|
Yeah, that's great and all echi but is it as cool as the Apple logo if it were made of cocaine? Doc Block fucked around with this message at 10:55 on Aug 9, 2015 |
# ¿ Aug 9, 2015 10:31 |
|
|
# ¿ May 22, 2024 14:56 |
|
there's an elasticity coefficient that determines how "bouncy" the particles are when they hit each other. the higher the number, the less energy they lose when they collide. for that simulation it was 0.05, which is pretty low. 0.0 = no elasticity, so if a particle were to hit, say, the ground, it wouldn't bounce at all, whereas a particle with 1.0 elasticity would bounce forever. they're also "bouncy" because of the way the physics system works. it's a discrete step solver, meaning it moves each particle just a little bit according to its velocity and direction, then checks for collisions, deals with those collisions (calculating new velocities and directions, etc.), then moves the particles a little again, and so on. a consequence of this is that particles can wind up overlapping after the move portion of each step, and if that happens they get forced apart until they stop overlapping, which can cause them to then overlap with different particles on the next step, in turn causing a cascade of particles overlapping and then being forced apart, which is part of the reason there's the big *POOF* of particles in that simulation. but i came up with an even cooler one, but which suffers pretty badly from a combination of bouncy particles and the aforementioned overlap cascade: i wasn't really sure what to expect with that one (~45 minutes to simulate & render), and was hoping for something like the effect at the end of each level in Doom where it would just slide down the screen and come apart, but this was pretty
|
# ¿ Aug 9, 2015 12:33 |
|
there are other methods, but like, there's swept volume collision checking. that's still done in discreet steps, but instead it extrudes each particle along its direction of travel by however much it would move in that step, then checks to see if it would collide with any other extruded particles, which has the advantage of letting you know if particles will collide in that step before moving them so you can only move them to the point of collision or whatever instead of having them overlap. but that's a lot of extra math and poo poo for what's just a toy app. the morgan freeman simulation had approximately 250 thousand particles and took 43 minutes to simulate. edit 2: "fun" morgan freeman simulation deets: Simulation time: 2584.749540 seconds Render time (frame only): 1.479263 seconds Render time: 11.084245 seconds Save to file time: 9.600084 seconds Total frames: 240 Average time per frame: 10.815991 seconds Shortest: 6.651670 seconds Longest: 55.615163 seconds Total time: 43 minutes, 15 seconds Doc Block fucked around with this message at 13:03 on Aug 9, 2015 |
# ¿ Aug 9, 2015 13:00 |
|
KoRMaK posted:gonna try agin with the goal of making cool webm's don't make webm's make h.264's
|
# ¿ Aug 10, 2015 17:24 |
|
qirex posted:you gott kern that poo poo bro, too much space between the j and e [and probably the e and t] w e s tj e t
|
# ¿ Aug 19, 2015 05:50 |
|
|
# ¿ Aug 19, 2015 13:28 |
|
Snapchat A Titty posted:e: lmao just found this old one of mine on my parents imac: I'd be stoked as gently caress if I woke up tomorrow and could draw that well edit: 1 month before 9/11
|
# ¿ Aug 22, 2015 04:36 |
|
|
# ¿ Aug 24, 2015 01:42 |
|
|
# ¿ Aug 27, 2015 05:23 |
|
almost certainly
|
# ¿ Aug 28, 2015 07:21 |
|
built an animation stand to do stop-motion animation for... stuff. gaze upon my works, ye mighty, and despair.
|
# ¿ Sep 11, 2015 01:20 |
|
poo poo and anyway it's "Look on my works" so whatever
|
# ¿ Sep 11, 2015 01:28 |
|
|
# ¿ Sep 11, 2015 02:35 |
|
v nice (that's a lot of air resistance)
|
# ¿ Sep 11, 2015 13:53 |
|
echinopsis posted:
the noise or w/e looks like little particles of dust floating through the glow
|
# ¿ Sep 11, 2015 21:44 |
|
echinopsis posted:yeah i know! and I can't think of an easy way to recreate it properly without manually placing dozens of little meshes Does Blender/whatever not have a particle system?
|
# ¿ Sep 11, 2015 22:39 |
|
Niiiiiiiice!
|
# ¿ Sep 12, 2015 03:09 |
|
that's the kind of thing you do at the compositing stage, so you can tweak it without having to wait hours for the result
|
# ¿ Sep 13, 2015 23:51 |
|
No, it's fine and a cool idea. this thread is for constructive criticism, circle jerking, and general positivity. anyone who doesn't like it can take their negativity and gently caress the hell off.
|
# ¿ Sep 14, 2015 18:16 |
|
ur so gonna burn in hell echi edit: i chuckled
|
# ¿ Sep 17, 2015 07:06 |
|
More experiments with my new AppleTV dev kit and Metal: particles edit: Ocrassus posted:so I'm not very good at photos but I was quite impressed with this one that I took of the Tokyo skytree. Nice! MY TABLES!
|
# ¿ Oct 9, 2015 06:02 |
|
65535 particles
|
# ¿ Oct 9, 2015 07:22 |
|
FPS on that GIF is set to 60, maybe you're computer is a POS? edit: you're context sucks Doc Block fucked around with this message at 15:52 on Oct 9, 2015 |
# ¿ Oct 9, 2015 15:46 |
|
here, watch it @ 1080p 60fps https://www.youtube.com/watch?v=RMrc4d5AKlc
|
# ¿ Oct 9, 2015 15:49 |
|
IF you have the new Apple TV AND IF you have it connected to a Mac running OS X El Capitan via USB THEN you can record video off of it using QuickTime Player (File->New Movie Recording, change camera and microphone to AppleTV) edit: this also works for iPhones and iPads connected with a lightning cable and OS X Mavericks or Yosemite and the video quality is way better than what you get with Reflector etc. Doc Block fucked around with this message at 16:10 on Oct 9, 2015 |
# ¿ Oct 9, 2015 16:06 |
|
Symbolic Butt posted:doc block watching this is making me sick
|
# ¿ Oct 9, 2015 16:11 |
|
Textured point sprites are fun In full 1080p 60fps glory: https://www.youtube.com/watch?v=2TBXu-EpSfM
|
# ¿ Oct 9, 2015 19:03 |
|
|
# ¿ Oct 9, 2015 19:04 |
|
|
# ¿ May 22, 2024 14:56 |
|
atomicthumbs posted:
2 scary 4 me Nice!
|
# ¿ Oct 17, 2015 07:30 |