Okay
Please read this with an open mind.
I read the thread on what houdini has to offer me.
I accept that the modeling tools are some of the best in the business.
But as I start to explore beyond modeling on my own I find myself lost.
The Particles,Vex,H script are not the most userfriendly.
I realise that I should do some of the video tutorials,but eventhough I have downloaded them all I find them very limited.
They deal with a very few aspects of Houdini.
So for somebody like me who has no coding knowledge
is houdini worth the buy ie can I pick up the scripting and other technical parts of Houdini with a bit of hardwork(the biggest problem I face,since I come from max, is that I either cannot find the method to do something I visualize or if I have to code it myself ,I am at a loss words(no pun intended)) ?


As always
Cheers
Lanchka

I would suggest you avoid the areas you mentioned for now. By definition, hscript and VEX are programming script/languages. You do not need to program in order to animate in Houdini…the power is there should you want to use it, but there has been a lot of work done to ensure that you can do everything you need in the GUI. VEX is only for advanced users that wish to expand the power of Houdini, and hscript is really a production tool that's used for various management processes, like batch rendering and clever macros.

Have you read through the online help? I would take that step by step. There is no question that due to it's powerful proceduralism, there are some new concepts. Try to work through the examples. I'm personally not aware of any basic training that requires you to use VEX or hscript, except for training in those areas. Try not to do it all at once…it's a complicated application(and topic!).

Cheers,

J.C.

Actually I come from the aspect that I use Houdini because I am more of an artist than a programmer. I feel Houdini's way of letting me choose which operators to use in which way I want allow me to visually do anything I want without coding anything. Heck, even Vex can be done in VOPS and not written. Can you take the coding side of Houdini and use it? Of course. Do you have to? I don't think you do (some may disagree tho). The only kind of written things i do are expressions because I find them to be apart of the whole procedural realm of things rather than writing code. I can't think of anything I've done in Houdini that didn't involve an expression. But still, you don't need to use them. You can animate your heart out if you want, I just think it's faster the other way for a lot of things that I do. And on occasion, I'll set keyframes here and there.

But anyways, IMHO, Houdini may seem to be completely different than other packages and therefor come off as “non user-friendly”, but I find it more user-friendly than everything else because, to me, it all makes sense. It's all structured and therefor intuitive and logical. I rarely get lost when learning something new in houdini (whether it be in a tutorial or not) because I know it's all there for me to dissect and i know the exact order of operations that has led me to that point. It may take a little bit of time to figure it out, but in the end, it will come together. I now find myself in other packages and feel like it's just sloppy to work in because I feel like I don't have complete control over the whole process of whats occuring.( And if the control is there, it's dug somewhere deep.)

So yea, stick with Houdini, get used to it's principles and you'll never turn back. It is intuitive and user-friendly once you understand those main principles. It's not only for programmers. And it's extremely flexible. Forget the code for now and come back later if you want to. Get used to everything else and see how you feel.

hth,
dave

I also come from traditional software, and although I still start to learn houdini. I seem to me the way a 3d program should be. I think I'll probably buy it once I know it enough to use it into production.
There are still things that could be improved if you compare it to other programs, but isn't that for all of them?

I'm also more a artist then a technical guy, the nodes system is understandable. But you really need to read the help file because it is really different from the “traditional” softwares.

B.t.w. This brings me up to a question I did not find yet a good tutorial or reference for. Im used to work with lightwave a lot for character moddeling. One of it's very handy tools is the fact that you can have falloffs, (weight selection or softselections in other programs) for moving scaling and rotation.
What I would like to do is to use for example an alpha of an image to weight select vertices and to transform them with the weight as a multiplier for the transform.

I assume you've tried the “Soft Transform” sop and “Soft Peak” sop as well as the edit sop, they all allow a falloff. It's not a complete solution but is good for lots of modelling ops.
I've also added a “Soft Ray” sop (available on the Exchange site)

There are other ways to do stuff with soft falloffs like using a image but they aren't built into the program by default. If you describe exactly what you want I'm sure I (or someone else) could make an otl to do it for you.

Simon
There are other ways to do stuff with soft falloffs like using a image but they aren't built into the program by default. If you describe exactly what you want I'm sure I (or someone else) could make an otl to do it for you.

Wow thanks Simon, I hope this can be done.
I think it would be a good extension for the group sop that could have an object selecting vertices by projecting a mask image along it's normals
So for example you would create a grid sop have a uv sop and apply a shader. And the intensity values of the shader are used to make a soft or weighted selection.

It's easy enough to do per-point blending using a painted attribute either the Point SOP or a tiny VOP network.

It's easy enough to do per-point blending using a painted attribute either the Point SOP or a tiny VOP network.

Agreed. I was trying to wrap my head around why you would want to even go the shader route. Seems like a lot more than what's needed. Paint sop and per-point attributes work beautifully.

The only thing I'm unsure of is the group sops ability to have weighted selections (which is why you would do it through attributes im assuming). If there is a way of grouping in that manner, i'd like to know how…

edward
It's easy enough to do per-point blending using a painted attribute either the Point SOP or a tiny VOP network.

Yes for static things this is usefull. But since Houdini is especially handy for doing procedural stuff this would come in very handy. You would be able to do the vex mountain stuff with just creating a softselection and moving the vertices along their normals. And if you animate the textures you can create the most weirdest effects easily wether you have an image sequence or a procedural shader.

You can easily do this just with a cop network, to read in your image sequence, and a point sop with the picni expression in the $TY field.

so

Create a grid
Add a UVtexture sop
Create a cop network with your image in it
Add a point sop to your grid and uvtexture sops
replace $TY with the following

picni(“/img/img1/color1”, $MAPU, $MAPV, D_CLUM)

make sure you edit the path to make where you have your copnwork and file cop.

Simon
You can easily do this just with a cop network, to read in your image sequence, and a point sop with the picni expression in the $TY field.

so

Create a grid
Add a UVtexture sop
Create a cop network with your image in it
Add a point sop to your grid and uvtexture sops
replace $TY with the following

picni(“/img/img1/color1”, $MAPU, $MAPV, D_CLUM)

make sure you edit the path to make where you have your copnwork and file cop.

Thanks Simon. This is indeed a specific example of what I would like to do with it. But what I would like to find out is how I can connect this to any parameter.
So it is like an open tool. for example that I can control the polybevelSOP relative inset. with the same parameter. So that the image controles for each point how wide the inset will be. I tried this expression directly on the relative inset but this does not work. Do you have an Idea how to make this more open to all parameters?

Just because almost all the nodes are able to read group selections i though it might be very usefull to allow groups to have soft selections. Like that one could be able to use an image, or something else to to control individual vertices or primitives for almost any sop.

In max you have a volume select modifier that does exactly what I discribe here. And it is very usefull. Only problem is that 3dsmas is not by far as open as Houdini is.

couple of comments/thoughts on this:

first, Simon: where the heck is “D_CLUM” coming from. I know it's supposed to be the color_type in the expression, but I can't find anything on D_CLUM or any other “color_types” to put in here.. if you could elaborate on that, I'd really appreciate it.

Now on to the soft selection discussion:

Operators like the polybevel don't support per point procedures.. only groups of points/prims/etc can be used for the operator to enact upon as a whole. (for example, you can't put $PT in to relative inset and expect the point number to act as the amount of beveling for that point. it would be cool, but it just doesn't support it for reasons i don't know.. can anyone else explain..?) if you wanted different beveling for different points/edges, you would have to use several polybevel's and each one would have their own point/edge selections with their own bevel settings/amounts..

so going on the idea that group selections should be weighted.. knowing that operators use groups as a basis to perform their procedures as a whole, then how should it deal with weighted groups? what does the operator base it on? so you have a group and Point 0 has a weight of .5. Since the operator uses the group as a whole, what parameter does it append “.5” to? You could easily say, well just have the operator do what it does at half the amount. Well, ok. But what if the operator's doing something that can't just “halve the solution”? For example, the poly bevel sees Point 0 as being a .5 weight. Does it append that weight to the relative inset, or the repetitions, or what about the bevel type? How can you have .5 repetitions? Or more drastic, a half Flat bevel type? This is where I can see things becoming an issue…

Don't get me wrong, I'm all for soft selection groups and more per-point operations, but it seems a lot more complicated than i can think of a solution for. As being somewhat young and not advanced as many of the people around here, I could be completely wrong in my thoughts, so I am definitely interested in what others have to say about it…


dave

is houdini worth the buy ie can I pick up the scripting and other technical parts of Houdini with a bit of hardwork(the biggest problem I face,since I come from max, is that I either cannot find the method to do something I visualize or if I have to code it myself ,I am at a loss words(no pun intended)) ?

Going back to the original question, I too come from a traditional background. My coding knowledge was limited, but I was able to get around houdini fine. The truth is that you can get plenty done with Houdini wihtout coding. However, what I have learned is that your life would be easier if you are familiar with the logic of computer programming and have an understading of math, specifically linear algebra and some trig.

When you want to become a more advanced user, you will realize you need to understand such things as vector mathematics. vectors play a big part in Houdini as far as tranfer of values or what not, like in attributes. If you want to build shaders or lighting networks, then you definitly need to know vector math to some level. Also, if you want to get into VOPs down the road, you will have to learn about for loops, if loops and quaternions. These are programming and math concepts which are prevalent in VOPs, albeit visually. Some people can pick up this stuff purely by playing with Houdini, but others with programming background VOPs is much easier to pick up.

But in reality, if you are willing to learn, then your options are much bigger. My 2 cents. 8)