i love chris's new tutorials, i think he really cracked the way to communicate the complexity of houdini in an effortless way, no matter how much i know about the software area he talks about, i get him.
Gotta finish some work but I'll be right back... I've been obsessed with the results of this MPM solver that I have to recreate this and understand what it does. Thanks in advance team
Hi this was nice tutorial @entagma loved it !!. Just one question i about how would i go about adding a specularity to the snow in xpu. I looked several resources couldn't find anything useful . This problem also happens when i am trying to render whitewater using xpu . There is simply no way i guess we can calculate gradient i guess in the karma xpu?
I'm by no means an expert in rendering snow, however what worked for me was: Find the right ratio for subsurface scattering (this hides a lot of those small glints), use a really fine, really strong normalmap, and play a lot with the lighting. This setup right here ended up not having a lot of glints because I wanted the snow to appear softer and had a particular lighting in mind, which didn't really emphasize them.
I am also looking for a solution for whitewater & snow in xpu - the old mantra method where we render as a volume with the gradient for normals is the industry standard and does in my opinion give the greatest photoreal results Please Entagma, we know you can solve this 😂🙏
You can use a point velocity or trail sop to calculate an initial velocity from the provided animation. Then set the start frame on the solver to the moment when you want to switch from the animation to the simulation.
I was under the impression that MPM runs on OpenCL and is therefore a GPU solver. Shouldn't the cache size under the solver settings be limited by your VRAM? I doubt that you have 64gigs of VRAM :P
Hey! The cache size parameter works the same on every DOP node and simple controls how many finished frames get saved into RAM for a fast playback in the viewport. Essentially it controls how long the blue bar in the timeline can get. This, I'm fairly certain, is in no way linked to VRAM.
![Lp. Сердце Вселенной #60 РОЖДЕНИЕ ЛОЛОЛОШКИ [Финал] • Майнкрафт](http://i.ytimg.com/vi/YoR0pAV9FVQ/mqdefault.jpg)
i love chris's new tutorials, i think he really cracked the way to communicate the complexity of houdini in an effortless way, no matter how much i know about the software area he talks about, i get him.
Awww thanks!
Cheers, Chris
Best description ever of how MPM differs from FLIP.
Thank you for the excellent tutorial, I really wanted to learn how to create snow, thank you for being so detailed.
Gotta finish some work but I'll be right back... I've been obsessed with the results of this MPM solver that I have to recreate this and understand what it does.
Thanks in advance team
Love it, thanks so much!
Amazing, thanks!
Love the content! btw, if you run a normalization filter on your audio, it’ll sound way better on the ears!
hey can we get rendering settings and stuff around rendering in general?
entagma tutorial without 35 attribute wrangles, is that possible :O
Hi this was nice tutorial @entagma loved it !!. Just one question i about how would i go about adding a specularity to the snow in xpu. I looked several resources couldn't find anything useful . This problem also happens when i am trying to render whitewater using xpu . There is simply no way i guess we can calculate gradient i guess in the karma xpu?
I'm by no means an expert in rendering snow, however what worked for me was: Find the right ratio for subsurface scattering (this hides a lot of those small glints), use a really fine, really strong normalmap, and play a lot with the lighting. This setup right here ended up not having a lot of glints because I wanted the snow to appear softer and had a particular lighting in mind, which didn't really emphasize them.
I am also looking for a solution for whitewater & snow in xpu - the old mantra method where we render as a volume with the gradient for normals is the industry standard and does in my opinion give the greatest photoreal results
Please Entagma, we know you can solve this 😂🙏
I have a query regarding the scenario where an artist from the preceding stage provides an animation of Snowball.
You can use a point velocity or trail sop to calculate an initial velocity from the provided animation. Then set the start frame on the solver to the moment when you want to switch from the animation to the simulation.
I was under the impression that MPM runs on OpenCL and is therefore a GPU solver. Shouldn't the cache size under the solver settings be limited by your VRAM? I doubt that you have 64gigs of VRAM :P
good question, this is the kind of guy to have 3 GPU's so not sure lol
Hey! The cache size parameter works the same on every DOP node and simple controls how many finished frames get saved into RAM for a fast playback in the viewport. Essentially it controls how long the blue bar in the timeline can get. This, I'm fairly certain, is in no way linked to VRAM.
Looks more like asbestos sadly