retime results in popping particles

Post Reply
ctjanney
Posts: 4
Joined: Fri Apr 29, 2011 3:07 am

retime results in popping particles

Post by ctjanney » Wed Mar 14, 2012 2:35 am

realflow 2012, windows 7, i7 6core, 24GB ram

I have 10 frames solved at 30fps that I then have retimed at a factor of 0.015 (2000fps). The retimed particles pop

so the original animation is 10 frames, and the ratio I'm using is 0.015 (30fps/2000fps), which gives me a new range of 0 to 645, and the pops happen on every 66th-ish frame, so it's the end frame interpolation that is causing the issue. I've tried kill on first, kill on middle and kill on last, all popping.

Suggestions, my fellow water movers?

-ctj

User avatar
FlorianK
Posts: 238
Joined: Fri Jun 11, 2010 9:56 am
Location: Cologne, Germany

Re: retime results in popping particles

Post by FlorianK » Wed Mar 14, 2012 7:56 pm

Only bad news. I remember having a similar issue, but I blamed Sheeter not Retimer (I was also using retimer in that scene). Unfortunatley, you don't seem alone with that problem. The guys should look into that.

User avatar
Alex
Site Admin
Posts: 143
Joined: Mon Jun 14, 2010 9:35 am

Re: retime results in popping particles

Post by Alex » Thu Mar 15, 2012 2:10 pm

Hi there,

We'd have to see the effect to confirm if it is related to particles birth or death.

Particles birth is not managed by the current version of the Retimer. If you have an emitter creating particles every frame and then you make it 10 times slower, the retimed particles will appear once every 10 frames.

Fixing this to make particles appear at their "right" time would suppose an overhaul of at least twice as much memory and time. But we understand that in cases when quality is mandatory and particles source is visible it might be worth the effort.

An extra issue for that case is related to the "where" to place the particle. For example, if a particle explodes right after being created (i.e. the first time it is recorded in the bin has a big velocity) the particle birth position will be assumed to be very far and at the opposite way.
That would be a different issue that could be somehow detected and "clamped". Otherwise expect some sort of a Star Wars interstellar rain of gunshots in that case.
Alex Ribao
RealFlow Team
Next Limit Technologies

Post Reply