PDA

View Full Version : Particle lag: multi-parented emitters? Pilot error? Something else?



jeric_synergy
10-03-2015, 10:22 PM
Here's a scene with particles, attached to a spirograph rig:


130157

Go ahead and let it play a second or two, and you'll see that the particles lag the emitter's position. They are obviously related to the emitter, but lag behind by a significant time.

Why would that be?

Thomas Leitner
10-04-2015, 02:41 AM
Hi,
it seems that they donīt lag behind in time, they are created in the wrong place. If you set the emitter nozzle to e.g. Object-Vertices it works. Maybe the problem comes from the way you set up your pivot points (if you parent a null to the cone object it is in the wrong place too).

ciao
Thomas

MonroePoteet
10-04-2015, 07:03 AM
Yes, as Thomas said, the particles don't lag in time, but are offset from the geometry. By using the nozzle type of "Sphere", the particles are emitted from the pivot point. If you either center the geometry of the pieces / layers in Modeler, or move their pivot points to their geometric center in Layout, or use a different nozzle type, it works correctly.

mTp

jeric_synergy
10-04-2015, 09:32 AM
Thanks, guys, will examine. :thumbsup:

jeric_synergy
10-04-2015, 08:26 PM
Well, Monroe, I'm not sure about that last part, because the pivot point WAS moved to the geometric center of the cone, and that's where the emitter appears (as expected) but the particles actually emerge from the UN-transformed location of the pivot point, i.e. the original pivot point.

I'm going to submit this as a bug, because no matter what, particles SHOULD emerge from their emitter. It looks like the emitter-drawing code is working and taking the pivot offset into account, while the actual particle emitting-code is not.

Thanks for the assist, fellows, I'm off to earn bug-hunting points!!! (Not a real thing.)

++++++
EDIT: Shouldn't children of an object always be relative to the USER'S defined Pivot Point, not the original LWM 0/0/0??? Perhaps someone can tell me about the subtleties of Pivot Point alteration.

jeric_synergy
10-05-2015, 11:12 AM
Rebel Hill, please weigh in:
LW3dG confirms this situation, and we're talking about the implications: FWIW, I always assumed that moving the Pivot Point, using the MOVE PIVOT TOOL, would result in children that related to the Pivot in the usual manner, e.g. if at 0/0/0 relative to their parent, would be right on top of the pivot. However, this situation is of long-standing, dating back to LW9.0.

The conundrum for LW3dG is: would fixing this 1) be avoidable (i.e,ignore it), 2) be desirable, and 3) would it break thousands of existing scenes?

My suggestion, and unfortunately it is huge, is that a new motion option be implemented (sorry for delaying 2016, guys!), and the default set to preserve existing scenes. The option could be of the form:


" [_] Respect Pivot Alterations",
or conversely
" [_] Use un-altered Pivot"
I think this would be the proper way to handle it, but I appreciate the magnitude of the change.

As the foremost expert on LW hierarchical motions, I'd like to hear what RH has to say-- and I'm prepped to find out I'm totally thinking about this wrong.

Of course, ignoring it is viable too. However, the location of the emitter icon could be reset to coincide with where the particles emerge.