Static Irradiance Particle + Animated Irradiance Particle?

[Narann]

Hi mr community,

I have an animated character in a room.

I use Irradiance Particle like Final Gather (0 bounds).

To avoid loose of time I would like to do this:

Generate a static Irradiance Particle map on the empty scene.

Add my character and compute Irrandiance Particles only on it surfaces.

Merge the two irradiance particle maps and do the render.

What could be the best way?

Regards,

Dorian

[Remydrh]

Is there a reason to use IP and not FG with the FG shooter? You may be able to avoid using maps and caching them this way.

[bart]

To further explain that suggestion, when using FG+IP, the final indirect results are stored in the FG map. Just like it does for FG+Photons or FG+FG secondary indirect.

[Narann]

Thanks for answers guys!

Is there a reason to use IP and not FG with the FG shooter? You may be able to avoid using maps and caching them this way.

I'm not sure to understand your last sentence. You say "I could be able to avoid using maps and caching", but this is what I actually want.

I want to actually have a static map of primary (and maybe secondary) rays to use it during GI computation.

when using FG+IP, the final indirect results are stored in the FG map. Just like it does for FG+Photons or FG+FG secondary indirect.

Tell me if I am wrong but Irradiance Particles (IP) are supposed to be projected for the whole scene right? And FG is only camera dependent right?

That's why I was thinking to use IP instead of FG as I would like to cache the whole static scene and don't rebuilt it each time...

But from what you say, I should use IP for shooting ray (and so, shoot rays in the whole scene), store this rays in a (for example) ip_static.map and use this map with FG for the render. Am I right?

If so, the workflow should be to:

- Activate IP only.
- Render the scene without character to create a "ip_static.map".
- Disable IP.
- Activate FG.
- Put "ip_static.map" in the "Primary FG File".
- Put FG Rebuild Freeze.
- Launch the render and it should use "ip_static.map" to compute FG.

Am I right?

If not, could someone explain (or give me some URL which explain) a step by step method to do what I want?


So now for my static+animated case:

- Once I have my "ip_static.map" file
- I create a "Secondary FG File" entry in Maya.
- I put my "ip_static.map" in it.
- I put "Primary FG File" to (for example) "default.fgmap"
- I put rebuild in ON (because it actually should rebuild only few part of the scene, the animated ones).

But I'm not sure all of this is the good way...

Note: Maya "mapViz" node doesn't work... I suppose this is due to the new map container of mr. How could I check my maps? Do you think, Nvidia could provide a OpenGL viewer to display .maps?
Note2: I also see there is no "Render mode" to only compute Irradiance Particles... I need to render the whole image (which is long...). Maybe it's hide somewhere... How do you deal with that?

[Kzin]

ip's are also view dependent like fg so you cant render ip's for the whole scene and cache them for all following frames.
but the fg_shooter can help to catch the whole scene with final gathering. create additionell cameras which all pointed in different direction of your scene so you cover all areas of your scene with these cams. then apply the fg_shooter shader and the fg will be computed from you rendercam and all the cams which has the fg_shooter shader applied. the rendered fgmap can then reused for all following frames.

[Narann]

ip's are also view dependent like fg so you cant render ip's for the whole scene and cache them for all following frames.

After some reads in the IP doc, I've also seen this. Thanks.

OMG, I completely miss what Remydrh would mean with FG shooter. He actually mean mip_fgshoot I completely forget about it! Thanks Kzin for making me notice it.

This open a wide variety of possibility.

I've also find information there and there. And in the mr production pdf.

So I suppose I need to use IP + FG with mip_fgshooter to store IP in a big fg map and use it later.

I will try that! Thanks again!

EDIT: I also notice the sentence is not finished in the mr production pdf (page 28 ): "This guarantees that even if the camera moves" and?

The problem also appear in the "Utility Shaders" page.

[Remydrh]

Yeah I'm not sure what that sentence was supposed to say without looking at the rest.

We have a UI for the fgshooter on the elementalray blog for you to try/use/modify. It makes adding virtual cameras easier. I assume your workflow is to eliminate flicker in an animation, this should help in most cases without making you cache files.

If you want to cache files for rendering later, this can still help.

[Narann]

Yeah, I've seen it and will try it.
Thanks for this tool.

I assume your workflow is to eliminate flicker in an animation, this should help in most cases without making you cache files.

This is what I'm not sure to understand. What is the point to use virtual camera for shoot many part of the scene if you don't cache them?

To increase FG quality on some specific parts? Where, for example, FG have some pickles? Like "portal" node is suppose to do? Am I right?

Sorry to use your time guys.

[Remydrh]

Using multiple camera positions for Final Gather improves the adaptability of the solution.

By looking at the scene from different viewpoints, you see more points at once for rendering the frame. This means the interpolation of Final gather points has more information to draw from. When using an interpolated technique, more information improves the result (you can better predict a result with more measurements, which this provides.) This means you should be able to render with a more stable FG solution frame to frame.

[Narann]

Ok thanks. So the "trick" only consist to add a lot more of FG points. Actually, shoot the whole scene with FG to avoid flicking.

By "interpolated technique", you don't mean "interpolation between animated frame" right?

Thanks a lot for all this precious informations. I will try by myself now.