Help - Rendering an animation with a static camera.

[super gnu]

hmm well in that example, youre really missing the effect of the sss object's gi on the floor, and on itself. in this case it would be better to use a per-frame lightcache, but i think for your character, this will be less of an issue.

basically there are about 100 ways to do this, its unfortunately necessary to try a few to see what works best.

[Emcs07]

hmm well in that example, youre really missing the effect of the sss object's gi on the floor, and on itself. in this case it would be better to use a per-frame lightcache, but i think for your character, this will be less of an issue.

basically there are about 100 ways to do this, its unfortunately necessary to try a few to see what works best.

Yep and I am missing the bright orange bounce from the cube onto the sss sphere. I think I'll test this on the large scale scene, will probably take 1 or 2 days though, since the renders take so long.

[Emcs07]

Well, I don't think saving the light cache method will work out. I tried it on a dummy scene with the character and a simple background. The standard render without anything pre-calculated renders in 11 minutes. Using the saved light cache takes 22 minutes. I don't understand why, but that's what it does.

[andybot_cg]

sorry if I'm kind of a noob at this, just learning Maya, so my thought may not apply. But if this is something you can bake the texture, you could basically just have a screen behind instead of having to calculate that complex surface over and over.

[ivaylo.ivanov]

Hello Vlado,

Thank you so much for your response. I've read through that tutorial but I am still somewhat concerned the render times will still be pretty high since the spheres would have to be big enough to encompass the shadow on the right side (I assume).

I'll try it and see how it goes.

Thank you for your patience,

Eduardo

It doesn't matter how big they are, but what geometry they intersect. However arranging them optimally can be tricky, but you will get the proper lighting. Yo can try manually do something like what sphere fade do, i.e. calculate the full render, but only for your character. This means to use the other geometry as matte objects, but still to receive proper GI from them, and cast shadows to them. It looks to me that this involves less cheating, and more things should work as expected .

[stezza]

I created a document relevant to our animation process here, it may help shed some light on the animation process.

As you have a character you will have to calculate maps at every frame.
Make sure everything is visible first for the calculations.
Then apply an object properties node to the background and turn on matte with -1 (you only need to render one image for the background then comp the character after)
If you can see the ground add another OPN for the ground and again turn on Matte with -1 and shadows with effect alpha.

In your case to help speed up the rendering you could just create an average colour of the rocks and apply it to a flat plane in the background as you dont really need to calculate all those little details to get the GI colour onto the character.

[Emcs07]

I created a document relevant to our animation process here, it may help shed some light on the animation process.

As you have a character you will have to calculate maps at every frame.
Make sure everything is visible first for the calculations.
Then apply an object properties node to the background and turn on matte with -1 (you only need to render one image for the background then comp the character after)
If you can see the ground add another OPN for the ground and again turn on Matte with -1 and shadows with effect alpha.

In your case to help speed up the rendering you could just create an average colour of the rocks and apply it to a flat plane in the background as you dont really need to calculate all those little details to get the GI colour onto the character.

Hi, thank you so much for taking the time to write that PDF. So, if I understand correctly, the first pass where you generate the light maps, you use an object acting as the wall (or even the wall itself without displacement and with a flat color) to generate them. Afterwards, you would use the real rock with a shadow matte and the characters in the scene using the previously calculated maps? And finally, render just the rock and composite the character/shadow matte render in post?

Sounds interesting, definitely a possibility!

Best regards,

Eduardo

[Emcs07]

I created a document relevant to our animation process here, it may help shed some light on the animation process.

As you have a character you will have to calculate maps at every frame.
Make sure everything is visible first for the calculations.
Then apply an object properties node to the background and turn on matte with -1 (you only need to render one image for the background then comp the character after)
If you can see the ground add another OPN for the ground and again turn on Matte with -1 and shadows with effect alpha.

In your case to help speed up the rendering you could just create an average colour of the rocks and apply it to a flat plane in the background as you dont really need to calculate all those little details to get the GI colour onto the character.

Alright I finally tested this. Your method seems to work wonders, for the most part.

However I am having trouble with the shadows, the matte material is acting a little weird and catching some really dark / unexpected shadowing.

Here is a composite render with the character and a moving ball composited on-top of the rock:



Here is the complete render straight from VRay:



The background:





Notice the harsh shadows on the composited image. Specifically the character's hand on the bottom left, and the characters head on the right are coming out too strong. I don't know why they are appearing in such a manner.

Thanks again for your help, it has been invaluable.

Best regards,

Eduardo

[sersia]

I created a document relevant to our animation process here, it may help shed some light on the animation process.

Thank you for the document.

[bigbossfr]

It doesn't matter how big they are, but what geometry they intersect. However arranging them optimally can be tricky, but you will get the proper lighting. Yo can try manually do something like what sphere fade do, i.e. calculate the full render, but only for your character. This means to use the other geometry as matte objects, but still to receive proper GI from them, and cast shadows to them. It looks to me that this involves less cheating, and more things should work as expected .

I have a tool for this, to do this in 1 click... I will release soon..