Friends who often use Lightscape or V-Ray to perform indoor performance，we often find objects in the scene that are more saturated in color than those in the same area. As a result of light transmission or photons rebound, the material overflow color is generated for other materials with low color saturation in the scene, finally, the effect diagram is distorted. Here is a brief summary of the common methods of overflowing color when using V-Ray rendering, let's share it with you.
As shown above, in the scene, the floor belongs to a relatively large red solid wood flooring material, because its color saturation is relatively high. So in the process of photon rebound, photons carry the red color information of the floor and bombard other normal material surfaces such as white walls. Finally, the color of the white wall becomes reddish, this is not right.
Here are some solutions:
1. Wrapped material method: Turn the material of the floor into the VRayMtlWrapper of the V-Ray renderer, select "save old material as sub material" in the replacement material panel.
In the new material panel, just lower the Generate GI value. For example, turn it down to 0.4, this means that the photons rebound only 40% of the original value, the energy rebound of photons has decreased, the phenomenon of overflow is also reduced, but this sometimes affects the lighting level of the scene. Be careful to use and match lighting or exposure parameters to increase the brightness of the scene, the following is the solution at 0.4.
2. Cross boundary material law: This method also takes advantage of the VRayOverrideMtl material that comes with the V-Ray renderer, also save the sub material in the replacement material panel that pops up, I will not tell you here. After you save it, you will find that the original wood flooring material will be preserved as Base Material, all you need to do is copy it intact to GI material, and adjust the surface color of diffuse in GI material to the grey white without overflow, of course, the floor map should be removed.
3. Rendering parameter method: This method is the simplest, all you need to do is reduce the Saturation in "Post-Processing" in the "Indirect Illumination (GI) Global Illumination Volume Display Bar in the rendering panel or set it directly to zero. This will directly seal all the color overshadows in V-Ray rendering.
But the disadvantages of doing so are extreme, some of the materials in the scene are overblown and fair, be careful to use, reduce the rendering effect when Saturation is 0.
Rocket Raccoon is one of
the Marvel Cinematic Universe’s most unlikely heroes. The genetically modified
creature is an ill-tempered, poor-mannered and belligerent career criminal. But
over the course of Guardians of the Galaxy, Guardians
of the Galaxy Vol. 2, and Avengers: Infinity War,
and his relationships with Groot, Peter Quill, and Thor, he’s matured into
an altruistic fan favorite.
The
character’s success is down to his believable characterization, hints at a
traumatic past, Sean Gunn and Bradley Cooper’s performance capture — and the
fact that he looks and moves just like a real talking raccoon.
One
of the people behind Rocket is Nate Shaw, a CG supervisor at Method
Studios, which handled the character for Guardians of the Galaxy Vol. 2,
and Avengers:
Infinity War. Nate’s role was to oversee Rocket’s fur pipeline and
his fur shader in V-Ray.
We
caught up with Nate after his inspiring talk at Total Chaos to ask him about
working with the sweet rabbit.
Were you excited to hear that you
would work with Rocket?
I actually did a groom of a CG dog on Divergent, which is a
movie Method worked on years before Guardians of the Galaxy Vol. 2.
It was a full CG shot, and I was working on that right when Guardians
of the Galaxy came out. I remember looking at my dog, and
comparing it to Framestore’s Rocket, and thinking: “That’s really inspiring.
And dang, they’re good!”
I never thought that I would get that chance to then work on Rocket, and I
learned a lot in the process. I can't take credit for it — a lot of people
worked on it. But it was very exciting. I feel close to that character for
sure. I spent so much time looking at him.
Do
you have any Rocket Raccoon toys?
I don’t! I have a small apartment, so I try to avoid as many things as
possible. But if I were to buy one, it would be Rocket.
Could
you explain Method’s relationship with Framestore?
Framestore did the original asset for Guardians of the Galaxy,
as well as the initial build for Guardians of the
Galaxy Vol. 2, so they just sent us a lot of data. They
sent us hair caches, which was every single hair. Then they sent us guide
hairs, although that wasn't as useful as every single hair. They sent us
texture maps, meshes for skin and suits, and they were really helpful. It was
nice working with Framestore because we would get on calls with them, they were
in London, and we were in LA, and we had a really good working relationship.
It
was so collaborative, even between studios, and they were very forthcoming with
any questions we had.
How did you make sure your Rocket
matched Framestore’s?
We had still frames from Framestore. They gave us their look dev environment
which had all their geometry, the Maya file, and their poses for Rocket, so we
would just put Rocket in there. Then render him for the same cameras, and go
back and forth in Nuke until they looked identical.
Rocket’s
movements are created from facial capture of Bradley Cooper, and
motion reference of Sean Gunn. Do you have to do much hand animation on top this
data?
For Rocket's performance, it was hand keyed, so it took a lot of animation time
to do that. We had face shapes from Framestore, and then we just made our own
if we needed extra blend shapes.
What
was the biggest challenge?
On Guardians
of the Galaxy Vol. 2, we had never done a furry creature that was
this hero, this close-up. He needed to perform in such an emotive way, and be
in so many shots. It was a volume of work that we had never done before, and it
took us to the next level of challenges.
Could you explain how you
animated Rocket?
The animators work with Rocket as a mesh, and then they have a fur volume,
which is a skin of Rocket, but replicating the dimensions of his hair. That's
how they would animate and then they would publish the animation, hit a button,
and export their animation to pass on to the next department who would create
an Alembic. Our pipeline bases a lot of things on Alembic, especially
downstream departments from animation. Even just the skin was an animated
Alembic cache.
Then we would do that hair bind, and that would create hairs. On both movies
that we had this post process to give the lighters this extra cache. In Guardians
of the Galaxy Vol. 2, it was Alembics out of Houdini that went
through a proxy. The second movie, Avengers: Infinity War, we
would run it and do the bind in XGen, and that was VR scenes that were rendered
and brought in. It was groomed whether we were using Houdini or XGen, it was
never going into Maya — we skipped it and went straight to V-Ray.
Any
grooming tips for aspiring CG hairdressers?
If you want to be a hair groomer, just look at real things. I develop a
fixation on real-world objects, whether it's buildings in San
Andreas, dogs for Divergent, or raccoons for
Rocket. Dig in deep and look at reference as you work and you'll see things
that didn't in the beginning. Eventually, you get to something that is photo
real.
Did
you use any specific V-Ray features here?
The great thing about V-Ray for Maya was the ability to use proxies.
We weren't working with a fur procedural; we were working with every single
hair. We had bound the hair in Houdini, and then we were caching it back out
because we didn't want to re-groom the character. As we were working with every
single strand of hair, it would crash Maya as soon as we tried to load up just
one section of his hair. Because it never really needed to be in Maya, we could
bypass that by using proxies.
What kind of changes did you make
between Guardians
of the Galaxy Vol. 2 and Avengers: Infinity War?
Just minor changes. The asset was looking pretty good from Guardians
of the Galaxy Vol. 2, but we changed some spec maps or spec
responses on his hair, to add some glints. We added a percentage of hair that
got more or less spec. We did some increase of some subsurface here and there
of the nose, as well as improving displacements.
What
did you think of Total Chaos?
Total Chaos was great. It was the first time I'd been to Bulgaria, and I would
go back. Everyone was really friendly, and there were tons of great speakers.
Any
particular highlights that you've seen?
One highlight was listening to a talk on machine learning, and how that can be
used for render farm predictions, which is super techie and nerdy. It's not
like the flashy Rocket Raccoon, but as a CG supervisor you're concerned with
iteration time, and getting information to then bid on work or knowing how
long's it going to take to render certain jobs. If we can use smarter
technologies, rather than just doing the math, then I'm all for it.
Three-dimensional animation is different from 2D animation. The three-dimensional animation is the image of a two-dimensional planes, move the picture through motion. This is achieved according to the laws of motion. How does the motion law of objects come into being in three-dimensional animation?
1.Time required for action and number of frames
The total duration of a 3D animation is usually predetermined，and the number of seconds in each shot, will be produced according to needs, the length of the film occupied by each action is the time of a shot, long time, slow speed, the number of frames is large. On the contrary, the number of frames is small. Usually in 3D animation, the rhythm is fast, and the shot is short. So, it's going to be a little bit more frames and there may be hundreds of shots.
2.Motion amplitude and space design
The 3d animation production method needs to design the objects and characters in the picture well. The range of the action and the image must be properly handled. Take the hyperbole or simulation techniques, It depends on actual content and the needs. Combined with the perspective of the background picture, you can show different depth and distance effects.
3.Change of speed
The change of speed determines the smoothness of the picture, there are three types of exercise: Uniform speed, acceleration and deceleration. The distance between each moving object is the same as that of the picture, and it is uniform. The distance from small to large, the presentation of the results from the slow to fast, and the reverse is from fast to slow. Through the visual distance in the picture, and the adjustment of the speed of the object, create a sense of beauty in sports , give viewers a visual impact.
The above three points are the key factors of moving objects in 3D animation, the change of lens is inseparable from the setting of these parameters.
