OctaneRender® for 3ds max® v2025.3 - 16.08

[remKa]

Hello,

I feel you, phoenix FD fire looks quite bad in any render engine that is not VRay or Corona...

That being said your grid is waaayyy too big for this small fire source, you need also to put a volume step % between 0,1 and 0,01 in your volume medium.

Check the remap option that neonZorglub put in this release, files are available in the forum

[coilbook]

Hello,

I feel you, phoenix FD fire looks quite bad in any render engine that is not VRay or Corona...

That being said your grid is waaayyy too big for this small fire source, you need also to put a volume step % between 0,1 and 0,01 in your volume medium.

Check the remap option that neonZorglub put in this release, files are available in the forum

Thank you! I wish otoy could create better adjustments for fire and smoke. So far there are lots of settings scattered all over and some duplicate each other. Octane should render naturally like naive embergen does in real time or vray.

[coilbook]

Why is 5090 not being used at 100% and runs too cool. is octane not taking full advantage of it during rendering? (using the latest drivers and octane)'

Also where is RTX on check box?

P.S. There is a problem with 3xxx, 4xxx, 5xxx cards where octane is not suing them at 100% unlike 2080ti. Something is not optimized right. On GPU load graph you can see that it's not at 100%

[coilbook]

Also when using tyflow parties (maybe 10000 simple mesh chuncks) times go up by 3-4 times for some reason. Very slow eval times and overall rendering. Something is not optimized

Like in the image attached adding 10,000 simple ty flow mesh chunks (All instances) turns rendering from 30 sec per frame to 2 min and 30 sec.  

Thank you 


P.S. Most of our GPUs are slaves. It looks like when using TY Flow data travels very slow to slaves and by then master with 2 GPUs renders it in 2 min 30 sec. (When Ty flow is off slaves pick up and render in 30 sec per frame) If you look at the slaves' graph it is flat and barely any GPU load. I think tyflow and octane are not optimized well with network rendering and data transfer. All our connections are 10 Gb

2 GPU master with tyflow render time 2 min 30 sec
2 GPU master and 20 GPU slaves with tyflow render time 2 min 30 sec. SAME

2 GPU master and 20 GPU slaves (NO tyflow) render time 30 sec.


Slaves are at 100% GPU render power without tyflow and around 15-20 when tyflow is on

[paride4331]

Hello,

I feel you, phoenix FD fire looks quite bad in any render engine that is not VRay or Corona...

That being said your grid is waaayyy too big for this small fire source, you need also to put a volume step % between 0,1 and 0,01 in your volume medium.

Check the remap option that neonZorglub put in this release, files are available in the forum

Thank you! I wish otoy could create better adjustments for fire and smoke. So far there are lots of settings scattered all over and some duplicate each other. Octane should render naturally like naive embergen does in real time or vray.

Hi Coilbook,
Nvidia has let Otoy team know that they know about an issue affecting other render engines causing some unusually long times for some kernels to initialise. From the users perspective this may look like Octane takes really long to start rendering or unusually long render times. In any case, I’m already reporting your issue to the developers.
Regards
Paride

[coilbook]

Hello,

I feel you, phoenix FD fire looks quite bad in any render engine that is not VRay or Corona...

That being said your grid is waaayyy too big for this small fire source, you need also to put a volume step % between 0,1 and 0,01 in your volume medium.

Check the remap option that neonZorglub put in this release, files are available in the forum

Thank you! I wish otoy could create better adjustments for fire and smoke. So far there are lots of settings scattered all over and some duplicate each other. Octane should render naturally like naive embergen does in real time or vray.

Hi Coilbook,
Nvidia has let Otoy team know that they know about an issue affecting other render engines causing some unusually long times for some kernels to initialise. From the users perspective this may look like Octane takes really long to start rendering or unusually long render times. In any case, I’m already reporting your issue to the developers.
Regards
Paride

Thank you but tyflow bug is on octane side I think

[coilbook]

This version thinks I have Windows 8 even though I’m using Windows 11, and the year in the top right corner still says 2021 instead of 2025

And where is RTX on checkbox?

[paride4331]

This version thinks I have Windows 8 even though I’m using Windows 11, and the year in the top right corner still says 2021 instead of 2025

And where is RTX on checkbox?

Hi coilbook,
I’m going to report the issue about the data shown in the info panel to the developers, but I think it’s just a formality. As for RTX, the option is no longer available because it’s now handled automatically by the render engine.
Are you already using the latest Studio 581.29 | WHQL drivers? Did you use the clean uninstall tool to remove the old drivers?
https://developer.nvidia.com/cleanup-tool
Regards
Paride

[coilbook]

This version thinks I have Windows 8 even though I’m using Windows 11, and the year in the top right corner still says 2021 instead of 2025

And where is RTX on checkbox?

Hi coilbook,
I’m going to report the issue about the data shown in the info panel to the developers, but I think it’s just a formality. As for RTX, the option is no longer available because it’s now handled automatically by the render engine.
Are you already using the latest Studio 581.29 | WHQL drivers? Did you use the clean uninstall tool to remove the old drivers?
https://developer.nvidia.com/cleanup-tool
Regards
Paride

yes. The newest drivers. GPU load is around 30% for all slave GPUs that are 3xxxx series. 2080TIs around 70% Is anyone else having problems with GPU load being low?'

Can it that there are so many GPUs they dont need to work hard? It seams that rendering with 12 GPUs vs 22 GPUs is not much difference

[paride4331]

Hi coilbook,
What about GPU 50.x load series performance?
This might confirm the issue reported by Nvidia. Could you try rolling back the drivers and check if you notice any improvement in GPU usage and render times?
Regards
Paride