As a Roblox developer, it is currently too hard to completely test the performance of projects using the “Future is Bright” studio prototype build @zeuxcg released for us this last RDC.
Testing games in studio, even with no plugins running, always has significantly decreased my FPS(30+ drop).
While I’m not sure if I’m alone in this problem, I’d assume running a game instance within a studio instance is not going to run as smoothly as just a player instance online.
Although the Future is Bright may not come to Roblox, it was released to us to prototype and test projects under the possibility that it will(and to create some neat things out of it). I’d like to not only build cool things with it as I’ve done, but also fully test the performance impact on certain instances.
I’d love to be able to join a game online through studio to mitigate the performance issues of studio, like we used to be able to do, but newer studio builds(and the future is bright build) no longer allows this.
If this would be added to a new future is bright build , or if we could be guided on how to do this manually if possible, it’d be greatly appreciated.
This doesn’t really make sense to me because as far as I know, the build is still a prototype and not optimised. So any performance testing you do will not be relevant anymore once a finalised (optimised/cross-platform) version of that lighting is shipped. I think confining it to a custom build is good for now as to not give people the wrong impression that it is a near-finished and fully supported feature.
I think most of the people using the build are people who know it’s not a near finished feature.
While I kind of understand what you’re saying, that should be our choice. Of course, if it was released, it wouldn’t be the same build.
But at least by trying it out we could understand an idea of the performance impact it would have for our projects. Unless the final release would be entirely different on performance, it’d be much more relevant and convenient than testing performance in a studio instance with significant performance impacts far greater than any deviation from those builds I’d imagine.
It’s not a matter of it being in-progress. The performance will entirely change before this hits production, making all of your test results obsolete. Testing performance in this state won’t be very helpful.
Also, ROBLOX is not going to release features which make existing games slow. They have old hardware and mobile devices to tailor to, so even though it’s not yet, they’ll be putting in a lot of work to optimizing it so it runs on all of those devices. Right now they’re just working on getting the feel of the lighting down though. You don’t have to worry about performance decreases in your game on low-end hardware.
True, however the project I’m working on isn’t your typical roblox project. I most definitely may have to worry. (If shadowmap is automated on better GPUs). If it’s an optional setting sort of thing then I don’t have to worry necessarily, but still would like to know how it runs(for instance with a lot of animations and a mesh and a texture heavy game). Future is bright is the best guess I have right now.
What would you do with this information? Lighting performance is not typically something you can influence easily apart from reducing the number of lights in the scene, if you have an obscene amount of lights. I don’t think map complexity or anything like that matters, maybe the amount of moving parts matters slightly. Even if you can improve performance by changing your map or the amount of lights, or reducing the amount of moving objects, it doesn’t make sense to do this right now because you don’t know how performance / edge cases will change between now and the final build.
For now it would be better to just read through zeuxcg’s posts about the new lighting build to understand what will affect performance and what not conceptually, and then try to avoid anything that would bottleneck performance.
Even if performance is a concern, the demo will not provide any usable results. Whatever conclusions you make now will be wrong when it hits production – you’ll have to make adjustments regardless of what you do. Don’t tackle performance until the new lighting hits production – spend your time working on what you can