Hi,

I've found this very old thread
http://www.sidefx.com/index.php?option=com_forum&Itemid=172&page=viewtopic&t=13&highlight=lsystem+motion+blur [sidefx.com]

in which Jeff states :

“When point attributes are created, the system's points are
automatically sorted by generation. This means point numbers
will match from generation to generation, allowing motion blur,
etc. ”

How is this helping for motion blur ?
Since the point's count is changing, I still get wild results on new points, when using a trail sop to compute velocities.

Am I missing something there ?
Thanks a lot for clearing things up.

Maybe you can create your own velocity attribute, using the info that l-systems saves in the points, use these velocity attributes to make velocity MB.
Hope this helps

Thanks for the answer.
Well, that's what I tried to do, but I still got weird results.

But I just noticed that the points were correctly sorted only for round generations, 1,2,3,4… not for fractional ones, like 1.1, 1.2, 2.5… so that explains it.
Noticed also that when I manually add a “generations number” attribute, using the t variable, the points which belonged to these fractional generations, had only a fractional part.
For example, I'm at generation 4.3, the last points have a t value of .3, instead of the expected 4.3, which would explain the wrong sorting, if houdini refers to this internally.

So, I got it more or less working, sorting the points using a custom attribute, however the frames where new points appear cause problem, but I guess not much can be done about this.

Maybe you can sorte the points relative to the first point of your plant using the sort SOP, maybe this will give you a more consistent sorting brecause usually this first point never change it's position.
Only guessing …..

this is another good case where persistent unique ids
per point would be useful.

lisux
Maybe you can sorte the points relative to the first point of your plant using the sort SOP, maybe this will give you a more consistent sorting brecause usually this first point never change it's position.
Only guessing …..

I did sort them using the integer part of the generations attribute, so the sorting is ok.

Only thing is that when new points appear I can't use their previous position to calculate their velocity since they don't have any! But that's inherent to the lsystem.