Hey everyone. Wondering if sidefx is going to surprise us with a social isolation gift of Houdini 18.5. Isn't there typically a release around this time?

Well, I think NAB would have started on Sunday this year. I think MAXON is making some announcements next week, perhaps even AutoDesk. Adobe just updated a bunch of apps, and so did AVID.

Right. What about Houdini SideFX?

There is also this in mid-May, but I don't see that Cristin will actually present:

https://www.sidefx.com/houdini-hive/houdini-hive-worldwide/ [www.sidefx.com]

now the daily version of H18.5 is around 200
and H17.5 released when daily version is around 170

I guess H18.5 will release in this month or next month or June

Hmmm. How can one get on the 18.5 beta testing to get those builds?

H18 was the most buggy version of Houdini I've ever seen. I hope H18.5 is much more stable. LOPs was a minefield.

—–> following!!

yeah! usually this should be the right period!!


hope to see soon what has been cooked for 18.5

Who will improve Houdini towards character animation to be better than maya ? noone ?
as better in character animation as it is in VFX..

Personally I feel that SideFX focusing on VFX would be a better bet. It would be very difficult to displace Maya unless you have something that's absolutely revolutionary from the workflow perspective, abstracting all the behind the scenes complexity.

no need to change existing houdini , it already has several versions
I would just add one more version for animators as a separate product
that will include animation , muscles , mocap , physics.

WalterSullivan
no need to change existing houdini , it already has several versions
I would just add one more version for animators as a separate product
that will include animation , muscles , mocap , physics.

They have that product, it's called Houdini FX

No they have not , it's not as good at character animation as maya is.
I know houdini have muscles but it needs depelopment otherwise it will become maya muscles
which no one gives a damn about and everyone looks at ziva today.

you don't understand what i'm talking about when I'm saying mocap.
it's not mocap for crowd animation or something like that..

jsmack

WalterSullivan
no need to change existing houdini , it already has several versions
I would just add one more version for animators as a separate product
that will include animation , muscles , mocap , physics.

They have that product, it's called Houdini FX

Houdini ANIM*

Still no news? Not even with Siggraph?!?

I am convinced that one of these days I'll just launch Houdini, and it will have a totally different splash screen and will say 18.5 without me having installed it!

LukeP
Personally I feel that SideFX focusing on VFX would be a better bet. It would be very difficult to displace Maya unless you have something that's absolutely revolutionary from the workflow perspective, abstracting all the behind the scenes complexity.

I don't agree. Houdini is already great in VFX and Autodesk is working really hard to displace Maya themselves. Their predatory pricing policy, lack of proper development and charging for bug fix releases makes users really hate them. Just look at the Adobe/Affinity case. It took very little for a small company with a good product (and without stock investors) to initiate mass migration. They offer nothing really revolutionary, just solid software for reasonably priced perpetual license. And Adobe is literally bleeding users. I'm on a little crusade to abandon Maya/Mobu myself and I think there are more people like me.

It might be hard to take on Maya from the perspective of ‘user friendly’ programs that work on higher levels of abstraction, like Blender and Modo. Houdini, however, is almost there in my opinion. And not only to take on Maya but Motionbuilder as well. It works fine with FBXes, it supports multiple animation takes properly, all the NLA operations can be done using CHOPs. All Houdini really needs is a quick IKFK control rig wrapper for joint based bipeds (yes, joints, not bones), maybe a straightforward interface for chop network that blends animation takes, a little curves editor polishing and it's ready to go. TBH I'm already working on mocap FBX pipeline to substitute my Maya/Mobu workflow. The moment I accomplish a common task at 70% of the usual speed, I'm buying a license.

n2e, what are the advantages of a joint based system over a bone one?

citizen
n2e, what are the advantages of a joint based system over a bone one?

https://www.sidefx.com/docs/houdini/character/bonesvsjoints.html [www.sidefx.com]

Enjoy…

citizen
n2e, what are the advantages of a joint based system over a bone one?

Universality. You don't work with bones in game engines, mocap data doesn't come with bones either. Generally speaking “bones” is an old idea, they might be more versatile inside the app you're working with but they're also specific to that app. If you want your files outside, you use joints, period. And Blenders example shows us, that converting bones to joints on export may cause many issues.

Fortunately the link mandrake() posted isn't necessarily true, I just found out you can turn a null into a joint by adding bonelink or cregion to it. It's also not uncommon that joints DO NOT orient their aim vector towards the next node in hierarchy so the lookat node may not be necessary. If the FBX exporter in Houdini works well with such setup then I'm fine with it. As for ik solvers and such - it shouldn't be a problem either as you typically want to separate your solvers/control rig from the main joint chain anyway.

A few more words in contrary to the statements in the description [www.sidefx.com] - you actually SHOULD start learning how to rig this way if you want into gamedev. Often times you'll get an already skinned character from a client and knowing how to control it will be essential.

*UPDATE*
Now, that I read my yesterday's post, I feel like it needs some clarification. The whole idea behind joint is the ability to be independent of it's parent. This is very important in more complex rigs where you want to write shape correcting solvers. Apparently bones in Houdini can behave like this, so the whole joints/bones discussion isn't really necessary. Which is good, because my experiment with nulls din't work out yet