• 0 Vote(s) - 0 Average
  • 1
  • 2
  • 3
  • 4
  • 5
p Flow follow op not accurate
#1
Hello,

I am using Fume to sim movement, then P flow to create parts.
but ...

p Flow follow op is not getting very accurate results. (parts moving at near 1/2 speed with little shape.

It seems as i increased the rate of movement in fume, i got much worse results.

I had much more accurate results with other setups, including earlier versions of the current set up I am working on.

any ideas
sam
  Reply
#2
my idea would be use Tp Wink i havent tested much the pflow op .. i gave u the only scene i had pflow in it lol
<!-- w --><a class="postlink" href="http://www.cgfluids.com">http://www.cgfluids.com</a><!-- w -->
  Reply
#3
yeah well, thats why I ask here, mr TP dude Big Grin


I imagine there would be a speed diff between
P Flow an TP Fume interaction

But what about accuracy between the 2 apps?

specifcly the follow op
sam
  Reply
#4
Hello,

Actually, it should not matter what particle system you use - operators for both particle system should work correctly.
For follow operator, make sure that Velocity Multiplier is 1.0 and Influence % is 100 if you want them to completely follow velocity field (make sure that FumeFX exports velocities!).
If you're still getting wrong results, please email me scene to mailto:kresimir@afterworks.com so that we can take a look.

Regards,

Kresimir
  Reply
#5
the problem seems to occur when u increase the time scale

the set up works just fine at a time scale of 1, but at a time scale of 5 the partilces move at near 1/4 speed with lil movement.


I will try and strip down a file to send u, heavy NDA on this one.




cheers
sam
  Reply
#6
time scale influences parameters like velocity multiplier. if u double the time scale, the velocity will be twice as strong.
VFX TD rsfx.de
CG Supervisor @ ScanlineVFX Munich
  Reply
#7
ok that seems reasonable.

So altering time scale will correspondingly alter particle movement
in a different way than the voxel grid?


thx
sam
  Reply
#8
the time scale parameter influences only the sim of course. if u alter this parameter, the sim produces a different kind of velocity vectors and ergo will move ur particles different. it wont initially change the values of ur settings in pflow. those settings have to be adjusted manually, to get similiar movement for instance.
VFX TD rsfx.de
CG Supervisor @ ScanlineVFX Munich
  Reply
#9
so does the time scale not effect the simulation, but the play back rate of the sim?

or the other way around?

cheers
sam
  Reply
#10
i dont recall having that probleme in tp .. we need to check that i think..

The info that Roman said is true for "In SImulation " behaviors and tweek .. if we have changing timescale , i dont know why the velocity Cached output would change to not Match the new sim basicly ( cause that is what would be wrong if that happened and would give wrong mblur as well ??.. )
<!-- w --><a class="postlink" href="http://www.cgfluids.com">http://www.cgfluids.com</a><!-- w -->
  Reply
#11
Hi,

just in case there is still something unclear:
time scale of X means that after 1 frame of simulation, fumefx will produce the result that would normally happen after X frames. So simulation will appear fast, but the exported velocities will still be the same in magnitude as you would get with timescale 1 and after X frames.

So you should set the velocity multiplier in follow operator to X also.

Vjeko
  Reply
#12
ah good stuff,

I will test that.
  Reply


Forum Jump:


Users browsing this thread: 1 Guest(s)