I have heard thoroughly how to make "GANTZ: O" for digital frontier Part 5 · Facial version

The movie that celebrated the second week of release "GANTZ: O". The third visualization following television animation and live-action movies was a full CG movie making, so I was waiting to see what kind of work it would be on the work, the work attracted regardless of the original read or unread of the original popped out It was.

In response to that frustration, in GIGAZINE the authors'Hiroya Oku, Director'sYasushi Kawamura, General directorSato KeiichiMr. interviewed. In addition, I was in charge of productionDigital FrontierI got the opportunity to ask the story of the detailed part from the staff of the 8 sections of. So farcharacter,background,Motion capture, Animation, but I finally wrapped it. That 5 · Facial version, I've been talking to Ken Ishiyama of the facial room.

Director Yasushi Kawamura (Kawamura):
With this work, with meAyako ToyoshimaIt is facial animation that CG producer put the most expectation and pressure on.

Production manager Toshi Funahashi (hereinafter Funabashi):
Since this time is CG, it is the best purpose to say that what is not incongruous in that, or to raise the quality of facial is the most important purpose, in that place I was in charge of the heaviest responsibility.

GIGAZINE (hereinafter, G):
Indeed, what kind of pressure was put on it?

Mr. Kensaku Ishiyama, Manager of CG Production Department Facial Office (Ishiyama):
After all it is full CG, in the workEerie valleyI think that it is quite a lot of such impressions such as disgusting or scary, but it was the best pressure to eliminate those things and to show the drama in which emotions are transmitted properly. If Toyosima-san and Kawamura-san seem to feel bad feeling it is useless.

G:
That's easy to say, but ... it's like that.

Ishiyama:
It is a delicate line that the original modeling is neither real nor anime, so it is quite difficult to think that it is not feeling bad by being seen by the general public, or I thought that the hurdle was expensive . However, there are also that, but this time it is "facial gambarouzu", and I got a lot of experience from shooting methods and performance capture for that, so I think I got a good experience. I was going one step beyond not only our own strength but everyone, because I was able to do what I wanted to do from all of the equipment, so I had the pressure, but I expected the expectation of "I can do it" was.

G:
I see.

Ishiyama:
I was doing animation and setup of facial, I think whether to see from the place of how setup controller works. There are such rigs, you can make your line of sight, or you can climb, but if it's a small place, you can get down without lowering the eyelids, and so on. Since the appearance of the character changes slightly depending on the angle and the layout, unlike live action, even if any angle around Kato it looks like Kato properly, is not it? Wrinkle also enters arbitrarily, so that slightly wrinkled wrinkles enter each firmly, so that you can feel some emotion even if you move which controller. After I close my eyes, I feel like closing more closely and going through such details in the beginning. Even just for now, even if "power comes in" properly, it is possible to reproduce as much as possible the place that "power is entering" not only empathy but also "powerful" or "weak", "face is loose" I am making it.


G:
It is such a moveable thing.

Kawamura:
Unlike the previous rig, when moving one place it all seems to move properly anatomically, so it feels like I'm really tinkering the face of a person.

Ishiyama:
I think that it is the same for any section work, but after all it means "to lie if you lie." If it is a face, if you do not pursue such real, such as "This moves when this moves," especially human faces are always seeing everyone's face, so they will feel uncomfortable immediately. It is to make sure that it does not come out as much as possible. Also, in order to simplify the controller as well, if you raise here, your mouth corners will be raised so that all animators are easy to understand. Looking at various rigs, there are things like a lot of controllers including overseas, but by making it as simple as possible makes it easy to understand. Character collapse is likely to occur if it is complicated. When there are a lot of controllers, the animator in charge will be responsible for something like "That? This character is not like that", so to make it too few, each controller is so quickly attached with each expression.

G:
If there are a lot of controllers, it will be far away.

Ishiyama:
It seems that it seems that you will not be able to remember which one is a real face because you are too tamed by yourself.

G:
That means that you have to do it to the minimum necessary to prevent it.

Ishiyama:
That consciousness was terrible. In the company there are six core members of the facial team, mostly animation teams and character team people who have never done facials come with help. If so, it becomes difficult for more controllers and it will be for advanced users, so we are making particular ways to make it simple.

G:
In "GANTZ: O", "The face of the original human being is roughly like this" "But this is absolutely CG" "No ... ... different, like a human being" I felt quite human-like It was. Previously, GANTZ was made into a live-action movie, but since live-actions are played by real people, "human-like" is commonplace. But, "GANTZ: O" looks more realistic than that, but why is that?

Ishiyama:
I think that Pixar's John Lasseter talked about "I should pursue the reality that everyone is envisioning, not pursuing real," I think that it was time for Finding Nemo, " Because there are various kinds of imagination in the children like fish doing like this "If you just make a real fish you will be deemed" bad "or" unhappy " , I was saying something like to pursue reality with a dreamy feeling. Everyone, I think that each of you thinks that when you see the work "This character is good" "This character is like this", is there. That is to aim for "good reality" rather than to strike a real stupid honestly. For example, if you imagine an actor saying "When Kato shouted, such a feeling might be cool", there are also moments when it is actually a disgusting face as it is peeling the white eyes. But, bringing it makes me feel really bad. There is a part that trial and error is still still in that area.

G:
The current story is easy to understand. When it is told, it feels like "Oh, I see." So, there are actually various characters in this time, so it will be done with most of them.

Ishiyama:
That's right.

G:
I think that it is an amazing amount of work ... ....

Ishiyama:
That's fine ... That's right.

Kawamura:
I decided that he was in charge of Kyou Yamasaki and he fixed the animator for each character.

Ishiyama:
When each character's character is different, when I become responsible for various characters, doing Kato in 90 minutes and doing Yamasaki and doing a Suzuki's uncle ... I will not know the reason any more I'm fixing the charge. It is like Kato team, Yamasaki team. Although I am making "GANTZ: O", I felt like I was thinking only about Kato as a result, as a result I thought like "What's I doing?" (Laugh), so Everyone is easy to do and you can truly pursue it there. Because each team is doing with the feeling that "I will make my character better" for each team, it is quite smooth if you collide with your emotions including the person in charge under that circumstance, you will not collapse the characters, There is a thing that everyone's feelings can remain quite high.

G:
It is simple, but it is effective.

Ishiyama:
I think that it was effective.

G:
It was amazing that he actually did it ... ....

Ishiyama:
I think that the character was moving at the rig now but I am taking all motion data of the base with motion capture. Even my eyebrows are just raised, this fine bean grain is a marker, but I lay it on my face about 300 points, I use this movement as the base of the rig.


Funabashi:
It's the purpose of analyzing where it moves.

Ishiyama:
Well, when you raise your eyebrows, foreheads are moving fine.

Kawamura:
Ears also moved.

Ishiyama:
It moves slightly with a picture. The base is as realistic as possible without telling a lie, from there we are making adjustments to feel like Kato or like Yamasaki.

G:
For example, what kind of thing is "like" of the adjustment of "like" feeling?

Ishiyama:
In Kato I have a wonderful V shaped eyebrow and it does not make me feel like moving it like this, so when I slightly adjust the shape or raise my eyebrow, I make it like this.

G:
Does that mean that each character has a personality of face movement?

Ishiyama:
That's right. Another thing is that it is a collection of facial expressions, like checking when you made a rig, or making some collections of facial expressions that Mr. Kawamura will check.


Kawamura:
This is a common anime.

Ishiyama:
Oka is after. Oka does not come out much, but I am making Tengu and Yamasaki respectively.

Funabashi:
It was a story of the main character now, but youkai also has all the facials.

G:
Does that mean that there are such things that everyone has a face?

Ishiyama:
Well, afterwards it is a feeling that I confirm the place where I became the final look as to whether there is no problem by rendering.

G:
Watching this, was the director saying "This expression is wrong"?

Ishiyama:
"There was a little more like this, ohhh."

G:
What was that like?

Ishiyama:
In Kato, there are things that it is safe to not wrink too much when it is originally such a character of a character, but director Kawamura says "I want to put this time", "Let's wrinkle a bit more here "Or" I am going to put a feeling more like a muscle here a little "here.

Kato's facial expression collection. Many expressions that fight desperately in the work are not expressions like "crying", but it has become a face that can be convinced by those who saw "If Kato cries it would look like this" .


G:
I see.

Funabashi:
Some of them changed the size of teeth when making a face, were not there?

Ishiyama:
I agree.

G:
Tooth size?

Ishiyama:
I was doing adjustment of how to see and how big the tooth was.

Kawamura:
When I first saw this, I told you that "your teeth are small," they were made big.

Ishiyama:
When I was making a character, I made it with a serious face with almost no mouth open, so I did not see the balance between shouting or actually clapping, so I adjusted the shape around that I did something.

G:
So it is something you will know when you make various facial expressions.

Ishiyama:
Yes, since I included facial expressions unless I put facial expressions, I used to go back to the character team and fix it with a facial while attaching facials.

Funabashi:
If you do not do this at this stage, teeth will open after the last rendering.

G:
Is this already done at the very beginning?

Ishiyama:
I agree. It is the first. I made Kato the first time, so at that point I solidified my direction and made wrinkles or such feeling, so after that it seems like it will do accordingly.

Kawamura:
So when you make a facial check with the final facial cut in, do not render and check with a gray face.

Ishiyama:
When animation check by all means it can not do rendering on cost and schedule.

Kawamura:
In other words, cost cutting at that place raises other qualities. There is a memory that I thought "I'm good" by asking people whether "Final Fantasy" properly rendered and checking was read by something.

Ishiyama:
We will imagine after seeing this.

Funabashi:
That is intracerebral simulation, also useful for lighting and so on.

G:
So when you do it firmly for the first time.

Ishiyama:
Because there is also it, so that it does not change so much with the rig that I talked earlier, or it does not make a strange face depending on the person in charge, I am pretty careful to make sure that the blur will not come out.

Kawamura:
Not rendering is a considerable cost reduction.

Ishiyama:
However, Ms. Kawamura is also checking, "May this be okay ... ...?" "...... OK!" May be mutual.

Kawamura:
With two people, "Hold the okay with all right" while floating some overhead on top of each other.

G:
I see.

Kawamura:
That's the state.

Ishiyama:
As I finally failed, I fixed that kind of thing, but they said to each other, "This is useless."

Kawamura:
"Is this a bit different" was about 10 shots?

G:
It is amazing to have almost hit it.

Kawamura:
I thought he was doing quite well.

Ishiyama:
It is easy to understand when you look at the gap, but there is a difference this much. This is how Maya looks on the hardware ... ....

G:
The impression is different at all with crying. Still it is a story saying "You can do it" by guessing "I guess that it will be like this" by looking at the guy who cries at the previous stage.

Ishiyama:
The best is to be able to render, though. Also, as the way of looking changes according to lighting, even if rendering is done comprehensively, if the lighting is incorrect, it is said that it is a troublesome work twice.

G:
They are intertwined and difficult part.

Funabashi:
So, the last thing to see is after rendering. Regarding simulation, there are things like imagination to some extent, but if you do not render, the last answer will not come out.

Ishiyama:
I agree. On the contrary, I think that there is also a place where I got a lot of lighting.

Kawamura:
I was expressing emotions in writing.

G:
I feel like that.

Ishiyama:
What is wrong, lighting seems to be fundamental, but facials are also delicate. It's just here, but it's a bit fine as the eyebrows are going up a bit, "It's a little different" or something. I'm about to work with controllers like around 0.05.

G:
Even ordinary people do not know because they are accustomed to the human face.

Ishiyama:
Because something disagreeable comes out, it is messy with delicately.

G:
In the neighborhood, even from the director's perspective, is there a place like "Is not this a little expression different?"

Kawamura:
I think that was there. After rendering, I heard that "Is that what Suzuki's facial something is different?" One cache before.

G:
The cache was one before! Is it?

Ishiyama:
There was.

Kawamura:
There are few mistakes really, but I occasionally saw it now.

G:
I do not even notice that as well.

Kawamura:
That has noticed.

G:
Indeed, as human beings are familiar with us, they are horrible.

Ishiyama:
It is frightening.

Kawamura:
I will be careful there.

Ishiyama:
After that there are also quite a few comics and other impressions.

G:
Although the manga is in two dimensions and its facial expressions are solid, when it is made into 3DCG it is strange as to why it looks like a facial expression so much, is that putting it on manga, or is it drawing on reality? Or is it something different?

Ishiyama:
Personally, many things are mixed too much and I do not understand at all. (Lol)

G:
Did you reach by repeating trial and error?

Ishiyama:
I also refer to manga, but there are differences depending on volume. Afterwards, various things such as watching Oguri Shun of "Close" which is a bit similar in live action, watching Hirosuse Suzumi if it is Yamasaki.

G:
Indeed, I'm going to combine various degrees of detailing up with the degree of various reality.

Ishiyama:
That's right. After all, if you think that Mr. Kawamura and myself are "good-looking" or "cute", it seems that people looking at it will be okay as well.

G:
It is not necessarily "this direction to make" unconditionally, it is attaching everything.

Ishiyama:
Another bigger thing is performance capture. Since I make it by utilizing the performance of the actor as it is, I think that there is considerable reality around there as well. Human beings have something to talk about with their eyes considerably and sometimes leads to "being dead eyes" is not a living thing, but it lives as data by capturing the movement of the line of sight I think that it is an element that is lively.

G:
I see.

Ishiyama:
This is the animation work screen, this is the window of the camera, the one for this to fiddle the face controller, this is the image of the head mount capture. Since it is almost all shots, even if 100% capture does not come, it is quite useful as a reference. After that there is a sub camera, which makes the face distorted a bit, so I can not see my mouth, but the nuances of facial expression look different. I also feel that cutting important things can be seen there. This is the screen that you are manipulating with the movement as a guide and what kind of emotions this is.


G:
It certainly does not seem to be said that we should reproduce the movement centering around like we mentioned earlier.

Ishiyama:
Yes.

G:
Motion of emotions when looking objectively objectively ... ... I see.

Ishiyama:
Based on various data, it is feeling pursuing places like "Kato like" and "Yamasaki seems like".

G:
Certainly, unless it is divided like "Kato special team", it seems that it is not understood.

Ishiyama:
It's the most efficient way to do it in a limited time.

G:
Really, various elements are intertwined. Even during motion capture there was a story about the head mounted camera, but also in the facial it was pretty good with a head mounted camera?

Ishiyama:
I can not do without this. Because there is a limit to the imagination of nothing state, I think that it can not be in a high area if there is no base reference or capture data even if it is colored. Also, the most important thing is how much to synchronize with the body, because the body and face are separated separately on our workflow by all means, because things like strange feelings are done separately There were many in his work. This is "there is a facial", but if the body and facial expression do not match, the facial work becomes very difficult. Something does not come nicely to wear an animation. This probably is also checked by Ms. Kawamura and I think that something was wrong. This is very incomprehensible because it is "not synchronized", but this time it is all synchronized and all of the movements of the body and emotions of the face fit perfectly, so please take a look I think that the people who are having it will be able to be seen smoothly.

G:
When I first saw the video at the preview, I felt the frightfulness that I can not put into words well but I can not say anything, I wondered what the identity is and this time I will do thoroughly to Digital Frontier I thought about talking to you but I can see that this "synchronization of the face and the body" is a big factor.

Kawamura:
The production period of "GANTZ: O" has been around for a couple of years now, but there are accumulations of know-how and intuition that have been utilized there for over ten years.

G:
I get the impression that the things I've accumulated have gathered in a splendid form this time.

Kawamura:
I think so.

Ishiyama:
This project itself was a long time pretty long time and it was just a little spacey so I was pretty enthusiastic. Finally, I brought a series of body animation, facials, and finals.


Ishiyama:
This is the scene 39, I almost almost checked it and it seems to be finalized. The body is made by the animation team, but while working while looking at the real face what the actor is doing in this scene, the facial makes a move about the face, and finally it is writing and so on It will be about the final. This scene has the most excitement. My team was done when I did this, but when the last final came up I thought that I was writing and compositing and making it more cunning. I was very happy.

G:
Amazing ...... This face is also completely reaching the area of ​​acting. The animation and the facial combined, and the final came out in a firm form like this ....

Ishiyama:
Perhaps, I think that all sorts of teams did not do something strange and came straight to create a synergistic effect, which made me feel very good.


G:
The word "synergy" to multiply two or more is exactly the job. Thank you very much for showing me great things.

· Continued
I have heard thoroughly how to make "GANTZ: O" for digital frontier Part 6 · Setup - GIGAZINE