I think the documentation about constraints [www.sidefx.com] is not clear, informative enough, and at times it's even cryptic.

One example : “Parent Blend will blend the transform of the current parent with other objects and apply the current local object local transform on top.” HUH? What do I make of this?

The doc provides more questions than answers. For example:
- Why is the constraining object called sometimes a “parent”, sometimes a “target”? Both are confusing.
- The blend constrain can add an offset by “keeping position”. It also gives the choice to keep this position before or after, what's the difference?
- What's the real difference between Blend and parent blend?
- What's the Maya or XSI equivalent for Blend, parent blend, point, surface, transform constraints ??
And I have more questions that I didn't write down yet.

So if there is some Youtube/Vimeo or paid video about Houdini contraints, it would be welcomed. I've been searching unsuccesfully so far, HELP!

Thanks
David

Hi, David,

I agree that Houdini's documentation - although much better than other program's docs - leaves a lot to be desired both for a newbie (as which I see myself as well) and a professional. Houdini seems to not only have grown over time, it also seems to have out-grown the docs-keepers and is always two releases ahead of them …

That said, for me the best approach to make sense of both docs and how the program actually acts (which often does not seem to exist in the same universe) is to create a simple enough scene that even I can understand it. Then tweak one parameter at a time and fully digest what it does.

Maybe I can help with parents-versus-targets, at least to give you a starting point: A “parent” is a node which transforms put a “child” node into a local-space (the child node lives in the parent's node's space). Since by using targets that live may live in their own space OR IN WORLD SPACE, you run into situations where you have to “blend” between “spaces”. Depending on your setup it is thinkable that e.g. your child object is getting its rotation both from its parent (local space depending rotation) and from a constraint that uses world space (by transforming the target's transform into world space and transforming from there into parent-node-space).
Again, the best approach to me remains to keep things as simple as possible and understand each component's dependencies. Then, step by step, make things worse …


Marc

Thanks Marc,

Well I created a scene with a parent blend constraint, a blend contraint with “keep position” = after, and a blend contraint with “keep position” = before. I don't see any difference in the behavior of the constrained object.
I find it unfortunate that I have to build scenes and use buttons to understand how things work. It's like I have to do a lot of research to understand. It is not straight forward. The best solution for Hou users would be to get a documentation improvement.

In every 3d app there is a difference between a parent and a constraining object:
- A parent is just a link between a parent and a child, so that the child will follow the parent. There can be only one parent.
- A contraining object can act on one of these hierarchy object, either a parent or a child. The constraining object will have authority over the parent-child relationship. Each constrained object can have multiple constraints, thus multiple constraining objects.

The constrained object is the target of the constraining object.
In Houdini, the target IS THE CONSTRAINING OBJECT which is much confusing.