Im doing a simple dissolve setup with particles piping into a fume channel via an age test.
All looks fine and renders well directly from pflow, but when i have saved the sequence out to disk, upon reloading with a PRT loaded - the results are incorrect, like its missing out parts of the animation.
I have no idea whats happening with this one. All i can think of is that the particle age may be causing trouble with IDs? And also i have offset my fumefx source start frame. woudl this screw it up?
When you say “missing out parts of the animation”, some of the frames are missing or do you mean that some chunks of where you expected there to be particles there aren’t any?
In the case of the latter, I’ve had minor freak outs similar to that and then realized that the viewport settings were set to load the first 1% which will show only a small piece (ie. not evenly spaced through the particle set to make a sparse cloud like normally).
If that isn’t it, can you post some screenshots to describe the issue further?
yes, i checked all the display options and upon render - results are incorect.
i actually solved the issue (though im not sure why this has caused the error) - as i have two position objects in my flow, one and instance of the other one. Upon particles being sent to the next event via a colliosn detection, they remain where they are still via the instanced position object…then are age tested to follow a FFX. I replaced the position object instance with alock bond and all is well/works as expected. So it appears the ID’s or positions are altered upon looking at the second position object. Though i cant see how, as its an instance of the first one.
Posting step-by-step explanations instead of large paragraphs makes it much easier to follow and debug.
Also, posting a screenshot of the PFlow is worth over a 1000 words
There are certain limitations to what you can do with the Krakatoa PRT Birth/Update. Also, you always need an Update in each event or in a global one to keep track of particle birth/death and refresh all data channels, and a Krakatoa ID Test to kill particles that have died in the original stream. Without looking at the flow, it is hard to tell if any of these rules have been violated…
yup - heres a screen grab. Its currently an extremely painful process to copy and paste a screen grab at the studio im working at! lots of restrictions and locks on the terminals and im running windows and max in VMware. good eh?
so heres an image. simple stuff as you can see. The second instanced position object seemed to be causing trouble.
Thanks for the image, I had completely misunderstood your initial post
Images = Good!
It is quite interesting that the PFlow would render correctly but not save correctly. Was the rendering of the PFlow done in Krakatoa or in Scanline?
Also, what is that second Position for? You can switch the Collision to “Continue” to prevent it from moving the particles if you are trying to just keep them in place and use it as a simple trigger for sending out to the next event…
The second pos. object is to keep the particles attached to the object still, as the object is animated travelling around. I still want the particles to remain on the object after the collision test, otherwise they are left behind.
Lock bond solved the issue - but i was interested as to why the partitioning was having problems with the second posobject.
