particle systems in Houdini/Maya

[seanmac]

Hey guys,

I have some experience using Maya's particle systems, and I am interested in experimenting with Houdini's particles. I've been doing some investigating and I've found that many people say that Houdini has an edge. I would like to know what they mean. For instance, in Maya, in order to get your particles to behave exactly the way you want, you need to control the system with expressions...that is, if you want complex behaviour and animation out of your particles. I read an opinion from somebody who prefers Houdini's particles: the reason they are better than Maya's particles is because they are easier to use due to far less expression writing...complex particle animation can be created right off the shelf in Houdini without writing nearly as many complex expressions as in Maya. Is this true? I am trying to understand this, and I am picturing the Houdini particle interface containing emitter tabs with thousands of extra attributes and way more option boxes than in Maya, that is, containing ready made expressions that one would need to write for themselves in Maya.

If anyone can shed some light, it would be much appreciated. Thanks, SM.

[TheDunadan]

I'm fairly new to Houdini as well, and know have some experience with Maya's particle system. Anyhow, one idea is wrong: Houdini isn't the program with 1000 of options in an emitter tab or alike, it's rather the opposite case, but combing all those Operators available you get an myrad of options. Quite a few represent typical "coding" structures and they are very general and therefore really flexible, they usually only care about the right data format (e.g. vector, float etc.).

Using the operator networks you kinda create code in a pretty visual way (this won't stop you from writing expressions here and there). Another thing that I like about Houdini, since these operator networks "represent" some data-flow chart you have very much control on when some operator will get evaluated.

The thing about Houdini that some love other likely hate is that it's never really guiding the user by narrowing him down in his options via long selection tabs or alike. I assume once you know Houdini well and you have a clear picture on what you want to create Houdini will get you there, otherwise you're not getting anywhere.

Jens

[kemijo]

I hate to continue the newbie fest as I am not even close to a very experienced Houdini person, but I'm a Maya dude and I might be able to shed some light.

As you said, Maya's main strength when it comes to particles is when using them in a per-particle fashion. Particle Expressions (and ramps, which is basically the same idea) are the only way to drive complex array attributes on particles and is very powerful, but of course is usually code heavy.

Houdini's POPs give you similar functionality right off the bat without code, in that pretty much all the parameters are already per-particle. Many of the calculations or functions you might use in a Maya particle expression has an associated node that handles this for you, and more. This is not to say that expressions are not used often - I find Houdini artists are expression whores (I mean that lovingly)

Another big one is that Maya particles are ALWAYS dynamic. Once the sim is working the way you want, Houdini particles are typically promoted to the SOP level where they can be manipulated as regular geometry. I find this to be where the difference becomes apparent, as particles and geometry become equal citizens. Many times a simple change needs to be made to the particles, and in Maya new functionality needs to be built into the particle expression to handle it, and the result re-simulated. In Houdini the operation is as straightforward as it would be with any geometry, like grouping or deleting based on rules, writing out to disk as a geo seq, deforming with any deformer (a new feature in Maya and XSI; it's inherent in Houdini). It can be treated as regular geometry, as the "arbitrary attributes per-point" in Houdini extend beyond particles and can be used on any geometry point. Maya would benefit greatly from arbitrary per-point attributes on meshes, even if expressions were the only way to control them.

What I find interesting is that particles seem to be Houdini's defining feature amongst non-users, where IMHO POPs benefits very little from Houdini's proceduralism. Most POPnets resemble a regular operator stack, and the true proceduralism is not really exploited until the SOP level. This leads me to believe that it's not Houdini's particles that are all-powerful...it's Houdini geometry and the way it's handled that is head and shoulders above most anything else. I won't even mention CHOPs and COPs and VOPs and...and... <sniff> it's been awhile

Anyway...not sure if that answers your question at all, but that's how I feel about POPs. If anyone wants to add more or enlighten me, please do...

[seanmac]

That makes sense, thanks. SM.

[edward]

I'm a newbie myself but the fireworks pop provides an example of a non-linear chain pop network. Go to pops, plop down a fireworks operator. Click on it and hit the enter key to dive down. I don't know how hard it is to create fireworks in other packages but its 10 nodes in Houdini.