first video about a prototype of everything nodes
i have been learning animation nodes & houdini lately and i have to say houdini is more simple to use, in AN to do a simple stuff you need a lot of nodes/math
first video about a prototype of everything nodes
i have been learning animation nodes & houdini lately and i have to say houdini is more simple to use, in AN to do a simple stuff you need a lot of nodes/math
Amazing! They found a way to add it in in the end
God damn are these people talented.
Now we’re getting Eevee on top of particle nodes and procedural modelling
That looks really promising. I really can’t praise Jacques Lucke enough. They could not have gotten a better suited developer for this. Also he lives the spirit of open source - he probaly could have gotten rich selling Animation Nodes on Blender Market (totally justified) but he didn’t.
This is going to be so exciting, Jacques Lucke will probably go down in Blender history as being behind one of the greatest Blender additions of all time (I’m serious, this is right up there with Cycles and Bmesh)
It is so good to see that they are going ahead with the Everything Nodes project. With stuff like this, they should consider producing more batches of USB rockets to sell.
I’m surprised they’re not just calling this Blender 3.0 by now.
And yeah, they need to make more rockets. I somehow managed to miss that they were selling them until it was too late to get one. I want a rocket!
i have been learning animation nodes & houdini lately and i have to say houdini is more simple to use, in AN to do a simple stuff you need a lot of nodes/math
Same. I’ve tried to dabble with AN a few times and just found it confusing compared to Houdini. I’m sure I could figure it out if I stuck with it, but seeing as Houdini is kind of a standard and AN isn’t, there isn’t a good reason for me to put the time in.
AN are able to re-use curve geometry properties, object modifiers and default particles system.
Their purpose is not to make nodal simple things that are quickly done by default UI but to to make nodal complicated things hard to follow without nodes.
Jacques did not try to clone Houdini nodes inside Blender but just to add nodes needed to have things more fluent in Blender. Of course, there are fewer AN nodes than nodes in Houdini.
But they are enough fundamental to cover most of needs by combining them.
With version 2.0, he simplified a lot UI of nodes and need of intermediate converter nodes.
Most of examples in the documentation are just using between 5 and 10 nodes.
Nowadays, Cycles will have lots of principled nodes (surface shader, volume, hair). It was not always like that.
When all basics of nodes will be solid ; logical evolution will be to have shared blend files, nodegroups, templates, presets of nodegroups and eventually, new nodes dedicated to simplified workflow.
Directly modifying Blender to obtain theses new nodes instead of making an addon should help to skip workarounds you may have to create with AN.
Yup, that’s what i miss from AN now, we can’t make a group node so we can just plug inputs and have a direct result, we need a tool so we can build our own group nodes, just like the MOPS from Houdini
( https://www.youtube.com/watch?v=zuE8wP-e2w0 ) .
i hope that working with everything nodes will be more comprehensive than working with AN, for example when i look at something like this :
it scares me a lot
anyway, i’m really hyped about everything node, more than eevee did
the nodetree of houdini may looks simpler, but if you dive one level deeper you feel like you enter the matrix as code it is far more complex but the tools creating workflow for nodes is really great in houdini. Aside from that animation nodes has become really easy to use in version 2.0 and if something like this will be included it will become even more powerful. remember, animation nodes is just an addon
I’m pretty sure the design for Everything Nodes will not be subject to the limitations people encounter if they’re making an addon that uses the node editor (as it will likely be hardcoded in C).
I also wouldn’t be surprised if some higher level nodes are coded as well for ease-of-use purposes.
animation nodes is one of the most powerful and amazing things I saw in blender, at least as an addon. This is extremely promising, to say the least.
oh what the hell, this is how I really feel:
Absolutely fantastic news!
I am always afraid when talking about nodes, I have always found it difficult to understand nodes, especially when there is math and logic involved. Anyway I am calm, with Jacques we are in safe hands.
Do you know if this would involve resolving old limitations of things that are not possible to perform with current particles system, for example collisions between particles or what “molecular” addon can do?
The video showed fireworks in the Blender viewport (which believe it or not isn’t exactly straightforward with Blender’s current particle system).
When it comes to actual VFX, the particle system in 2.79 is limited enough that it shouldn’t take a lot of work to really improve its functionality (especially with the power of nodes behind it).
In other words, being able to emit a secondary particles system from a primary particles system at conditions defining a specified event.
For example, when a drop particle collides this object and dies -> emission of 3 droplets particles.
Dayyuuuum ! Such power ! Such flexibility ! Yahaaaa thanks Jacques !
I am curious how far this could be extended also into modeling commands like the way Houdini works
I would be already super happy with basic object (not mesh) manipulation and be able to control modifiers via nodes like connecting one value node into the bevel size or subdivision level input.
Svershok is I think getting closer to generative modeling.
Here is git repo for everything nodes…
I’d give up my horse to see every mesh operator converted into a node…