Community
Bifrost Forum
Welcome to the Bifrost Forum. This is the place for artists using Bifrost to ask and answer questions, browse popular topics, and share knowledge about creating effects procedurally using Bifrost. You can also visit the Bifrost Community on AREA to download an array of ready-to-use graphs, read Bifrost news and updates, and find the latest tutorials.
cancel
Showing results for 
Show  only  | Search instead for 
Did you mean: 

Try using Bifrost's particle system to make a magic effect

9 REPLIES 9
Reply
Message 1 of 10
Bruce.Lee.
1048 Views, 9 Replies

Try using Bifrost's particle system to make a magic effect

Hi,

When making a magic effect with Bifrost's particle system, I found that the particle system does not release memory. In the process of production, when the memory is full, maya will crash. There have been many crashes.

Particle systems still lack a lot of basic facilities, but I believe that particle systems should be perfect soon.

9 REPLIES 9
Message 2 of 10
Bruce.Lee.
in reply to: Bruce.Lee.

In my opinion, turbulent noise is very useful for making programmatic effects. Bifrost should have more different types of turbulence noise nodes.

Bifrost's turbulence field is too simple. Houdini has many different types of turbulence noise nodes.

I hope Bifrost's second episode can add more turbulence noise nodes 🙂

Message 3 of 10
mjcg91
in reply to: Bruce.Lee.

We could implement some by ourself I believe, but the main problem with custom influence is that if one wants to implement one, the user has to make the said solver editable first, and then dig deep inside of it in order to add some customs effects. It works really well, but this is not user friendly at all, and for distribution it is even worse from a user stand point.

 

The thing is, you can't actually build a compound and use it like any other influence, because all the influence compounds does are settings a bunch of properties, which are then used deep inside the solver, where the actual influence computation is done. So, custom influences compound would require to be be distributed along the modified solver as well.

 

It would be nice if there was a mechanism that would allow a pre-solver influence compound to be computed inside the actual solver. For example, you make your math logic inside a compound, and that part of the graph would be computed where it's supposed to be. It would be quit convenient for both distribution and usability.

Maxime Jeanmougin - Technical Artist
https://maximejeanmougin.com

Join the Bifrost Addicts community on Discord:
https://discord.gg/bifrost-addicts
Message 4 of 10
marcus.nordenstam
in reply to: mjcg91

 

Totally hear you and we have some things coming to address this.  The first one that will show up are the new fields system.  Then we have some more hard-core engineering solutions for this as well, further out.

 

We will start talking about the new fields system post Episode I launch.  Stay tuned 🙂

 

Message 5 of 10
Bruce.Lee.
in reply to: mjcg91

I used to sneak into the AERO solver to modify the field solver, and then made a few custom influence fields. This approach is really cumbersome. However, Bifrost can now create simple custom fields without diving into the solver. It is to use VDB field or volume field.

I remember a long time ago, Duncan said that Bifrost will eventually have a thing called a deferred execution graph that allows the user to make any custom tools outside of the solver.

Message 6 of 10

Wow! Marcus, I am looking forward to the next update! 🙂

Message 7 of 10
ryan.harvey
in reply to: Bruce.Lee.

Hey Bruce,

 

We found a memory leak type issue with particles (and other bifShapes too actually) + Arnold rendering, but in order for the leak to happen, Arnold needed to be running a render session while frame changes were happening.

 

We haven't seen a leak with just Bifrost/Viewport and particles. Would you be able to attach your scene?

 

Thanks!

Ryan

Bifrost QA at Autodesk
Message 8 of 10
Bruce.Lee.
in reply to: ryan.harvey

Hi ryan,

I have figured out what went wrong. The reason is that I used two Bifrost graph nodes. Because I want a Bifrost graph node to be used for calculation, another graph node is used to handle the cached particle size. The result is that Bifrost quickly eats a lot of memory, and when you return to the initial frame to start the calculation, the memory will not be released. No matter what you do, the memory will only increase and will not decrease. There is only one way to clear the cache. Restarting maya will clear the memory used by Bifrost.

Thanks.

 

 

 

Message 9 of 10
Bruce.Lee.
in reply to: Bruce.Lee.

This is an attached file

Message 10 of 10
ryan.harvey
in reply to: Bruce.Lee.

Thanks Bruce, this is really helpful to us -- I can reproduce this here with your scene. I've logged it as BIFROST-3244. 

 

Cheers,

Ryan

Bifrost QA at Autodesk

Can't find what you're looking for? Ask the community or share your knowledge.

Post to forums  

Autodesk Design & Make Report