When will roblox add grass interaction!!!
Or water interaction physics!!!
Come on man, like this resource is good and all I love it but… Its just not usable for games.
I hate how unmodern roblox is…
Oh also, you could probably modify the octree to increase performance and use .presimulation to be more accurate to how the engine grass works…
this is very usable for games, as I am using it for mine and it works perfectly fine. Its easily modifiable as well. This guy made it very well for others to use
Well yes it’s usable, but at a very high performance cost.
Especially with the regular amount of terrain:area you need to cover. That will be way too costly.
I’m just saying, it’s not performant but it works for sure
its not that crazy of a performance problem, its optimized pretty good for what it does. It having bad performance depends on the scripter. The tool itself performs very well. Ive been testing and modifying all day today. The tool is very good.
Little question though.
How are draw calls?
Not entirely sure how Roblox internally handles skinned meshes but I recall that animated meshes for instance likely cannot be instanced due to how graphics rendering works which would result in a lot of draw calls.
What’s the status on this?
In the provided example I see a patch / lawn of grass but I’d like to see this in action on a huge mountain for instance, or a dense forest.
What are you talking about?
At 180fps. With over 5 percent usage for 1600 skinned meshes. This tanks performance by a good 10fps for just that.
Mind you that’s out of box. If you modify it and make it larger or add more. It will only get worse.
Also, my specs are a i5 13600k and a 7800xt.
So low end devices will suffer even worse!
It’s well optimized for what it is. But it isn’t performant.
on a side note, could we get your scene rbxl place? it looks really ncie
just wanted to make a more performant alternative to similar modules on the forum, in most cases it’s not suitable for grass at a large scale
static parts perform better than static skinned meshes I assume because of batching, but dynamic skinned meshes using Transform outperformed dynamic parts with BulkMoveTo even with the increased draw calls (also not sure how roblox handles behind the scenes)
here’s a couple larger scale showcases
Can you please make a tutorial on how to create thousands of blades and bones skinned meshes in Blender automatically? The most I can make by hand is 20 blades/skinned mesh xd
Alright, let me fix what I said, it is performant for the capabilities of roblox. Highly unlikely to make anything large scale that is what you think is perfomant, not because roblox is bad but it was never originally engineered or intended to handle it in the first place. Hence there is no shader support on roblox and never has been, which is what most performant grass algorithms use.
That’s more like it!
I apologize for my confusion and or ignorance. I’m also glad we came to an agreement.
I do hope that Roblox adds grass physics interaction natively…
No, what I meant was to be able to make the SkinnedGrass bones have the “shockwave” effect im looking for. Kind of like the clip 3:56 in this video https://www.youtube.com/watch?v=a6d83wG82Mg but for grass instead of cloth.
Yeah i know what you mean, the script is very easily configurable/modifyable to make this
Haha youre not ignorant man, its just a discussion on programming! Glad we could agree as well! I hope so too.
ummm
nvm found the problem
Hi,
You have given some great performance information. Do you think this can be used on small scale areas for grass and not take performance hits? Also if it has FOV , then why would grass not in the FOV or streaming area cause performance issues?
Thanks
OH MY GOD ITS THE WII SPORTS GRASS!!! /srs