New tutorial added:
Object information data delay filter
Credit goes to Sparky for this one, I just drew it up. Used for making an object follow the same path as another object with a slight delay.
thanks m8, just corrected, this is because the increment and test was being done after the OPE was calculated, making the final offset value reset before the OPE could use it.Originally Posted by Blastergamer
nice, this is the same sort of method snake games useOriginally Posted by JaminBen007![]()
@nannerdw:
your single axis facing tutorial is horizontal axis, so if I want it to veritcal, what should I change?
I tried mixing it with X/Y, Y/X, Y/Z, Z/Y with the OIDS, but it didn't work.
edit: Now I used your two axis tutorial and I wonder why the vertical axis needs to be local?
PS. @dasraiser: It worked very smooth and nice with your suggestion!
I'm gonna make a tutorial for this!
I agree with sparkierjonesz about the vertical movement. I'd just use a glued object so that the axes line up with what's shown in that tutorial.Originally Posted by Blastergamer
Otherwise, I think you'd have to change the object position events from "Pitch" and "Yaw" to something else depending on how the object is oriented. Also, you might have to remove the Negative operator from the Y axis if the up/down movement ends up getting reversed.
The way the axes in my tutorial are set up, it's basically like this:
Global Yaw = Object sitting on a spinning platter. Global yaw is the platter's rotation angle.
Local Pitch = The object tilts away from the platter's center as it starts to fall off, but it's still turning with the platter. Pitch is local, because its axis of rotation is turning along with the platter.
Local Bank = The object is doing cartwheels while still rotating around the platter, like this.
Don't think of the platter as being turned on its side, but rather the object is oriented differently on top of the platter.
You'd only have to mess with the arctan2 operators if the object is moving like in the "Local Bank" example.
When you glue objects together you have two options, usually you would highlight all the objects then select the "glue" option to stick them together.Originally Posted by Blastergamer
The other way is to spawn a "glue" icon ( I think it's in the common tab under physics, near where you find the physics joint) and stick items to it.
Then what you can do, is align the glue joint to the axis you need, stick stuff to it and the axis will stay where they are, so you don't need to change any logic.
OK,
I think I understand now.
Nannerdw, can you add to your single axis OPE tutorial the way, if I want it on vertical axis please.I have tried out to make it to vertical with the OIDSs and Yaw->Pitch but I couldn't get it right.
btw, instead of horizontal axis, you described vertical! That's an issue!![]()
I just want to make sure I understand this correctly. By "vertical" you mean like a gun pointing at an airplane as it passes overhead? That's going to require a new method, I think, so that the gun won't spin around 180 degrees when it passes vertical. Is the object that you're rotating aligned with the global xy or zy plane, or is it positioned at an odd angle?Originally Posted by Blastergamer
I thought of a good way to rotate an object at any arbitrary angle with vectors, but I'm still having trouble figuring out how angles work in the editor. There are data sources for angle, direction, side, and up, but I'm not sure what these actually represent.
It says "Direction, side, and up are the object's orientation vectors." Is that like some common notation in 3d graphics? "X, Y, and Z" would make way more sense, IMO.
I think it might be something like this:
[code:1lft7k05] side(x)-> [xx, xy, xz;
up(y)-> yx, yy, yz;
direction(z)-> zx, zy, zz][/code:1lft7k05]
Up is definitely Y, but I think I might have Side and Direction mixed up.
I is confused
Does anybody understand how to use rotation matrices? The fact that x and z are horizontal and y is vertical are really throwing me off, lol.
hey nannerdw
the Up, Direction, Side are the objects normal vectors to the world axis, in terms of computer graphics these are handy for simple lighting, if a surface is visible (eye to object vector dot normal<0) or calculating the distance to the object surface plane and so much more.
in trials these represent the world rotational cosines to the world planes XY, XZ and YZ.
i think:
Direction is on the YZ
Up is XZ
Side is XY
unscrambling these to form useful rotations in trials is quite difficult to do. i mostly use a simplified method, one dayi may figure a nice method to find the full rotations.
totally agree.It says "Direction, side, and up are the object's orientation vectors." Is that like some common notation in 3d graphics? "X, Y, and Z" would make way more sense, IMO.