I find it somewhat ironic that some people seem to miss the breadth of utility that Erindale creates. His videos are the most consistently useful CG content I have come across in three years of self-study. Thank you so much for what you do!!!
Really, you're brilliant ❤️❤️ You explain so beautifully about node and all the mathematical functions in such a simple way. I am a physics student now studying animation, and your videos are really good for me.
Hm I do have a spare channel in the vector I’m sampling across there. Could also do fixed spacing just by overriding the length instead of making different curves 🤔
This is so awesome, thanks. I'm curious as to what would be a good solution for making arrays to fill surfaces, ideally just being able to add more instancing on y. I find it awesome that I now actually start to understand what you're saying in your tutorials, thanks again.
To fill out a surface is definitely a more complex problem than doing linear arrays. I’d probably look at making curves instanced on a curve to create a configurable grid and then maybe look at sampling the position on to the curves via UV map (Sample UV Surface)
@@Erindale Thanks for the answer! What I needed now was shingles on "japanese-styled" temples so I just made some insets on a plane and separated the diagonal lines. Made it really easy to shape too and align the rotations, works perfectly!
Amazing tutorial!!! I learned so much as usual. I've played with your setup and customized it (good exercise!) by adding position and random position. It was quite a challenge to find how and where to add the vectors and how to modify them so it takes in account the sampled surface and have it work for both Deform and non deform on curve. But I'm happy I found how to do it! I have a question regarding your setup: Have you tried to draw curves ? If you draw a C shape and an inverted C shape, you'll see that one has a perfect alignment and the other one has the C vector at roughly 37 degrees (I found this number by manually trying to align haha). Scaling the one curve -1 in X solves it though (object or edit mode). I couldn't find a solution for that, would you have an idea about it?
Great set of tutorials, many thanks Erin! Do you know is there any solution to distribute the instances aligned to the start and end points of the splines? It is made for sampling all curves lenghts together, so there are gaps or extensions on each spline. I tried some setups evaluating on spline indexes and domains, but no success so far. (something avoiding the usage of the repeat zone). Something like it works in curve deform, but for instances...
finally created another point field for attribute statistic of the max length, last point leading minus trailing, and made correction scale multiplier to the sample curve lenght and instance scale coming out from the instance packer. Repeat zone involved;(
Check the video here please ua-cam.com/video/HXAPVQD2P-A/v-deo.html The unrealised instances (non deformed variant) should start and end exactly at the starting and ending points of the splines. Dont know why but it does not work 100% for one spline. With random seed it works well. Is this possible without the repeat zone? thanks! (reason - performance for large scenes - keep instances but ends aligned like in later curve deformed part)
Very nice. Nerdy Q for you. If I accumulate rotations, using, say, chained 'Rotate Euler' s.. I read backwards from the result.. which is what I would expect. If I 'Align Euler to Vector' , I read forwards to the result, which is nice, but seems inconsistent. (For example, you Auto align X to Tangent, and *then* pivot about locked X for tilt. ) I can't figure what's going on, here, internally.
Hm I still think of rotate eulers as left to right. They’re essentially a stack of matrix multiplications. Best to jump on blender.chat and ask Hans directly, he’ll know much more under the hood info here
@@Erindale 'Talking to teddy' might have sorted for me... one is transforming the vector, the other is transforming the coordinate space, therefore inverse? But hopping onto the chat would be a good idea. Thanks.
All hail the giant pointer!
May it guide us all!
I find it somewhat ironic that some people seem to miss the breadth of utility that Erindale creates. His videos are the most consistently useful CG content I have come across in three years of self-study. Thank you so much for what you do!!!
Thank you that means a lot! I try to bring the value I missed when I was learning proceduralism
I'm a geometry nodes junkie and you're my dealer. Thank you.
The only addiction that’s healthy
@@Erindale is it tho? My back kinda hurts...
I have no idea if I'll ever catch up to all of your teachings but darnit I'm trying.
Lmao good luck amigo
Reading the docs like a book helps along so much- the most concise writing I’ve ever seen
Solved a big issue in my professional workflow with this one. Thanks!
Oh the timing for this is perfect. Thank you so much.
Your videos are truly something
Thank you so much! I hope they’re helpful!
@@Erindale very much. You're the best in blender geometry nodes there are on UA-cam.
I was just thinking this. Build Tools.
great! Cool tool with many use cases, thank you for showing how to build it!
Really, you're brilliant ❤️❤️ You explain so beautifully about node and all the mathematical functions in such a simple way. I am a physics student now studying animation, and your videos are really good for me.
Thanks so much! Glad you’re enjoying them!
WHOOAAAAA !!! 00
Invaluable ! Thank you !
I have to watch this content! lets go! :)
Erin, why did you use sample nearest for the tilt when you couldn't have just passed it through the sample curve node?
Hm I do have a spare channel in the vector I’m sampling across there. Could also do fixed spacing just by overriding the length instead of making different curves 🤔
@Erindale I see later on you used that socket for the spline index. (But as you did for the other thing, you could of used a vector 🤷♂️
This is so awesome, thanks. I'm curious as to what would be a good solution for making arrays to fill surfaces, ideally just being able to add more instancing on y. I find it awesome that I now actually start to understand what you're saying in your tutorials, thanks again.
To fill out a surface is definitely a more complex problem than doing linear arrays. I’d probably look at making curves instanced on a curve to create a configurable grid and then maybe look at sampling the position on to the curves via UV map (Sample UV Surface)
@@Erindale Thanks for the answer! What I needed now was shingles on "japanese-styled" temples so I just made some insets on a plane and separated the diagonal lines. Made it really easy to shape too and align the rotations, works perfectly!
Amazing tutorial!!! I learned so much as usual.
I've played with your setup and customized it (good exercise!) by adding position and random position. It was quite a challenge to find how and where to add the vectors and how to modify them so it takes in account the sampled surface and have it work for both Deform and non deform on curve. But I'm happy I found how to do it!
I have a question regarding your setup:
Have you tried to draw curves ? If you draw a C shape and an inverted C shape, you'll see that one has a perfect alignment and the other one has the C vector at roughly 37 degrees (I found this number by manually trying to align haha). Scaling the one curve -1 in X solves it though (object or edit mode).
I couldn't find a solution for that, would you have an idea about it?
Great set of tutorials, many thanks Erin! Do you know is there any solution to distribute the instances aligned to the start and end points of the splines? It is made for sampling all curves lenghts together, so there are gaps or extensions on each spline. I tried some setups evaluating on spline indexes and domains, but no success so far. (something avoiding the usage of the repeat zone). Something like it works in curve deform, but for instances...
finally created another point field for attribute statistic of the max length, last point leading minus trailing, and made correction scale multiplier to the sample curve lenght and instance scale coming out from the instance packer. Repeat zone involved;(
I think I don’t quite understand. What is the result you’re trying to achieve?
Check the video here please ua-cam.com/video/HXAPVQD2P-A/v-deo.html The unrealised instances (non deformed variant) should start and end exactly at the starting and ending points of the splines. Dont know why but it does not work 100% for one spline. With random seed it works well. Is this possible without the repeat zone? thanks! (reason - performance for large scenes - keep instances but ends aligned like in later curve deformed part)
Very nice. Nerdy Q for you. If I accumulate rotations, using, say, chained 'Rotate Euler' s.. I read backwards from the result.. which is what I would expect. If I 'Align Euler to Vector' , I read forwards to the result, which is nice, but seems inconsistent. (For example, you Auto align X to Tangent, and *then* pivot about locked X for tilt. ) I can't figure what's going on, here, internally.
Hm I still think of rotate eulers as left to right. They’re essentially a stack of matrix multiplications. Best to jump on blender.chat and ask Hans directly, he’ll know much more under the hood info here
@@Erindale 'Talking to teddy' might have sorted for me... one is transforming the vector, the other is transforming the coordinate space, therefore inverse? But hopping onto the chat would be a good idea. Thanks.
Where is the list node ? Like in sverchok.
There are no lists in geo nodes