Please don't apologize for the video taking too long. I can't speak for everyone but I personally really appreciate the pace and how you break things down. It's really powerful to see how you approach each problem and your voice is easy to listen to as well. Please keep these coming!
I love that you're explaining the problem and how a specific node will solve it. Really good for fundamentals that people can use for their own problems too!
Good explanations, very in detail and comprehensible. Just a quick note for people who want to reduce the amount of nodes in their setup: when doing something like checking the distance between two points for collision, it is not necessary to use two Vector math nodes, first to subtract the vectors and then get the length. The Vector math node has the Distance function which is doing that in one step, because the Euclidian distance between two vectors is the same as the Euclidian distance between the difference of those vectors and the origin (0, 0, 0), which is what the Length function calculates. I'm sure you are aware of that, I thought it might just be interesting for others who do not know this relation.
Hi! Thank you for your attention! yes you're right but In that situation, It was necessary to have a 3d vector for each points to give it to the "set position" node. that's why I needed to use the "subtract" for achieving that vector. otherwise, yes. we can use "distance" instead of "subtract" + "Length". Thank you again for paying attention!
Wow ... what can I say? You are just brilliant, such detailed explanations, all logical, everything perfect, I thank you a lot for your effort and for sharing!
Excellent tutorial and well presented. This setup makes an interesting tool to model ways for galaxies to form. I added an additional force of negative gravity (although his may have a similar effect as the controller and added a switch so that one can choose to have outer particles rotate faster than inner particles or the way you set it up. Large galaxies have outer starts rotating faster than inner stars due to proposed dark matter. Thanks again for an excellent presentation.
--To be clear, the Node Group you created contained the "Set Position" node that was within the "Simulation Zone" AND all of the math (noise, subtract, scale) that fed into it, correct?-- Never mind, it was more clear when you made the second Node Group what you did.
Subbed only because of that cool first image, i'll be sure to watch the entire tutorial, seems like you're explaining things very well if i take a look at the comments 😊 👍 will comment later for real feedback 😊
amazingly well explained dude... please keep doing it because tutorials like "put this into this" are soo poorly explained... you are teaching us to understand geometry nodes, the only way to use it efficentliy instead learning by heart what a single node can do... subed and liked....
Thank you so much! It's very heart warming to hear. because for my personal learning, I always look for the tutorials like this that they explain every thing and it is my main purpose on this channel. But I also understand that some people don't like this kind of tutorials and that's totally OK.
well people who dont like tutorials like this are never really going to understand geometry nodes wholy - their loss if you ask me... the best thing about tutorials like this: we learn much how math is used in such enviroments like geometry nodes,but this knowledge you can transfer to shaders for example, even to another node based programs like touchdesigner or unreal engine... you are educating us, not only teaching geometry nodes, and thats really worth every seconds watching it buddy... as I said, just keep doing it like this and you'll get subscribers in no time@@blenderpath
@BlenderPath Thank for this detailed tutorial and in-depth explanation! There's a detail that I can't reproduce, neither with Blender 3.6 nor 4.0.1: The points never move up the Z axis. They follow the attractor but only horizontally. Did I miss anything?
Thank you and please forgive me friend. honestly I couldn't understand exactly what you are talking about. If you still have your question please send to my email (blenderpath.contact@gmail.com). Maybe we could talk about it more in depth there.
Thanks! eevee doesn't render point clouds, so if you want to use eevee you should use "instance on points" node and an ico sphere for instancing at the last step.
Nice tutorial Blender Path (Liked and Subscribed since this vid is one of the best geometry nodes physics tutorial). Though, I have one question about the implementation of the rotation. You could have also implemented it in another way (100% accurate) and all you need to care about is the angle of rotation. So, why did you choose to go with the sampling method (the one in your video)? does it provide a better rendering speed? The technique I am talking about is explained by Fraya in this video: ru-vid.com/video/%D0%B2%D0%B8%D0%B4%D0%B5%D0%BE-7j5yW5QDC2U.html But long story short: p = p.x * (cos(a), sin(a)) + p.y*(-sin(a), cos(a)) Where: `p` is the point you want to move. (one of the distributed points in your example) `a` is the angle you want to move the point with. ---------------------------- The only work, you need is to precompute the sine and cosine of the angle you want to move with. Thanks again for the great video, looking forward to see new videos with such great explanations!
Hi! Thank you so much! well there are 2 reasons: 1. Since blender GN didn't have any proper rotating and twisting tool, I made such asset with the method you mentioned for my self. but that method is best for static situations, not dynamics necessarily. because in dynamics, we need one vector AT EACH FRAME for the points (or any thing else) to move them along those vectors. in dynamics we need to have more control for mixing multiple vectors together at each frame to make more artistic or random movements and for many other reasons. this method you mentioned which I used that too, is very nice but dynamics is more about vectors at each frame. for example in this dynamic method you can interactively move you center object and see the changes in points movement, while you can't have that in static mode. 2. well this was simulation tutorial and I needed to explain basics of simulation along with making a cool effect :)) also needed to explain how to increase accuracy and other stuffs. Also thank you for asking this nice question. I hope the video was useful for you.
@@blenderpath I spent the last hour implementing the other technique (couldn't sleep without checking your point :P ). Because I was sure that both are dynamic (in sense they respect the center). But, one pushes particles away (not perfect angular motion), and the other rotates them with perfect angular motion. And it works even with a moving center. Will share it on my channel and post here the link. My conclusion is: the only difference between the two in my opinion is the application. The sampling technique is better to simulate hurricanes like forces as the velocity is not literally angular. Hurricanes tend to throw away objects (outward-direction). However, to simulate a perfect angular momentum (mostly seen in dense-sticky liquids rotations such as honey) the angle-based rotation seems to look better as particles are not thrown outward.
Oh... I didn't think that you meant to implement that method in dynamics, because I didn't do that for myself since it needed more nodes and would make the tutorial so much more complex =)). but yes. we can do that also in simulation nodes but honestly I didn't even think about this method in here and that it results in accurate angular movements but yeah, it does. so that's nice 👍and thank you for recommending that. Now I'm interested to check that myself.
Yep, my geometry nodes panel looks hairy now 😂 Anyways, it was fun checking out the simulation zone. I never thought that this is possible in Blender! Thanks for mentioning it
