PDA

View Full Version : Discussion of "Up, Forward, Right" nodal inputs? / use, node tools, etc

jeric_synergy
06-06-2014, 10:03 AM
I'm watching RH's "06- Nodal Foundations: Vectors" again (fourth time?) to specifically get a grip on the "Up/Forward/Right" orientation inputs.

I'm not to the end (again) yet, but IIRC these are rather sketchily covered, and I was wondering if the forum users knew of a resource that discussed their use in greater detail.

I seem to recall that there was some conversion nodes required before going to the destination node. (?)

Ryan Roye
06-06-2014, 11:15 AM
Right = X positive
Up = Y positive
Forward = Z positive

My guess in the mindset of the node setup for "item info" and related nodes is that they wanted to conserve space by only including 3 directions, expecting the user to know to use the "multiply vector" node or equivalent with negative values to achieve backward, down, and left directional inputs. Keep in mind that simply using an invert node will not work like you'd expect.

If the LW user manual covers this, I haven't been able to find it... but knowing how these inputs work makes taking advantage of raycasting and other directional functions so much less frustrating.

jeric_synergy
06-06-2014, 12:53 PM
Right = X positive
Up = Y positive
Forward = Z positive

My guess in the mindset of the node setup for "item info" and related nodes is that they wanted to conserve space by only including 3 directions, expecting the user to know to use the "multiply vector" node or equivalent with negative values to achieve backward, down, and left directional inputs.
Is that as simple as using "-1/-1/-1" as multiplicands?

:grumpy face: It's not like LW3dG couldn't make a "{node name} VERBOSE" version of nodes-- sometimes redundancy is perfectly valid. --That is, 'saving space' is a terrible reason to not have some conveniences in place.

Keep in mind that simply using an invert node will not work like you'd expect.
One thing I'm not sure LW users (but maybe it's just stupid me) are taking advantage of is SAVING Compound Nodes -- Jen H. demonstrates that several Compound Nodes are included in the 11.6 Content files-- or just saving Nodes in general. We could easily be exchanging complex node networks instead of screen grabs of same.

E.G., I loathe building those (very useful) "rainbow gradients" nodes, what a PITA that is, so I just saved one and re-use it all the time. It's part of my default "aaa_EmptyProjectDirectory" folder, along with various "Utility Objects" and "Utility Images". Compound Nodes are like saved networks, on steroids. --In fact, here it is: 122202

If the LW user manual covers this, I haven't been able to find it... but knowing how these inputs work makes taking advantage of raycasting and other directional functions so much less frustrating.
BeeVee?

See, in my dream-world of MODERATED USER-ENHANCED DOCUMENTATION, at the very least user could insert links to, for example, useful websites on trigonometry and vector math.

RebelHill
06-07-2014, 04:37 AM
Up right and forward represent rotation (as a rotation matrix) on an item... thats it, basically. Each one is a vector and thus, each describes a direction... those directions equate to the XYZ direction vectors of the given item. Ordinarily ofc, you're used to X and Y and Z each being a single number... but they can also be vectors as they each "point" in a given direction in world space ( so X = 1,0,0 Y= 0,1,0 and Z = 0,0,1 ).

As such, by getting direction vectors for axes and "sticking" them together, you describe the overall rotation of an item... which you can then convert to a set of rotation values (using the matrixHPB node).

You will also see the same labels on other nodes that take 3 vec inputs, such as the multiply/transpose/etc and other matrix tools... but here they're just labels, and you can simply think of them as equally meaning top/mid/bottom with reference to row vector inputs.

bazsa73
06-07-2014, 09:40 AM