Agreed this is a fantastic resource! However, I'm confused by the use of cos(a) and sin(a) in the quaternion rotation about a vector definition, when all other sources I've seen use cos(a/2) and sin(a/2). What is the reasoning behind this?

the tetrahedralizer doesn' t work very well proved in this video kzbin.info/www/bejne/imiYhKiZd8iZg9E where the generated object haves missing faces that needs to be patched , i also had the same problem with my generated object i hope you see this and either update the blender plugin or address this issue in the description

I dont understand the constraints. Are you simplifying by using springs between vertices or something else? I see you have some kind of max and min of lengths as well as volumes.

Looks great! Any plans on video about Shallow Water Simulation?

Great channel, but I fall at the first step, more with the simulation than the physics. Take the cannon ball piece, works great but I have cpompletely failed to be able to speed up or slow down the animation. Suppose I wanted it to run at 8 fps, or 50 fps. Is that so much to ask or myself? Spent two evenings on this which is hardly a lifetime but any chance of a hint? I looked at requestAnimationFrame and setInterval, what do I miss?

Very interesting channel you have here. Congrats! Can you give me a quick tip on how to add and remove water from the system? Thank you.

Thank you for the great video! I tried extending your implementation by adding ground collisions, essentially applying a correction the size of the penetration depth whenever it is greater than zero. Unfortunately, this behaves very unstable for me (opposed to the mouse grabbing updates), and collisions seem to add rotational energy to the system. Are there any important extra steps one has to take when adding rigid body collisions with XPBD? Very much looking forward to the next video in this series!

haha you were one of my idols as a high schooler, I read so many of your papers, I hardly understood most of them, but they were very cool, and I tried my best to implement some papers, especially for liquid physics.

Hello Matthias. Thank you for this tutorial - it's very helpful and informative. Currently I'm playing with it and I noticed that you probably wrote it in GL and then adapted to simpler version for make it compact and be able to fit in about 200 lines. What struck me at the first time was that the code in a pure JS is many times faster than in Python, thanks to specialised typed arrays of floats. I also noticed, that changing tables to 64bit floats reduces conversions in JS and let save another bunch of FPS of frames. Adapting code to workers and WebGL also made it so fast that in practice it's the best solution for visualising and presenting physical simulations with relative small effort.

No gravity When a plane wing speeds through the air, it does not curve air but compresses it, and force is felt by the wing. We, earth, solar system, galaxy, we are racing through a very rarified "atmosphere" called space; rarified at a macro level, but at a quantum level, not so empty. A plane wing does not "curve" the air, nor curve any temporal measure. Massive bodies do not curve space nor time; massive bodies instead are racing through space, compressing everything, including at the quantum level, like the Higgs Field, bosons, etc. But compressing due to the quantum "stuff" flowing into all the matter of the body and finding resistance from the quantum energies of all the mass in the massive body. We are pushed to the center of the earth, yet held on the surface by the macro structures of the body. All material motion is from collision, and at the quantum level our mass is being collided with by "Higgs Vibrations" equalling about 32 ft.per second per second from where I sit. Space, with all its fields, is being actively compressed as we turbulently swirl and soar through it at a couple million mph. There is no attractive force called gravity - that would not be collision, so it cannot be. All material motion and change is from collision; there is no attraction. The Higgs Field is the atmosphere of Space. Galaxies and their stellar bodies are turbulently swirling, moving through this atmosphere of space at several million miles per hour. There is turbulent disruption of the Higgs Field, with vast compression increasing around all massive bodies. This is the dynamic reality of the universe. "Curved Space" is a photographic snapshot of the compressing of Higgs Field, frozen in time, just as a photo of an ocean wave looks like a solid ramp. But no curvature. The Higgs Field is forced into each body, interacting (colliding) with the quantum fields of the stellar objects, pushing them toward the center of mass (but the mass of objects resist each other, giving the "gravity" appearance). Only light waves seem not to collide with the Higgs Field, but it does curve as it passes through the collapsing density of the Higgs Field around massive objects. No gravity. No curved space/time. All material motion is from collision, of matter, of fields, etc.

The PyOpenGl website appears to be removed.

Thanks Matthias!

Hello, I have been paying attention to your TenMinues, thank you very much for your efforts.I would like to ask whether the rigid body simulation given PBD or XPBD is better in performance and stability than the traditional mode?

So how to implement an agular bending constraint? There is no infjrmation about it in the further tutorials

result is terrible

My previous comment contained the link, which appears to have been deleted. Thank you very much matthias for your explanation. I learned a lot of useful knowledge through this series. In line 277 of the 22-rigidBodies.html file, determine whether pos is defined (if (pos == undefined)). This conditional statement does not seem to be triggered under any circumstances. I don’t know what it is written to handle this condition.

3:00 Should have mentioned PBD trades stiffness troubles for high-mass-ratio troubles.

thank you once again for your work. i don't have much luck prosletysing with young developers, i've been attacked dozens of times for apparently suggesting that people who were moments ago saying 3d is too difficult aren't prefectly valid being incapable. terrible! but you're lovely.

me: just index by grid and go through the grid. humanity: add more computations but not rnd = 196314165 * rnd + 907633515; learned from you but you talk about this like it finds collisions if it's lucky. i'll have to wait for a bigger project to actually think about the difference there.

8:40 I'm wondering why you excluded the cross term in torque calculation for a rigid body. The full equation would be, Torque = I*alpha + omega cross I*omega. Excluding the cross term surely works most of the time, but there wouldn't be a way to explore (for example) gyroscopic precession.

This is amazing, I've wanted a series like this for the longest. Can i ask for one extra physics topic, the 4 bar linkage problem. Most tutorials on that isnt very good plus it's always missing code samples. Something id like to learn and add to my inverse kinematics projects to perform procedural movements. Again, thanks for the rigid body series ❤

Thank you, so much! I've been waiting for this

Physx is a powerful engine but the documentation is very lacking…

Pure gold!

JavaScript 🤮

Say give write code sumall to create cloth simulation

"phys simulation in JavaScript" killed me

The return of the KING 👑

Wwoooww so happy to see update from this channel

The velocity/position update has something that is bugging me. If at t_i we say that velocity is now v_i=v_(i-1)+∆t•g and then the position is x_i=x_(i-1)+∆t•v_i, this means that a body in free-fall starting from rest would cover a distance of ∆t•g on the first time step, but it should cover 1/2∆t•g on that time step. Am I getting something wrong?

No way, a new video from Matthias himself!! Just discovered this channel and was blown away by the content. Thank you!!

Thank you!

he's back!

Wow what an amazing breakdown of rigid body simulation! Makes it more understandable for mortals like myself. One thing I’ve always struggled to understand with PBD is how an example of collisions would factor into the simulation?

been waiting so long for this. Thank you!

Excellent content as always.

Bro where have you been finally new video

I think there's a minor typo In the slides on PBD Integration and velocity update (page 18,19 in the pdf, 10:30 and 11:00 in the video). The vectors should be [ωx,ωy,ωz,0] and [Δqx,Δqy,Δqz] respectively. In the slides, the subscripts are x's. But the code looks good 👍 Impressive demo

OMG Hes back. We thought you got filthy rich from Nvidia shares and would spend the rest of your days on some tropical island instead of making videos for us plebs.

Awesome! I can’t wait to dive into this example!

Godot's creator will be please 😂

I've been waiting for this! Almost all of the current tutorials on that issue are way to complex, thanks for breaking down the topic

Yess more videos! Loved your PBD videos!

Good to have you back

Im too stupid for own physic simulation, i prefer rapier library. Thx for another great video.

at 14:10 you say "explicit integration", but you write "implicit integration". I'm not sure which one is correct, or whether it matters, but in any case cannot simply memorize it, since I have a 50% chance of memorizing the wrong thing :) (I'm actually not sure what the difference is to be honest)

the goat is back

I was learning about this years ago and came across some papers you wrote/where involved with. So happy you are making videos on this too

I was waiting so much for a new video, I was afraid you were not interested in posting anymore 😅

It's good to have you back! 👏