Hello, I'm quite experienced in Quartz Composer but struggle to understand the fundamental concept behind the Houdini graph at all.

If there is anyone out there with an ok understanding of both Quartz Composer (or maybe max/msp) and Houdini? I think they could help me a lot and i would be very grateful.

For example, why aren't node inputs labeled in Houdini? How do you know what to connect to where? How do you know which outputs can be connected to which inputs - that is, what are there types? Say i create a Sphere primitive how would i connect another node output to the Sphere radius? Attributes - i don't understand what is actually going on if i, say, take the output of a Sphere primitive to the input of an Attribute Node. Is the Attribute added to the Sphere?

Many questions, i know. I don't expect them all answered. I just welcome any pointers on understanding the core functionality of the graph.

EDIT: For anyone unfamiliar with Quartz Composer:- Nodes have labeled inputs and outputs, each with a specific type (eg, Number, Transform, Geometry, etc). A Sphere node might have a Radius input and a Geometry Output. Any node with a number output, say an Oscillator could be connected to the radius input. The Output of the Sphere could be connected to any node input of type Geometry, say that of a Deform node.

This model isn't helping me to comprehend Houdini at all, actually it is hindering me and i can't find any good overview of how Houdini's nodegraph functions at all.


Thanks for reading.

Though both Houdini and max/msp use nodes - they differ in quite some aspects.
Eg: In Houdini nodes don't act as sliders/inputs them selves, but can have sliders - which are called spare parameters.

Means: you don't connect a slider to a node to modify the radius of a sphere, but you can read the value of a slider on a different node to modify the radius of a sphere - look up “channel referencing”.

In SOPs you pass the complete GDP - i think of this as the complete geometry - from node to node.

Probably best to watch some of the beginners tutorials.
There isn't hardly anything you can transfer 1:1 from Quartz/msp to Houdini - apart from: yeah - there are nodes …

Thankyou. I have done the beginners tutorials and i have a couple of books, i ‘get’ the tools and the gui - i just can't get a solid grasp of the nodes. What's GDP?

Have you done this one?
http://www.sidefx.com/index.php?option=com_content&task=view&id=1399&Itemid=261 [sidefx.com]


GDP is the internal representation of the geometry. Don't quote me on that, though.

Node labels aren't visually labeled in Houdini, but I believe you can hover over an input and get it's name.

In Houdini, nodes take in a specific type of data, operate on that data and output the same type of data. The operation on that data can be controlled by input parameters, but in Houdini those parameters are not patchable in the graph the way they are in QC. You create flow control and logic graphs in Houdini using chop networks which you access from any nodes parameter using an expression.