I've been messing about with this growth type setup, basically having particles randomly spawn new ones and creating trails behind them - and this scene is just one of a variety of setups I've tried today - but as soon as it gets the least complex, stuff just breaks and I really don't get why. I tried everything, keeping things in groups, keeping them in streams etc, I've tried using VEX snippets for choosing random id's to spawn from, I've tried the POP group node, but no matter what approach I take, nothing works as it should.

So is it me or are the POP's just not very stable?

Or perhaps I'm using the wrong approach? Is it better to stack POP's for this kinda thing? Or can I stack solvers in the DOP context, piping the POP solver output into new POP nodes and onto yet another solver, etc? And I'm used to doing pretty complex stuff in Pflow and TP but I just don't know what approach to take doing similarly complex setups in Houdini…

Maybe this tut from cmiVfx is helpy:

https://vimeo.com/83814122 [vimeo.com]

Promo Mail from cmiVfx… from 20.03
Step 1) Visit http://www.cmivfx.com [cmivfx.com]
Step 2) Select as many videos as possible (Here are some links below)
Step 3) Enter in this code: 70thisweek (case sensitive text) in the store cart

I actually bought that way back but haven't watched it yet - so I'm doing that today then, I'd totally forgot about it. Tnx for reminding me.

With that said, having watched the promo, I'm pretty sure he creates the effect in a SOP solver, so still not really the solution I am after - but it might take me in the right direction. The thing is I'm learning Houdini so I'm not always looking for workarounds but for a solution in a particular context, in this case POPs, as I really need to learn how that context work, “inside and out”, so to speak.

From your network, this is my best guess as to where confusion is arising:

In the old POPs, if you referred to a particle using the group field, that particle will imported regardless of whether it is in that logical stream or not. In the new POPs, the group field will *only* select from the current stream.

With this in mind, the next question is what is the “Stream” downstream from a POP Replicate. Nodes after the replicate refer to only the replicated particles. Not to the source particles. This is a bit odd when you have stream_gen_1 in the field, but makes more sense if you had blue_particles in field, you'd expect only replicated blue particles to be affected.

In this case I'd suggest using the POP Stream node. You have a bunch of operations (noise, cling, trail_generations) which you want to apply to both stream_gen_1 and stream_gen_2. So instead of setting the group field on each of them, you can make one new POP Stream as a new chain. Set it to stream_gen_1 and stream_gen_2. Then wire those three nodes after it and into the merge.

If you skip the POP Stream you can have a stack of nodes whose stream is all particles, thus letting you mix and match groups as you wish. The tricky part is that Sources (including replicate) create streams, so nodes below them are restricted in what particles they effect.

I don't know what other things you tried, but hopefully the attached is similar to what you intended in this file?

So that file is not working for me either, this is what the Gen_2 node. (H14 281)



But I feel the question in part has become, is this a reasonable approach for the kind of setup I'm looking for, having particles create a growth pattern like this? I'm starting to suspect I might have been way too concerned about doing this type of setups inside a single POP network instead of doing it step by step in SOPs…

Here's a setup I did stacking POPs instead of trying to do it in one network…

Yes, the problem is I'm not sure what you intended to do. :>

Stacking it in SOPs is quite a sensible solution if your particle systems have a one-way interaction.

Yeah, so I'm starting to realize I probably been way too into having everything in one DOP but splitting it up I suddenly feel I'm back in the workflow I fallen in love with, working in Houdini.

But I also need to get more familiar with what works and what doesn't in one single POP context - but it's really annoying when things break without any clear reason, like in my first file, but if I understood you correctly, it's breaking the stream, so I'm going to try setting this up, splitting the streams in the network, so to speak, see if I can make it work.

So I tried using stacked POPs and it really worked out great for this setup…

http://vimeo.com/123357428 [vimeo.com]

…but I've gotten a scene file with a working setup using the groups for doing this and I will dig deeper into it and make sure I really get how they work. I just love TP and how you work in groups and if you can use the streams like that, shuffling particles in and out of streams, it's just something I need to know inside- out.

sterams = groups.

Just dive inside any of the POP * DOP micro-solvers and there you will find SOP-like workflows and plain old geometry groups.

If you want to know POPs “inside- out” then dive in to the POP micro solvers and dissect. It's just SOP geometry workflows supported inside DOPs.

You can wire POPs in to any other solver pink inputs as well and massage any “Geometry” bound to any sim object this way.

jeff
streams = groups

Well, that's what I thought, but grouping something in the same stream doesn't break it but as you can see in the file I posted, one stream can/will/might (?) break another. So seeing streams break when conflicting streams are in the same flow, they at least work differently in that sense.

Now, I haven't dug very deep into the streams setup Ian posted but I got the gist of it and the idea is to replicate my current setup [vimeo.com] (w. stacked POPs) but keep it in one POP.

And thanks for the tip about the microsolvers, I'm hell bent on really learning how to best utilize Houdini's particles so I really need to get all this figured out fully.

So it took me a couple of days to figure out how stupid I am and how, as the stream is evaluated every cook, there's no need to split it into more than three groups/streams, one to lead, one to spawn and one to trail - for this particular setup - and it's no problems having that in one flow.

And I also figured out (with help from Ian Farnsworth) how to split out streams without breaking the main flow, using the reference stream input in the POP replicate node.

And thanks for the help & info. Finished setup in the attached hip-file.

Can anyone help in this type of effect can can i give that variance in lifespan and that particles dopping effect and replicate particles after a parent particle dies,am using pop kill node for this but its creating a replication for every parent particle