding video game streaming to be playable on tablets, smartphones, etc., will Brigade support this feature?

Will your Otoy cloud service implement this technology?. Streaming is the future of extreme photorealistic videogames.

[Andygraph3D]

I really wonder where this idea comes from, that Brigade is much faster than Octane. It's not. The direct lighting kernel of Octane is fairly close to what Brigade does and has similar speeds.
Because Otoy don't use octane as game engine if are not different there ?

But there is not much point for that in a normal production rendering environment.
The render quality (renderman) of first Toy Story 1994/95 i much better of Brigate now .. ?

If a clean rendering is at least many seconds or minutes, it doesn't matter if the first frame takes 20ms or 200ms.
if I want create a short movie for web from seconds to minutes make big different, very big.

On the other hand, Octane has a fairly complex node and material system and scales very well in multi-GPU setups, which is not the case for Brigade. And again ..
I hybrid version octane+brigate i mind too is not simple but maybe a new brigate friendly version for octane users ?

[darkline]

Ok Marcus I see your point now, you're saying one (of many things) is that having complex materials would make brigade a similar speed to octane. Fair enough.

Brigade aside I guess what I'm saying is that I'd welcome some sort of 'cheating' kernel in octane that maybe faked things a bit like element 3d does. I think it would be sought after by many here. AO is cheating anyway so is there a way you could implement some similar kernel that would produce noiseless AO renders quickly?

[marchermitte]

What could be interesting for archviz , have a brigade license for real time archviz presentation with specific animations capacities (walk through camera, flyover, sun anim, etc.) The client would sit an visit his project in real time, that would be pretty impressive. Couple that with Nvidia 3D glasses and... wow!

[Goldorak]

Without getting into specifics (since things may change over time), an Octane+Brigade bundle is a possibility at some point down the line.

Right now, there is still much work to be done to make something like this useful to end users, but the idea would be for Brigade to complement Octane for use in real time games or scene navigation, where final render quality is of secondary importance to refresh rate/interactivity that Brigade is tuned for. Many of the ideas in this thread are good. You may see some of them implemented.