Xparticles Explosia FX render problems : channels bounds mismatch

[Anonymous]

Hi !

Trying to render an explosia FX object with arnold while using temperature and velocity channels to drive rendering style and from frames to frames it flickers (arnold lost track of the channels) and got this in the log ( for 10 frames out of 100 for example) :

[c4dtoa] Explosia FX field 'temperature' bounds mismatch (expected: [-168.000000 0.000000 -192.000000] - [192.000000 360.000000 216.000000], has: [-168.000000 0.000000 -192.000000] - [192.000000 360.000000 204.000000])

and the same for velocity, uvw and color channel

tried disabling adaptative bounds in ExplosiaFX, but didn't solve the problem

- Another thing to note is that it happens only with a cached sim, if I'm rendering without caching, it works, renders fine, no errors in log.

don't really know if it's on the Arnold side or Xparticle side on this one though...

[peter.horvath6V6K3]

If you can send me the scene I'd take a look. Maybe the plugin does not handle a use-case correctly, although it's strange that the cached and non-cached sims are different.

[Anonymous]

Hi Peter, thx for your help !

here is a simple scene with the problem, i put as helper the velocity channel in the scattering color of the shader, and as I said, as long as I don't cache the sim, the velocity channel clearly send color information to the shader, but as soon as I cache it, the smoke turns black and I got these errors in the IPR log

and the link to the scene (somehow can't upload it here, c4d file not allowed)

https://drive.google.com/file/d/16lyQY2DlZzJC1PT5mtlbtkjDh1bcmmSW/view?usp=sharing

[peter.horvath6V6K3 · Accepted solution]

The bounds are indeed different in the cached simulation. Smoke and fire have 80 x 88 x 112 voxels, while the other channels have 80 x 88 x 104.

I can try to work around it in the plugin and get the intersection of the grids, however it would be nice to check with Insydium if this is intended or a bug. Would you mind to report it to them as well?

[Anonymous]

hi Peter, thanks for your help and feedback ! yes I'll send a bug report in the next minutes to insydium on that matter !

[peter.horvath6V6K3]

A workaround for this issue has been released in C4DtoA 3.3.10.

[Anonymous]

thanks a lot !