OctaneRender for Carrara 3.0.0.1 [Test]

[PhilW]

The Volumetric Clouds work fine in "cloud" size. But when I shrink them (approx 50 cm side length),
I cant see them any more...
So I assume, thy only work in large sizes.

Eddy - because it is a volumetric effect, when you shrink it, it becomes too sparse to actually see, but it is still there. If you shrink to 10% of the original size, you need to increase the density 10x to get the same apparent density. So for a very small size, you need a correspondingly high density setting to be able to see it. I hope this helps.

[PhilW]

Sorry, I seem to be the one coming up with issues - Wendy mentioned on a Daz thread that volumetric clouds were not working with replicators and I can confirm this - they work in Carrara but this is not being passed on to Octane. Same for replicators and surface replicators. This is quite a biggy for me and I imagine others.

[Sighman]

The Volumetric Clouds work fine in "cloud" size. But when I shrink them (approx 50 cm side length),
I cant see them any more...
So I assume, thy only work in large sizes.

If you reduce the size of a cloud volume then you also need to change the volume step length in the kernel. Warning, if you choose too small a step length then your render time will take forever. So make sure you don't mix small and big clouds in the same scene. A good rule of thumb for step size is 20% of your smallest edge. For your 50cm volume you should set step size to .01 (10cm).

[Sighman]

Sorry, I seem to be the one coming up with issues - Wendy mentioned on a Daz thread that volumetric clouds were not working with replicators and I can confirm this - they work in Carrara but this is not being passed on to Octane. Same for replicators and surface replicators. This is quite a biggy for me and I imagine others.

Ctrl-D instances are working, I haven't tested replicators.

[Sighman]

Thanks for the info - the emitter effect sounds like it might be cool and very useful.

I have attached an illustration of the "flat fog" in the environment medium, there is no sense of it receding as it is applied uniformly across the whole image. In the previous original version, I was able to get good receding atmospheric haze in a scene like this. I am just guessing here, but maybe this is done by having an inward facing "bubble" around the camera, and somehow the normals on that have got inverted?

It looks like you have not increased your medium radius to take in the scale of your scene. Note that, as you increase the medium radius you also need to reduce the power of your medium shader. I have found no difference between the way this and the previous plugin version behave so there must be something you have not set in your scene.

[rashadcarter]

Sighman,

I am personally anxiously awaiting the Baker Camera. There are a myriad of possible uses. One usage I have in mind is to produce SSS maps. I want to render out transmission maps for use in applications like Bryce that don't have SSS. Using a pre-baked transmission map to plug into the ambience glow channel of my favorite biased render engine will actually go a long way toward making it look as if it is receiving lights under the surface. Currently I use a similar approach but the map has been hand painted. I am certain it isnt as accurate ans a real transmission bake would have been.

On the Environment Medium, I am not certain if this feature is operating the way it should. I will look into the standalone to see if the issue persists, but it very well may be an issue with the standalone. One of the reasons I hated the old system of inverted spheres is because I found the result was uniform, same amount of haze in the foreground and in the background. The atmosphere in nature doesnt work this way. Haze should appear to thicken with distance from the camera. I have found that even adding in absorption doesnt solve the issue.

I will investigate the standalone and report back. A uniform haze isnt very useful to me, it's hard for me to believe the current effect is what the masses intended when they asked for an environment medium. IT has to reference zdepth data in some way or it will not look very natural. Back soon.

[Sighman]

The way atmospheric medium works is by having two spheres, one around the camera with inverse normals and one that contains the medium. You can set the size of the medium sphere and you should set it to enclose the area that you want the fog to cover. You then need to set the medium. Setting absorption to black and scatter to white will give you white fog. You then need to adjust scale to get the effect you want. The larger the medium radius, the more the distance fog effect will be seen. If you are pushing the limit of the scale value then you should reduce the scatter color to a darker gray.

[PhilW]

Ah, hah! The light goes on - thanks for the explanation of how this works! I must have just been lucky with my settings first time around. This is an image with a 1000 feet square terrain, 1000 m radius on the environment medium (would it be a good idea to set this in feet as well for consistency?), and Medium Scale of 0.01 with scattering a mid to darkish grey slightly tinted towards sky blue.

[Sighman]

Sighman,

I am personally anxiously awaiting the Baker Camera. There are a myriad of possible uses. One usage I have in mind is to produce SSS maps. I want to render out transmission maps for use in applications like Bryce that don't have SSS. Using a pre-baked transmission map to plug into the ambience glow channel of my favorite biased render engine will actually go a long way toward making it look as if it is receiving lights under the surface. Currently I use a similar approach but the map has been hand painted. I am certain it isnt as accurate ans a real transmission bake would have been.

The way the baking camera has been implemented in Octane means that a mesh can only have a single, non-overlapping UV map. Something like Victoria, for example, cannot be baked properly because it uses overlapping UVs (head, arms, torso, etc.) all overlap in the same uv space, they just target different texture maps via materials. Try loading Victoria into the standalone and assigning a baking camera to see the results.
In order to work around this issue, I need to assign object IDs for each material ID. Then you can collect the object IDs into baking groups and generate the single texture maps that you need. Sounds good but how do I design a UI (in Carrara) to let you configure such a thing. I am open to suggestions.

[PhilW]

Anyone having any success getting emission in a volumetric object? I've tried a range of values and can't see any emission effect. Are there any tricks to getting it to work?