Houdini 16 - Can we borrow some ideas for 2.8

Houdini 16 video is out. Was wondering what features they have that we can also request for Blender 2.8.


https://vimeo.com/202435893

I think it would be easier to make a list of features that the BF shouldn’t borrow from Houdini.

Rarely do you see a 3D DCC vendor doing almost everything right development-wise like SideFX.

The auto rigging toolset looks really nice as well as the terrain tools.

I personally would like to be able to see tiled terrains in blender just so we can bake them down to heightmaps.

the ocean in Houdini is awsome, Blender are far behind

Houdini should be the only example for Blender development if any.

There are many plugins that you can buy for blender that do some of the things in this video. Maybe I should make a list.

SideFx make 3d development sexy :stuck_out_tongue:

We can not really request Houdini 16 features as new features to fill a lack in Blender.
For everything demonstrated in this video or showed as new Houdini 16 promo pages; there is already a corresponding solution in Blender.

All we have to wish to Blender 2.8 is the ability to improve flexibility, simplicity of existing Blender tools.
Houdini was built since ever with procedural process in mind. But it was not as generalist as Blender.
Some of Houdini 16 new tools are present in Blender since a long time.
( Adaptive Profile of Loopcut, Volume Snapping for Bones, Reroutes nodes, …)

Node Editor have already alignment tools.
Fracture modifier can handle arbitrary cutting planes.
Basically, Houdini features went in same direction than Blender ones.

Of course, in many areas, houdini is going further. We could pick some ideas for node editor UI ( node shapes, warning icons…)
But about what is impressive, in Houdini 16, they basically take their existing tools and tried to add a layering system to each of them (to ocean generation, to hair grooming, …).
It is the same idea that contains “everything nodes” paradigm.
Being able to mix patterns obtained from different procedures.

If modifier nodes are done in 2.9, Blender will look like Houdini at that moment.
Efforts are made on viewport for 2.8. Here, it is also important topic mentioned by Houdini devs, too, about their work to facilitate lookdev.

Again, Blender is on the same way. We don’t have to request anything. We just have to support current work.

I will second the “everything nodes”. That would go a long way toward this goal.

One thing Blender REALLY needs to improve is integration between tools and features. As @zeauro said we already have most of those things implemented in Blender in one way or another; the problem is they’re not very easy to use in conjunction with other tools (or sometimes not possible at all). Simulations and particles being the most obvious example.

I’m just guessing, but probably because of the nature of FOSS development, some awesome addons or modifiers get added to the program, but they rarely get done thinking about how to integrate them with other existing features. So we are left with half-working tools all over the place.

Let’s hope for 2.8 and/or 2.9 to improve on this… The “object nodes” project sounds like it could be the answer :smiley:

Very cool, I especially like the ocean and liquid sim stuff>

Also, did anyone else find it weird to see Lee ‘that head model guy’ Perry-Smith with hair? :smiley:

I’m personally not a big fan of using Nodes for everything. I tried modeling in Houdini, but it took way to long to model using nodes. Nodes are just not a fluid way to be an artist. Rigging, FX, Layout, materials makes sense with nodes, but those can be considered TD task.

I like what you said here. I totally agree.

Yeah, I feel like it’s really easy as a CG artist to look at what the other guys have, having never used that software, and think, “Ooooh, that looks so cool!” In actual use (I’ve used Houdini before but never on an actual job) it kind of

For me, the issue with Houdini is that it can be needlessly complex. There are so many situations where they supersede the node network by forcing you to have to create and keep track of all kinds of variables and expressions. So much of what makes Houdini so flexible is this ability to get down deep into the scene graph and manipulate things just the way you want.

However, in practice, this can be a nightmare. Especially for Blenders target audience! I mean, just think about how difficult it is to setup a driver in Blender. It seems needlessly complex for the average beginner hobbyist user. Now imagine, this is exactly how the whole program works. Imagine that you went to the physics page to add a cloth effect and were presented with 500 parameters (I’m exaggerating of course) but you really only ever use about 5 of them most of the time.

In my opinion, this is actually one of the main the strengths of Blender. As a developer you need to think about the easiest way to add a feature without overwhelming the users. The general effect is that users wind up being faster because there is less to tackle went you’re creating or trying to solve issues. This is also the age old issue with just adding features verbatim from Maya or Houdini. Often times people ask themselves, “How do they do it in Maya?” when trying to solve a particular issue. But sometimes, the solution is far to convoluted to actually expect users to figure out.

I guess what I’m trying to say is, for the most part Blender developers get it right with keeping a balance between simple to understand and a lot of flexibility/deep control. Any feature from Houdini would be welcome but as long as you keep this in mind and didn’t just copy verbatim. It needs to be balanced.

Loved the high contrast UI design, look at the input fields. Top notch.
Also noticed they have same high contrast in some nodes, almost dark black where sockets/labels are. Nice.
I really really like that clean 3d view, clay gray neutral with dark wires. So clean and easy on the eye.
Invisible rigs looked straightforward.
Muscles looked really perfect for blender in the video, no super over-the-top technical solution.
Looked quite simple and manageable on blender (small studio scale/individual) … how they just draw the muscle, then scaled the thickness.

What I’m still wondering though is whether you can see the effects of the muscles in real-time as you pose the character or it’s another one of those solutions where you have to bake the animation beforehand (as the muscles are simulated).

The reason I ask is that muscle simulation historically requires some rather complex calculations in order to avoid the impression of balloons under the skin.

i have heard there are both options a simple one and a FEM based solution but the proof will be on the 21. feb :slight_smile:

We are slowly moving to Houdini at work, it is not really complex, it is just really different, the main advantage, at least for me, is that everything, and I mean everything, works with everything else, and you can absolutely use anything to animate anything, this is just amazing, as for procedural modeling, I understand that from a Blender point of vue it may seem inefficient, but peoples need to realize how powerful and flexible that is, it is a huge time saver in the long run, and it allows for the construction of assets that can be used and modified on the fly in different productions, like, you build a library of models that can be modified with sliders and used tons of times.

Mantra is close to Arnold in flexibility and quality, even renders faster on some scenes, to the point where we may ditch Arnold completely.

We’ve been beta testing 16 for a while here, and I just may buy an Indie license for my personal projects, although I would miss Blender if I did, Blender has something to it that even Houdini does not have, for one I absolutely love Cycles, although I doubt that it could render the massive scenes Houdini can dish out, Blender modeling tools are better then those of Houdini, although 16 comes a long way to ease that, but the sheer quality of Houdini 16 is just too much to not consider, it can handle huge data sets, so huge that 15.5 introduced a compression node to make those smaller so they could fit on hard drives, and just like Blender it has everything in it, down to a stellar compositor that resemble Nuke.

Power is better than ease of use in production 100% of the time. If you want ease of use in your content creation tool, grab an Etch-a-Sketch, because CG production is never going to suit those needs regardless of the app you choose. Houdini is the biggest shaker in the industry today, which can only be a good thing.

I don’t think that is true, the world is moving towards getting more for less, or in this case, better visuals without the complexity. Eventually, we’ll get to the point of dreaming it and it will be created for you in real-time.

mmm, in combination with blender… i dream of a bridge…

@thomascheng
complexity is the bridge over the uncanny valley… world of human is evolving into such an incomprehensible and complex organism an illusion became the necessity decades, centuries ago (ie: Tesla) & has become a need for the democracy to strive, also to control, to prevent the outbreaks of fear - from the unknown: “Ignorance is bliss.”