vfx testing testing

Here is a compilation of some fluid fx I've been learning.
Fluids can be anything from smoke to fire to water. Figuring some of this stuff out is such a challenge but also so much fun! Enjoy!

 
{tapping on the mic) Is this thing on?
Not Listening Dumb And Dumber GIF
 
Sorry, I was planning to answer this thread, and got distracted. Really cool showcase of fluid effects.

I used to use fumeFx in Max for this stuff, and then I used a system in blender for a while, and now I'm using a few new plugins in UE5. I really enjoy working with fluid simulations, but the reason I don't do it more often is because it always takes forever to run the initial simulation. I remember waiting literally days for Max to build out a 10 second water event in 2010. The new stuff doesn't look as good, but it's a lot more fun to work with because it's literally hundreds of times faster.

Here's some of my old blender work.



I'm still working out a fluid pipeline in UE5, it's a lot more complicated this time because I have to simulate a lot of different real world scenarios, make them stable and repeatable, and have them run under 3 seconds a frame with simulation (non baked). I do run baked sims for certain things, like vortex portals.

Here's a quick video I made showing one of the new systems, for another thread a while back where someone was asking about reverse engineering things.


I'm glad to see someone else on here doing this stuff, great work man!

As far as the comment silence, high intelligence work has kind of an inverse response curve. You're on the right track. This kind of work pays off big over the long term, but doesn't get the quick kneejerk responses that you get from joke videos. I totally get the frustration you and buddy feel, spending weeks of hard work to do something amazing and then watching a video of a drunk guy dropping his iphone into an aquarium get 6 million views and 40k comments.

 
Last edited:
Thanks, Nate!
Your stuff looks pretty cool. I can see we both like to experiment to see all the things fluids and dynamic sims can do.

I'm about 10 pages from finishing the first draft of a screenplay that I think I can shoot. The story features CGI characters comp'd onto live action plates. Right now most of my FX tests are to help me make sure I can actually pull off some of the stuff I'm writing.. It's not the best way to work but it keeps me from getting carried away with myself. Sure, I can write a scene that shows aliens leveling a city a-la War of the worlds but do I have the ability to pull it off realistically? As a CGI generalist who works a 9 to 5 job too, there's only so much I can know about EVERYTHING. That's why I run tests as I write.
 
Thanks, Nate!
Your stuff looks pretty cool. I can see we both like to experiment to see all the things fluids and dynamic sims can do.

I'm about 10 pages from finishing the first draft of a screenplay that I think I can shoot. The story features CGI characters comp'd onto live action plates. Right now most of my FX tests are to help me make sure I can actually pull off some of the stuff I'm writing.. It's not the best way to work but it keeps me from getting carried away with myself. Sure, I can write a scene that shows aliens leveling a city a-la War of the worlds but do I have the ability to pull it off realistically? As a CGI generalist who works a 9 to 5 job too, there's only so much I can know about EVERYTHING. That's why I run tests as I write.

I disagree, I think it is the best way to work. This is exactly how I think, lol. What good is it to write something I can't show on screen? So I test. Usually, the test doesn't turn out anything useful the first time, but with each test I pick up skills that make that scene more possible in the future. After just over a year of this process full time, results have improved dramatically.

Here's what the Save Point pipeline looked like 1 year ago. You might actually find this video amusing, I used audio from an actual phone call, made sometime in the 90s. I think my favorite line is "I understand that deer love animals"


This was the first experiment simply prioritizing speed way over quality. Behold the 23 minute animated film with 100 lifetime views, lol.


Here's what it looked like after a year of these fx tests. You can see 3 different sims working together here at under 2 seconds per frame render.

 
Last edited:
Forgot to mention, there's a new movie that came out recently, I think it's called "Chip and Dale, Rescue Rangers" where they did the film with animated characters on live action backgrounds, and you know, it worked out pretty well. The movie wasn't that bad either, considering the expectations I have going into a streaming only movie about a cartoon I watched as a kid. Roger Rabbit is the classic example of course. You may want to consider running an art pass over your 3d characters with Dall I 2 when it becomes available. It's how I'm planning to kind of pave over the uncanny valley when we go into production on the main product. You can see in "Arcane" where they've blended styles, and managed to diminish the off-putting nature of cg actors to a large extent. Just a thought.
 
I've actually engineered the story to match the level or realism I'm capable of producing. Photo-realistic is doable, but With my resources, I would never try to simulate humans. Claymation is what I will be simulating. I love the look of stop motion animation with clay puppets but I have no patience for doing that sort of work. I've tried, and after an hours worth of animating, I was ready to jump out the window.. Using mocap and some clever editing of the animation curves, you can get pretty close to the look of hand animated stop motion. One interesting difference between stop motion photography and what I'm planning is motion blur. Because of the nature of the story I wrote, even though the characters are stop motion puppets, they are moving in the real world which means a camera that is filming them as they perform would capture motion blur. In traditional stop motion one of the noticeable traits is the lack of motion blur, so even though the animation will have a deliberate herky jerky animated look to it, it will be somewhat smoothed over by the inclusion of motion blur.
 
Sounds cool! I've got some PBR clay materials, but they are ue format, so I doubt there's an easy way for you to convert them. If you ever want to try out rendering un UE5, I'll send them to you. Have you ever tried out octane render? It's really good for PBR stuff. I'm pretty sure they have octane for Maya. Way faster than Vray, etc.
 
Thanks Nate!
I use V-ray and I've grown to love it. For a while, I was using Arnold and I thought I loved it until I switched to V-ray. After being away from Arnold for that time them going back to it, I realized I didn't love it as much as I thought I did. LOL! As far as PBRs go, I don't see any noticeable difference in the results between V-ray and Arnold. V-ray is just more user friendly. Speed-wise, I haven't done any back to back tests. Never tried Octane. I'll have to read up on it.

I have a clay shader that I bought on-line that looks very good though I may use a V-ray standard material and just dial it in to look just as good..
 
Back
Top