Particle Deflectors not working the way I think they should.

Hi.  So my issue is this.  I've essentially created a box out of five deflectors (bottom,back, front and two sides) that holds the particles that have fallen from above it (using a 'force' for gravity).  When the box is filled to my satisfaction, it is my intention to tip the box and have the particles fall out.  All of that is fine and it essentially works except for one thing: when I move my box horizontally (no tipping) all of the particles fall off (not through) my bottom and go through the side that that is opposite of the movement.  I made sure that all sides of the box are 'going through' one another so there are no gaps.  I guess my problem as I see it, the side is not acting as a deflector one the box is on the move.  Any thoughts?  I will include some images once I figure out a place to store them online for this purpose if it helps to visualize what I am saying.

Thanks loads,

Adam Raitano


  • NormanPCNNormanPCN Website User Posts: 3,950 Enthusiast is an easy way to upload images and get an image  link to use in a forum post. They even have a tray app to capture screen shots.

  • adamraitanoadamraitano Website User Posts: 62

    Thanks @NormanPCN

    Here are the images:

    The top on is when the box is spilling (not exactly a perfect box but there are no gaps in the seams).  At this point on the timeline it have move horizontally to the right.  The bottom one is after the box has been filled with the particles and the emitter is no longer active. It should also be noted that the angle here is due to the camera position and not the box's location in 3D space.  It is perfectly level (x rotation of the bottom green layer at 90 degrees)

  • adamraitanoadamraitano Website User Posts: 62
    edited October 2017

    Here is a different angle.  You can see that there are no seams and that the particles are moving right between the blue and green layers.  Strange.

  • Triem23Triem23 Moderator Moderator, Website User, Ambassador, Imerge Beta Tester, HitFilm Beta Tester Posts: 18,307 Ambassador

    Can you upload your project? I just did this sort of thing myself--build a box of plane layers and stuck an emitter inside. When I move and rotate my box I'm not getting any unwanted spill. Deflectors working as desired.

    Here's a link to what I built. This is Hitfilm Pro 2017. The project should open about 4-5 seconds into the animation. All my planes are parented to a "Box Point," so if you move and rotate the Box Point you'll see that all the particles are following the box perfectly. Rotating the box lets the particles spill out. Box.hfp?dl=0

    So I guess I'll want to look at your project, since I wasn't able to replicate the problem.


  • adamraitanoadamraitano Website User Posts: 62

    Hey @Triem23.  Thanks for posting that.  I will save you the trouble of having to look at my project since I was able to replicate it in yours.  Try this:  move the Box Point X position from 250 starting at  0:01 to 1050 at 0:02.  I think this illustrates the issue I am having.  At least it does so on my machine.

  • tddavistddavis Moderator, Website User Posts: 4,192 Moderator

    @Triem23 Hey, just wanted to let you know I downloaded your project to play with know, I'm sooooo weak on particle systems and not just Hitfilm; all programs.

  • PalaconoPalacono Website User, HitFilm Beta Tester Posts: 3,442 Enthusiast

    Could it be that the step size of the box's movement is bigger than the particles and they get left behind somehow?

    It used to be with early video games that cars could go through walls, bullets pass through characters etc. if the speed was such that steps were large enough for there to be actual gaps between updates.

    In this simulation, God has a much faster frame rate, so doesn't suffer from that issue.

  • Triem23Triem23 Moderator Moderator, Website User, Ambassador, Imerge Beta Tester, HitFilm Beta Tester Posts: 18,307 Ambassador

    I'm on day one of a 4-day 14 hour/day production job, so I probably won't be able to directly look at this for a while.

    Try building a box rig using cuboid forces. Annoyingly these have to be manually keyframed.

    What I'm thinking is position might not be computed on subframes, and the infinitely thin plane might be "jumping past" the particles. I'm wondering if a cube would catch it.

    Other thing to try--increase frame rate of the comp. If it's 30fps try 60 or 90. This will increase the temporal resolution and, hopefully keep the planes from jumping past the particles.

    If you up the framerate of this comp then embed this comp in another comp that's at your export framerate. That will conform your final render correctly, but give finer control of your particle sim.

    Those are my thoughts without being on a computer. 

  • adamraitanoadamraitano Website User Posts: 62

    I will give those ideas a try.  Thanks, @Triem23.  As usual, your help is invaluable.

Sign In or Register to comment.