Thinking about reasons for the discomfort of Anime 3D Charactors.

アニメ表現された3Dキャラの違和感の理由について考えてみた

“Feeling uncomfortable with cell-look 3D Characters.”

Of course, there are some animations that use 3D expresson and usage that do not fell strange. However, there are probably many peaple who fell that “cell-look 3D characters” are more “strange” than characters expressed by “Sakuga”.

sawada

I’m one of them.

I have been working in the Anime CG industry for about 10 years, and I’m aware that there are “areas that can be expressed by Sakuga or Photorealistic 3D but not by cell-look 3D characters,” and there are many “issues”.

So, I have compiled such “issues” concerning “3D characters in Anime expression”. If these issues are resolved, it will become easier to see them as a “method of expression” and many peaple will no longer feel uncomfortable…
This is the beinning of my blog.

The term “3D character with Anime expression” is generally reffered to as “cell-look 3D”, but it is also called “toon rendering,” “non-photorealistic,” “painterly,” and many other terms when touches and various textures are added. Since this is a bit cumbersome, i will refer to them here as “Anime 3D Characters” for simplicity.

Let’s explore the “Identity of uncomfortable ” of “Anime 3D Charactors”.

TOC

1. Physical phenomena are not properly drawed

This is the most basics, but I don’t think we have done this properly first.
Even in “Anime 3D Characters” which less uncomfortable, I have nerer seen a Japanese film that can draws this properly. (There are some foreign films that do.)

For example, “cloth” and “hair” are probably the most obvious. Even for a single “wrinkle,” the changes in wrinkles when the body is moved are not drawed properly.

Have you ever seen an “Anime 3D Character” like this?

The use of deformations is an attempt to express a feeling to some extent, but it is still “just like that”.
Compared to the “Sakuga,” it can be said that the “Resolution” is low.

Other deformations of muscles and human body structures are not properly done. The effects of “Subdivision-Surface” and deformation can easily round out various areas. If it’s a “Long Shot” it is fine, but when it is “Close-Up Shot”, it becomes tougher.

I think the reason why the CG of Pixar, Disney, etc. does not feel uncomfortable in the same 3DCG character expression is first of all bacause this area is properly done. (Of course, it is also bacause of tha budget, but I will not worry about that for once here.)

If this area is not firmly in place , first of all, the subject drawed does not look like it should.

2. Controlling Lighting Expression

In terms of pictorial expression, it is “shading” and “highlighting”. I think there are fewer films where this is properly controlled. I think the most common cause of discomfort is “fixed texture shadows that don’t move” or “lots of weird shadows that look creepy in lighting shadows”. There are not so many shots in “Sakuga” that are so particular about lighting expresson, but in the case of 3D, it becomes “unnatural” unless it is controlled very intentionally.

  • Shader Development
  • Texture
  • Controlling Normal Vector
  • Objects
  • Drawing
  • etc.

Various Methods have been developed, but i think it can be said there is still no “definitive method that is intuitive, efficient, and looks natural”.

If I must say, i think that when we can do something like Disney’s Meander Animaton Tool, we will be able to express ourselves more. I would like to sell it, but I don’t think Disney will open it to the public, so I am hoping for “Blender”, “vector-intensive 2D software such as CLIP Studio or CACANi”, and “PSOFT”.

I think for shadow shapes that require fine control, it seems that the fastest and highest quality way it to draw them in some way. Looking at Arcane Style’s Tutorial and Making of Disney’s Paperman, it seems that they are drawing and controlling something.

3. Lack of cotrol over the amount of information

“Sakuga” is a “Drawed Picture”. And “Anime 3D Character” is an imitation of that “Sakuga”.

It comes naturally when you draw, but for example, the amount of information is different when you draw a person in the distance and when you draw a close-up. This kind of thing is sometimes described in the “Drawing Precautions”.

More specifically, in “Sakuga”, “Long Shot” and “Close-Up Shot” are of course differentiated, as are the mood of the body, depending on the “feeling of the character” and the “dramatic presentation”. I think that such changes are particularly strong with children’s characters.

I think it’s evident in, for example, “Ghibli’s animation”. Naturally, each “Sakuga” artists own habits and preference depanding on the shot, but the “impression” seems to be well controlled.

Simply drawn, appears low isometric.
It is drawn in a realistic and lifelike manner, and appears to have a higher isometric value.

Quoted from KiKi’s Delivery Service – Studio Ghibli

It seems they actually made a conscious effort to make those adjustments.
He say “When I was scolded by Director Miyazaki, I ould hardly understand what was being sayd. However, I did understand three things.
Bust shots of the heroine should be drawn with broad shoulders (to give a sence of realism).
If you are lonely, get tacky. (Don’t make an empty picture just because the parts aren’t in the setting).
When I painted the stone wall as if it were a precision drawing, I was scolded very much. He told me that I should make them feel as if they existed without being painted.”

This tends to be almost constant with 3D. This is because the 3D model is basically one and depands on it. This is also an advantage, and one example that quality is more stable.

It’s just right, but it’s not enough as an “expression”. Since it is a “picture,” you want it to “look more like what you want to express”.

Of course, there are films where the skill of the “3D Animator” is able to “make it look more so”. However, when asked, “Is it as good as the Sakuga?”, most of them are doing their best within the realm of deformation.

What’s not good about having a constant amount of information?

I think that the weak point of “Anime 3D Characters” is the “Close-Up” Shot. Sometimes Sakuga’s detailed drawings for make the “Close-Up” Shot look more impressive.

This creates contrast in the infomation and makes it more interesting to look at. Since “Close-Up” means that the subject is more in focus from a production point of view, a picture with more “resolution” is more likely to convey the emotion of the subject.

Since 3D is created based on the basic “Setting Picture” standard, the infomation is constant unless some element is added to the shot with some intention. In this way, the characters are shown in the same picture each time, so the contrast as an image is low, and I think it is easy to get the impression that the picture is “boring” compared to the “Sakuga”.

There are many other things to consider, but I will write about them another time because they are lengthy.

4. Not pictured, Limited range of expression

As descrived in 1 to 3 abobe, the “expressiveness of the screen” as a “picture” is still inferior to that of “Sakuga”. This may be discomforting, but I think that “little change” can easily contribute to something like “screen boredom”.

In the case of drawing, depanding on the work, symbols and deformed expressions may be added for fun, but in the case of 3D, they are tied to assets, which can be time consuming and difficult to work with due to the workflow and various other restrictions.

There are many processes involved in 3DCG, and compared to “Sakuga”, it is currently more difficult to create smoothly. Even if there are small adjustment or elements that we want to do, it is sometimes too much work to adjust them in 3D, or it is technically difficult to do so.

sawada

I think it is difficult to make the screen interesting unless you create a system with a great deal of awareness of various elements, or unless the site is open to generalist improvisation and originality.

5. Easy to cause discomfort when mixed with “Sakuga”

“Sakuga” and “Anime 3D Characters” are two different things.
This is true for movement, and it is also different when viewed in a single picture. We can “get close” to each other, but not together, even with the quality of the lines, and it seems pointless to put effort into it.

As I have written so far, if you look at “Anime 3D Characters” in their entirety, you may not feel so much discomfort. It is related to “amount of infomation” in 2, but as long as the “quality of infomation” is uniform, the viewer can see it as “that kind of thing,” sof there is no sense of discomfort.

It may be a “don’t mix” or “be careful” kind of thing.

What is important here is the power of “direction”, and it can be accomplished if it should look different. Common examples include using 3DCG for robots and monsters, making all the “recollections” into “Sakuga”, and using “Sakuga” for “effects” and “some of the costumes”.
Other times, “mobs” can be treated as “different” because of the “directional intent” to “show them as a crowd”. Since they are usually small, it is easy to make them look not so discomfort.

What is not good is “a half-baked mishmash that does not include a production intention or final picture finish that has not been worked out.

  • “Sakuga” Artist can’t draw, so draw at “3D”.
  • It would cost too much to make 3D assets for something that will only apperar in a few shots., so we are going to draw at “Sakuga”.

It would be worse if, after these business circumstances are taken into account, “Sakuga’s Character and 3D Character live together in the same shot while both move” in a “relatively Medium Shot”.

Summary : 3DCG is a lot of work.

If i think about it, we just can’t do what is natural for the “Sakuga” Artists to be able to draw.

In summary, hare is what i think.

3DCG has a large amount of visual infomation (easy to add), but a small amount of expressive infomation (difficult to add). Therefore, in terms of “expression”, “Anime 3D Character” tends to look cheap compared to “Sakuga”. This is the reason why viewers feel “uncomfortable” with it.

However, this is not a fault of the Anime 3D people, but simply something that in its infancy.

For example, “poor resoponse time”.
Of course, if it is a drawing when you want to add a quick touch, you can draw it there and it will be reflected immediately. With 3D, it takes a lot of preparation to do little thing, it takes a lot of work, and it takes a long time to see the results.
I think creators repeat trial and error from small to large, but this “poor response” is the only reason why it does not lead to quality. It means “poor intuitiveness,” so to speak.

sawada

3DCG is difficult!
It takes a lot of work to try to do something.
Prease give me the magic tool, fast!

The history of 3D Charactoers in Anime is only about 20 years (?) old. So, I think that the way of making 3D characters will change rapidly in the future. In the beginning, animated “Sakuga” ware not as realistic as they are now.

A little excuse.

This time, I dared to focus only on the “discomfort” part, but ” Anime 3D Characters” also have many merits.

  • Creating a picture that add a lot of infomation (For example, you could make a painterly picture.)
  • Strong in handling retakes
  • Easy to make complex movements while moving the camera around
  • Movement using a full number of sheets is relatively less time consuming than “Sakuga”
  • It is easy to make a design that is difficult to put into “a design that can be drawn by Sakuga”

“Anime 3D Characters” can do thing that are difficult to do in “Sakuga”.

Another advantage of using “Anime 3D Characters” is that one person or a small group of people can create a complete make a film if they are not concerned about quality. I have an image that many Japanese famous CG creators are individual artists.

In addition, although I focused only on “Characters” in this article, I think that “3DCG” which the larger sense is now an indispensable part of Anime.

  • Taking live-action realistic layouts
  • It is easy to try and error various layout with director
  • Moving BG’s
  • Creating effects unique to CG

Finally.

In the first place, since it is only pseudo-reproduction of a “Manga Picture” in 3D, so I think it is ultimately impossible to convert “3D” into a “picture” at this point. In that sense, “Sakuga” and “3D” are two different thing, but I think we need to think about how to express animation expressions that are not currently available and how to improve the quality.

sawada

I prefer to use CG technology extensively to create a hybrid of 2D and 3D expressions that have never been seen before, rather than reproducing “Sakuga”.
In the case of traditional Japanese arts, they sey that you should first firmly inherit the techniques, and then do somothing new in accordance with the times without being bound by the concept.

Eventually, somethink like “there is more demand for imformatve expressions that can be done with Anime 3D Character” may happen.

sawada

By then, I would like to resolve these issues.

アニメ表現された3Dキャラの違和感の理由について考えてみた

If you like this article, please
Follow !

Comments

To comment

TOC