Hey! anyone have any cool images created with Stoke? Need it for our PR!!
email me chrisbond ]at[ thinkboxsoftware.com
thanks!
cb
Hey! anyone have any cool images created with Stoke? Need it for our PR!!
email me chrisbond ]at[ thinkboxsoftware.com
thanks!
cb
This is a quick one I did as a learning experience, I’ll let you decide if you want to use it or not. The main thing I think is cool about it is that it’s a Naiad simulation (~300,000 particles at the last frame) that I used as a velocity field for Stoke to generate 4,500,000 particles (last frame). One thing I immediately feel Stoke is good for is if I don’t have enough particles from my liquid simulations for foam/splash I can just make more with Stoke (and have some level of control as well rather than just multiplying the amount of particles). I’m working on generating some more examples that were made with purpose and direction now that I feel I got a hang of it.
Hi Tobbe,
Could you render the last frame with the original Naiad particles only, and post it and the Stoke particles’ last frame so we can show them side-by-side as a comparison?
Absolutely, exact same render settings but using the Naiad PRTs. I used a thickened version of the spline and converted the normals to velocities so that any particles that left the Naiad velocities would get pushed back in towards them. That’s where I used the single frame thing that we discussed in the other topic. Otherwise I’d get a lot of particles that would leave the Naiad velocities and have no velocity which didn’t look very nice. I tried deleting them by velocity but this approach looked nicer.
www.tobbeo.net/Naiad-Stoke_before.png
www.tobbeo.net/Naiad-Stoke_after.png
I just left it plain like this so you can decide the formatting you want.
nice! I wish that we had one with like 10x or 100x more particles so there weren’t any individual points left.
bobo, is that something you could do with the file if we could get access to that last frame?
cb
Nope, I would need to simulate the whole thing from the beginning, so I would need all frames.
I just ran the same Stoke sim with 15,000,000 particles and am about to render it out. I’ll post it here as soon as it’s done. The stoke sim took just shy of 30 minutes, which is pretty crazy impressive for that many particles. I could also render out the Naiad particles so you’d have a) Naiad particles, b) 4.5m Stoke particles, c)15m Stoke particles.
Don’t forget to reduce the Final Pass Density between 3 to 4 times to produce the same look, just smoother.
Does this work for you or do you want them as separate videos?
tobbeo.net/Stoke-NaiadSpline_Combined.mp4
tobbeo.net/Stoke-NaiadSpline_CombinedB.mp4
Edit: I added one with lower density.
Definately something I’m missing between Stoke and SR’s Repopulate… With Stoke we can add extra perturbation on the vectors to produce higher frequencies and reduce the “blurry” look that Repopulation provides. Of course, Repopulate is HI, so that’s nice, too. Maybe a two-stage setup?
This is probably meant for a separate topic but I feel that since they put Repopulate in KrakatoaMy and KrakatoaSR then it should be in KrakatoaMx as well. It’s what’s most fair to consumers. That’s my 2 cents.
repopulate is intended to come to MX. no shifty business here, just resource allocation.
cb
Ah, I got a different impression from Bobo when I brought it up in a topic over in the Krakatoa channel. That’'s great news, thanks.
yeah, we want krak to be unified across platforms, no question.
cb
Hi Tobbe,
I have a few questions about your demo video:
*How long did it take to simulate the original particles in Naiad?
*Do you know how long would it take to simulate the same with 15 million?
*How did you do the spline deformation? PRT Loader + PathDeform? Or Ember?
The full simulation is 300 frames, I just don’t show the first 100 because they are boring.
a) 13 minutes – end frame is actually 170,000 particles not 300,000 (I’m updating the video).
b) I ran a test and while every frame gets longer and longer, a rough estimate would be around ~10 hours for 15m particles and that’s an extremely generous estimate as each frame that’s generating 50,000 particles is taking ~2 minutes in the beginning. The longer the sim goes, the slower each frame would get.
c) I am not using deformation, I’m using a spline, converted to PRTs with PRT Hair and exported to Naiad, emitting from a sphere at the base of the spline + some tricks. I wanted the turns to be simulated by the flip solver to get the nice motion.
Hope that helps.
Wicked!
Awesome, I have uploaded it to our YouTube channel as Unlisted.
Here is the link:
youtube.com/watch?v=Ed6rrI_o … detailpage
This is a great illustration of the time-saving potential of Stoke, thank you very much for the help!
Cool!
Here are a couple of Wallpapers I quickly put together:
tobbeo.net/Stoke_wallpaper1.png
tobbeo.net/Stoke_wallpaper1b.png
tobbeo.net/Stoke_wallpaper2.png
Don’t know if you’ll find them useful but anyways!
Tobbe