Fumefx follow issue

Become a member of the CGSociety

Connect, Share, and Learn with our Large Growing CG Art Community. It's Free!

THREAD CLOSED
 
Thread Tools Search this Thread Display Modes
Old 01 January 2013   #1
Fumefx follow issue

Hello all
I have just been testing some fumefx work flows and I have run into this problem. So I have done a simple sim, saved, velocity, temperature and smoke channels. When I use a Fumefx birth and a Fumefx follow it works as shown on the left image. However its quite slow.

With the image on the right I am using a normal birth, position icon ,and and a fumefx follow and I am getting this odd behavior where its almost like its going in the opposite direction and small. I thought its supposed to work without fumefx birth.

Thanks.
Attached Images
File Type: jpg fume.jpg (88.5 KB, 22 views)
 
Old 01 January 2013   #2
FumeFX Follow only applies the velocity to particles that are inside the grid. Once they exit, they will continue to move with the last known velocity unless deleted (there is the option to delete particles that leave the grid).

I suspect that your regular birth is emitting at the edge of the grid, and it is quite possible that the velocities there are sending particles down and out of the grid. When using the FumeFX Birth, the particles are seeded at positions inside the simulation based on the channels thresholds, so they never end in such border regions, they are always inside a valid voxel.

I would suggest moving the PFlow emitter slighting up to be fully inside the simulation- if both the bottom of the grid and the emitter are exactly in the XY plane, it is possible to get strange results.

Of course, I could be wrong.

Also, when you say it is quite slow, can you qualify this in numbers? How many voxels, how many particles, how many frames and how long does it take?
I did a reflow of up to 100 million particles yesterday (emitting a million per frame from the surface of a mesh and advecting with FumeFX follow over 100 frames) and it took 30 minutes.

If you could have a commercial solution that does exactly the same, but faster and without PFlow involvement, how much faster would you expect it to be before considering it as an extension to your FumeFX workflow? (This is market research for Thinkbox, obviously )
__________________
Bobo
 
Old 01 January 2013   #3
Hey thanks for the response, but unfortunately it didn't work. I moved up the icon and resimmed but I am still getting the same behaviour. The only change is that now its position is inside the container but its still confined to one corner and its not close to the original motion.


"Also, when you say it is quite slow, can you qualify this in numbers? How many voxels, how many particles, how many frames and how long does it take?



Is there a script that I can use to spit out calculation times regarding pflows and loading times etc? When I say slow, with the Fumefx Birth even when emitting a mere 1500 particles I find the response a bit sluggish its not unworkable but it gets annoying, everytime I make a change or scrub through it take 2-4 second for the pflow to update. When using a normal birth this dosent happen , in this case even upto 10k its pretty responsive.

I did a reflow of up to 100 million particles yesterday (emitting a million per frame from the surface of a mesh and advecting with FumeFX follow over 100 frames) and it took 30 minutes. "

That sounds about rightbut you must have a real good machine. Is that to save or to make a preview? When I reflow from fume to krakatoa I have never done 100mill but it takes about 2 hours to do 6mil on my machine. Which is fine as I just set it and leave it running and once its in prt format its just smooth sailing, but if I wanted to make adjustment to the flow after fume like I want to do now it would probably be a nightmare. Using PRT fume is also quite fast it would be great to get some velocity variation in that, but i suppose its possible through magma.
 
Old 01 January 2013   #4
So I followed Bobos advice and made sure the pflow emitters were inside the grid. What iam trying to do is make the particles go along the green curve as shown in the image. In the original pflow which is being simmed I am using a speed by icon path constrained to the curve and I am getting this curving smoke motion. But I can't see to get the Fumefx follow to curve like (the green ticks) it just continues upward. If I just animate the speed by icon going up straight the fumefx follow works fine. I have tried incresing the velocity mulitplier but that just spreads them out more rather than have them follow the line.
Attached Images
File Type: jpg fumeB.jpg (86.7 KB, 19 views)
 
Old 01 January 2013   #5
Originally Posted by zoubein: So I followed Bobos advice and made sure the pflow emitters were inside the grid. What iam trying to do is make the particles go along the green curve as shown in the image. In the original pflow which is being simmed I am using a speed by icon path constrained to the curve and I am getting this curving smoke motion. But I can't see to get the Fumefx follow to curve like (the green ticks) it just continues upward. If I just animate the speed by icon going up straight the fumefx follow works fine. I have tried incresing the velocity mulitplier but that just spreads them out more rather than have them follow the line.


Because it's velocity field. Look at yours. I bet you don't have perfectly spline-shaped tanget vectors in it.
Here is example vel field:

As you can see while you have vel vectors following curve you also have portion of vectors pointing up at turns.
Think of vel field computation as a car racing - you can't just turn without inertia.
 
Old 01 January 2013   #6
Ooooooh Thanks it makes sense now, I was thinking it was as if it didnt save the positions . I will try to blend it with the speed by icon and see if that works.
 
Old 01 January 2013   #7
Thread automatically closed

This thread has been automatically closed as it remained inactive for 12 months. If you wish to continue the discussion, please create a new thread in the appropriate forum.
__________________
CGTalk Policy/Legalities
Note that as CGTalk Members, you agree to the terms and conditions of using this website.
 
Thread Closed share thread



Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

vB code is On
Smilies are On
[IMG] code is On
HTML code is Off
CGSociety
Society of Digital Artists
www.cgsociety.org

Powered by vBulletin
Copyright 2000 - 2006,
Jelsoft Enterprises Ltd.
Minimize Ads
Forum Jump
Miscellaneous

All times are GMT. The time now is 03:12 AM.


Powered by vBulletin
Copyright ©2000 - 2017, Jelsoft Enterprises Ltd.