You mean AskNK stole the Houdini 19 Sneak Peak video, removed the SideFX branding in order to get views?
Yes, he did, but the guy has a sizeable following, meaning that a lot of peoples are seing the updates, and he does link all the SideFX URLs, this is part of the media jungle, but at least he's doing a decent job at it IMHO!
You mean AskNK stole the Houdini 19 Sneak Peak video, removed the SideFX branding in order to get views?
Yes, he did, but the guy has a sizeable following, meaning that a lot of peoples are seing the updates, and he does link all the SideFX URLs, this is part of the media jungle, but at least he's doing a decent job at it IMHO!
Why remove the SideFX branding, as if 'AskNK' authored the video?
I'm hoping that SideFX gives Mac users some love too....not expecting full Metal support in Karma XPU, but at least a wee bit of ARM optimization in Houdini 19 would be much appreciated. I know that Apple has contributed code help to SideFX in the past, so I hope that trend continues.
Daryl Dunlap Why remove the SideFX branding, as if 'AskNK' authored the video?
Like I said, media jungle, but it does give Houdini 19 some attention, I wonder how SideFX sees these videos!
I am sure they have nothing against it. Its a free marketing that only benefits them and yea the guy who did a video, so in this case both side can be happy.
I saw your thread with the carriage and I hope you don't take it as a personal attack, but that's just not a very compelling showcase of modeling prowess. And even if it were, because I'm not saying that one can't model just about anything right now, I'm talking about the road of getting from A to B in Houdini, which is rather assiduous as it stands.
It looks to me that the users defending Houdini's DM capabilities, assume that the users criticizing it, have just not tried enough of Houdini's modeling. If only they'd change their mindsets...
having also come from C4D, while his carriage is simple on the surface it is all quad, a rarity in Houdini modeling where many use Boole modeling and don't bother to clean up tris and ngons. even H19 sneak peak video has a model which has horrible geo (6:52) granted it was Solaris and Karma. you are correct when modeling in H it can be an adventure getting from one end to the another. I will say modeling tools have improved since H17.5 when I first started using H.
ctedin If Houdini has full support for animated GLTF/GLB export to Blender's level, I'd feel a bit better about dropping Blender. Motionbuilder is still the standard for motion capture work, but Houdini's KineFX is really impressing me. I've been looking into Modeler 2.0 from Alexey Vanzhula, but wondering if the new tools in 19 would close the gap. I might pull the trigger if my next project needs more traditional modeling. I LOVE Houdini for most of my modeling needs. It's really flexible, although the interface is not as smooth as I'd like.
GCharb the new pie menu in Blender comes directly from Houdini,
they both got it from maya Blender has pie menus 5 years earlier than houdini to be fair
In all honesty I have no idea, they just look pretty much the same, so I assumed that it came from Houdini, just like I assume the blenders' geometry nodes come from Houdini as well, although I might be mistaken!
Sorry, I should have been more clear. Missing in 18.5 is any animation exports, especially multiple named clips/takes. As far as I can tell, it only exports static meshes. I use animated GLTF files to three.js files quite a bit. It seems that version 19 might support this, which would be fantastic. If anyone can find the animations export setting for GLTF in 18.5, I'd be pretty jazzed.
If you compare the stock exporter in Blender to Houdini 18.5, there is really no comparison. Lots more options in Blender. This is just a small sample. You can scroll up and see more. Blender has been my tool of choice for Three.js work. Houdini could really blow it out of the water with only a little work.
Sorry meant melt edges not melt deformer , specifically doing subd diamond pattern, changing the edge flow, etc... usually provides all kinds of weirdness which takes you down the debug the flow path, and elsewhere be it maya, 3dsmax, c4d or blender is usually less involved process...
hMonkey Sorry meant melt edges not melt deformer , specifically doing subd diamond pattern, changing the edge flow, etc... usually provides all kinds of weirdness which takes you down the debug the flow path, and elsewhere be it maya, 3dsmax, c4d or blender is usually less involved process...
Houdini is a low level DCC, meaning that you get very fine control over everything, which usualy also means a more involvbed process, but it also means an incredible level of controls, all you say can be done in Houdini from what I understand of your question, like you could use the labs edge color tool to select edges is any sort of ways, and then melt them, if that's what you mean!
What GCharb said. I use Dissolve or polyDraw when I need to delete edges. Also I would not say that C4D tools are poor. Those that are there are quite solid, its just there are many things that are missing.
hMonkey Sorry meant melt edges not melt deformer , specifically doing subd diamond pattern, changing the edge flow, etc... usually provides all kinds of weirdness which takes you down the debug the flow path, and elsewhere be it maya, 3dsmax, c4d or blender is usually less involved process...
Houdini is a low level DCC, meaning that you get very fine control over everything, which usualy also means a more involvbed process, but it also means an incredible level of controls, all you say can be done in Houdini from what I understand of your question, like you could use the labs edge color tool to select edges is any sort of ways, and then melt them, if that's what you mean!
That was something that I tried when I started using Houdini (and am still long way from knowing it, still using c4d and blender as a crutch), however I just retried the same approach, which is as you would do it in other apps which is kinda wrong for houdini and just for fun I used foreach prim block which doesn't seem to preserve groups, instead of polybevel inset individual elements which is something that starting out is going to trip you up... anyway it's an adventure
However making a complex pop sim no issues at all, but modeling always seems to take me down the wrong/time consuming path esp. when you need to just make something that does not need to be procedural :/
for the not-that-kind-of-an-app argument - having things under one roof would be better...
hMonkey However making a complex pop sim no issues at all, but modeling always seems to take me down the wrong/time consuming path esp. when you need to just make something that does not need to be procedural :/
Earlier I shared a modeling workflow that is pretty much like modeling in blender, pretty simple and straightforward...
"Just press CTRL-B to bring your 3D view to full screen and model as you would in Blender, the blenders' pie menu comes directly from Houdini, just press C to bring it up, use 2, 3 and 4 for points, edges, and poly selection, and model away, the nodes will be created in the background, and if you ever need to go in the network editor for tweeking, you only need to press CTRL-B again, no need to switch to edit mode and all that jazz, its right in the interface, I do it all the time!"
Could someone clarify whether these points from the announcement mean that vellum will now be able to interact (both ways) with other solvers, like RBD and Flip?
'.Now a fully complete, unified multi-solver. .Native support for XPBD fluids, rigid bodies, and plasticity.'
I've heard some conflicting views on this so I'm not sure what to think.
pu - Karma ROP for instant rendering of non USD Scenes -
karma working in obj context?
Looks like something like that, I guess we will have to wait until next week for some answers
SideFX confirmed that Karma ROP is just a wrapper for the Scene Import LOP into USD, since Karma is a USD render engine. I'm assuming they invested quite heavily in the Scene Import LOP having parity with OBJ context for H19?
I interpreted the Karma ROP like "Yes, it's still based on USD but we take care of it for you. The conversion is in the background as ifds for Mantra". Thought it was a method to make the transition smoother to the stage context.
Faitel I'm pretty interested how XPU would be scaled over cluster. It's multiple cpu, multiple gpu or both at once, Hm...
Or can you use a cluster to render locally to mplay/viewport?
Possibly on the basis of cpu/gpu availability, that's what's attractive about it. So, perhaps combined gpu/cpu on local workstations for fast iteration and then cpu when it goes for overnight renders.