Hello, I'm attempting to follow a tutorial for Houdini Indie that unfortunately uses version 17.5.327 - in his video, he sets up an arrangement of nodes that I was able to duplicate, but got stuck on Transform, and Points From Volume.

Ref: https://www.youtube.com/watch?v=oRzcGI4kbsQ [www.youtube.com] -- Voxelization Workflow, Creating a Digital Asset (HDA)

I've done searches on the forum, the tutorials, and other documentation -- but I haven't found one geared towards the workflow I'm trying to emulate in the above video. I know the nodes exist, because I've used the built-in help and read the docs, but I'm a bit lost on how to wire up a FBX or OBJ subnetwork node to achieve my goal.

I'm at the point where I'm willing to pay a bounty in Bitcoin to get this done, as I realize I'm asking for work that may be non-trivial in execution.

Just a bit frustrated. Houdini is massively impressive and I want to do custom things, I'm just having a rough start.

Any help is appreciated.

Seventy views, and not a peep. Wonderful.

From the outside, Houdini looks impenetrable and very complex.

All tutorials are based on revisions-ago versions, and even after going through a good number of the tutorials and documentation, its clear that everyone is just too busy down the rabbit-hole of Houdini's internals to even surface for one moment to help a n00b.

Thing is, the n00bs are the people who become new users.

Usually.

I might figure this out, or I might not. What I will remember is that even when offering to pay for help -- no one here had the time or inclination, and that diminishes you all.

the problem is not whether you pay or not, nobody here expects money for help and people are very helpful

the problem is that you are very vague with your question, its not even clear what you are asking, you linked a tutorial yet didn't describe which part you are not able to replicate
it may not matter that it's for 17.5 or it may, but you have to point put which part is giving you different result than the tutorial

or are you asking how to bring geometry from multiple objects of object subnet into SOPs to even be able to start voxelizing?
for that you can do Object Merge with the pattern like /obj/subnet1/* or /obj/subnet1/*/*, ...

or you may be able to bring fbx directly to SOPs using File SOP even though it may not support all features, or if its a skinned character using FBX Character Import SOP

It was pretty clear if you took the time to look at the video.

He steps through the process of using a mesh - in his case it was a pikachu - and establishing a node network that first pre-processed the model to recenter and align, and then creating a grid of points conforming to the mesh volume.

I'm sorry, I thought by asking the question -- "Hey, can anyone translate this guy's network to v19 equivalent" would be clear enough what I wanted to do.

The subject in posts are limited, so "Voxelizer Tutorial - v19 Equivalent" was as concise as I could be - with my further explanation later.

I said clearly that I was trying to emulate the video -- get the result he did in v19 of Houdini.

If that is confusing, then I don't know what to say, really.

TallTim
I said clearly that I was trying to emulate the video -- get the result he did in v19 of Houdini.

upon clicking through the video I don't see immediately why it wouldn't work in H19
what was the part you got stuck at? do you have a file up until that point?

if it helps, he posted the hda on Orbolt (at least for even older version, but seems to work in H19, which suggests that the tutorial should also be possible to follow in H19)

if you want to check finished his HDA https://www.orbolt.com/asset/Kino::Kino.mVoxelizer

tamte
if it helps, he posted the hda on Orbolt (at least for even older version, but seems to work in H19, which suggests that the tutorial should also be possible to follow in H19)

if you want to check finished his HDA https://www.orbolt.com/asset/Kino::Kino.mVoxelizer

I didn't see that, and thank you for your patience in pointing that out.

My frustration stemmed from not being able to recreate the node network he was laying out - for instance, once you dive down into a subnetwork node, there isn't an option to use "Transform" for example. There's "Extract Transform" but it didn't seem to have similar parameters. Same thing with "Points From Volume".

I'll take a look and investigate -- thank you again.

TallTim
once you dive down into a subnetwork node, there isn't an option to use "Transform" for example. There's "Extract Transform" but it didn't seem to have similar parameters. Same thing with "Points From Volume".

Sounds like you are in the Object level Subnet, so you see just nodes that belong to Object context

Mentioned Transform and Points From Volume are nodes in SOP context, so you need to dive into a Geometry node first

tamte
Sounds like you are in the Object level Subnet, so you see just nodes that belong to Object context

Mentioned Transform and Points From Volume are nodes in SOP context, so you need to dive into a Geometry node first

First, I'd like to report that the HDA you linked does indeed work in v19 -- so that is fantastic. Also, drilling down on the node network there are distinct differences, which makes me feel better that I wasn't missing something simplistic.

As for levels in the network editor, your point is well taken, I was following along step-by-step with the video -- so in his version making certain nodes at certain levels would work, but for v19 -- forget it, no way.

Its a good thing to keep in mind, you are correct of course.

Thanks for the help, and if you want a tip -- hit me with a Bitcoin address and I'll oblige.

Cheers.