Baku Hashimoto

橋本 麦

CATEGORY
Diary

「動かし方」の分類

なにかを動かす時、操作する物理量が何回微分されたものか — 具体的には位置/速度/加速度のどれをアニメーションさせるのかによって「動かし方」を分類できるのではないかとこの数年考えています。

ゲーム内での自機の移動を例にとると分かりやすくて。Pongでツマミが操作しているのは位置。スーパーマリオ64で3Dスティックが操作しているのは速度。アーケードレースゲームでアクセルとブレーキが操作しているのは加速度。

映像制作の中で「動き」について考えるときも、この分類はなかなか有効だと思う。手描きアニメや、キーフレームアニメーションは位置の推移をプロットしていくことで動かすし、ストップモーションやクレイアニメは、フレーム間の変化量(速度)だけ対象物をずらしたり変形させることで動きを与えていく。アニメだって頭の中で「速さ」とか「タメ/ツメ」をイメージしながら描くじゃないか、と一瞬思うけれども、結局紙の上に残していくのはあくまでもそのコマにおけるキャラクターの位置。

実写撮影だと、慣性が強く働くものを身体で動かす際には加速度を操作することが多い。ドリーを手押しする時は、台車に力(≡加速度)を与えることで加速/減速させる。ジブもそう。手持ちやステディでカメラを動かす時も、基本的には加速度を操作していることになるのだけど、重力との合力になるので少しわかりづらい…

そういう風に考えれば、3Dソフト上で実写らしいカメラワークをつけようにも、どこかぎこちなさが残るのは、アニメーションさせている物理量の次元が違うからだと説明がつく。

そもそも多くの映像ソフトが、速度や加速度に対してキーフレームを打つことを許さないのは(パーティクル系のような例外はあるけれど)、現在の状態をフレーム0から漸化的に計算しないといけなくなるからで、時間を前後させながら編集したり、並列処理させるには不都合だからだったりする。またユーザーにとっては、変化量をアニメーションさせては動きの終点を厳密に設定できないという問題もあって。そのために、AfterEffectsの速度カーブなんかは2つの固定されたキーフレーム間の速度の配分のみ編集できる仕組みになっている。

アニメーターやモーションデザイナーのような、「動き」について考える人達の多くが、パラメーターの変化量ではなく、実は絶対量の推移を描いて(打ち込んで)動きを設計していることになる。それは先に挙げたようなソフトウェア側の都合でもあるし、一度紙に描いた線を指でこすってずらせないように、そのメディアが持つ物性によるものだったりもするので、良し悪しの話とはまた別なのだけど。考えてみれば「動き」という概念自体、文字通り変化量のことなので、そこにちょっとした矛盾も感じる。

「ここでシュン!っとなる」の「シュン!」は速度のピークを表しているのだけども、それを実際に紙やアニメーションカーブ上に再現するときには、一旦頭の中で積分して位置に置き換えてやらなくてはいけない。現実でものを動かす時、摩擦で慣性が働きづらいものには速度を、滑ったり転がったりするものには加速度を力として与えることになるので、そのワンクッションは、人の身体感覚からすると不自然なことなのかもしれなくて。パラパラマンガで跳ねるゴムボールを描くことすらとても難しく、アニメーターには鍛錬が必要なのも、本質的にはその部分の違いに由来する反直感性にある気がする。

この1、2年、クレイアニメやフィードバック系を使ったグラフィック作りを試していたのは、(比喩としてではなく、物理量としての)違う次元から「動き」について考えてみたかったからなのかなぁ、と今になって思いました。

とか衒学趣味じみたことをダラダラ考えつつ、こういうUIがあれば面白いぞとoFでスケッチしてたのが、上の動画でした。

Oct 3, 2017

そのジャンルごとに在る, 大概みんなこういう質感や空気感を良しとして作っているよね. みたいな傾向と, どういうスタンスで構えれば自然と距離を保てるかを延々考えてしまう.

Oct 1, 2017

‪過去の制作の反省でもあるのだけど, キーフレームで動かしたものを1コマづつフィジカルに出力してコマ撮る系全般, そこまで心にグッとこない. レンダラーとして実写を使っているにすぎないというか.‬

‪コマ撮りの良さは見た目のアナログというよりむしろ, 隣接するコマ間の変化量だけで動きを考えることを強いられる不自由さ, それゆえのぎこちなさにある気がする. (描き送りのアニメ作画に近い) キーフレームベースで何度もプレイバックしながら組んだ動きは安直に気持ちよくなり過ぎる.

Jul 19, 2017

汎用ツールの使い方に意識的に制約を導入することで一貫性をもたらすというより、楽器の形や可能な運指がその楽器で奏でることのできる旋律や和声に一定の法則性を与えるように、ツール自体の特性を変えることで手癖を変化させられたら良いなぁ という気持ち

Not Moving Sofa Problem

This piece visualizes “the moving sofa problem” as a kinetic sculpture. The “sofa” moves through the L-shaped corridor forward and backward repeatedly.

I mainly used Cinema4D and Python to calculate and design this piece. Cutting all parts with a laser cutter, sanding its surface for three days.

Some of people might think this as a kind of visualization, educational stuff, or digital fabrication. However, I just wanted to make super smooth surface and super soft sofa as I’ve said over again. Because I regard it as a kinetic sculpture, not a physical computing stuff. Neither getting familiar with mechanics nor transplanting mathematical notion to the physical world could be my subject.

In my career, I often feel wired that people in the context of ‘art & tech’ have some certain biases. For instance, when I asked some artists to explain their idea, they tend to talk about its system just like using Kinect, RaspPi, connecting via OSC, and so on.

But I’d always been wondering why they weren’t telling the aesthetic aspect of their work, just like its color, texture, and how we can feel from it.

I’m not telling just an idealistic way of thinking. As I’ve studied and worked with video production without any code, I just feel such a way of thinking as natural.

I have some friends and co-workers who’re making music videos, animations:

I don’t know how they had thought and felt at that time, but I believe their main concern were in its aesthetics. Not some specific devices and frameworks.

When we meet and have a conversation, we almost spend time on highly abstract topics just like “don’t you feel mesosphere is so emotional?”, or “the rainbow reflection of gasoline on water is so stunning”.
Art is not topology. Art is not a specific combination of concepts, systems and devices, which is like a point cloud and wireframe. But it’s rather like vapor and it’s filling and enclosing an whole piece in a very abstract way. Even if there’re few works using completely same technologies, frameworks, they should have certain differences. Art is what differentiates them.

Some people might argue I should observe more intrinsic part of their work. Or some might say the system and ideology themselves can be the part of art. Of course, admittedly, these can be that. But I think it’s only if the work has certain aesthetics as feeling. Otherwise, there’s not reason these should be the form of art, not an essay or paper.

I don’t want to be exclusive to other people on art. There should be diversity on stances for art and I respect all of them. I just wanted to write down how I’m thinking so far.
Neat wiring and a caption with a clean edge are still important. A way of exhibiting is not just a way of presentation of one’s idea. But it’s an art itself.

The “extremely smooth surface” is still an essential part of my work. Because it strongly affects the impression of my work whether the surface is smooth or roughly cut by laser and interlocked with teeth, even though it’s completely irrelevant from its system and mechanics.

Through the moving sofa, I might want to express such my stance for art and technology. — It didn’t work at all after all, though.

Aesthetic Coherence

Admittedly, the natural of computation has always inspired me and it’s much more than an enabler, but it can be neither subject nor motif of my work. Programming would unleash us the limitation of commoditized software just like Adobe. On the other hand, it imposes us another limitation of graphical expression at the same time and tends to distract us from aesthetic improvement.

It’s much less than just an idea to think about only which technology and devices to use and what concept to argue. It’s not important compared to its look, feels, and atmosphere, after all. In other words, the technology and concept behind it could be important for viewers only if the work has aesthetic coherence and beauty.
Art is not topology. Some of planners or media artists tend to focus on connecting context and context, tech and tech, buzzword and buzzword too much but this only means constructing its topological structure. I think art is continuum and the feeling of material of the nodes and branches themselves.

I’m so bored about the context of ‘art & tech’ but I think artists arguing such a statement should do R&D in term of both technological and aesthetic aspects simultaneously, so to speak. The art produced by who distinguishes tech development and artistic depiction feels like just a stylish demo of cutting-edge technology.


Capture from Please Say Something by David O’Reilly

Here’s one of texts I’ve deeply been impressed ever. It describes very important notion not only for CG animators but also for new media artists.

Nov 16, 2016

As making both of generative stuff and videos using keyframe-based animation, I become not able to distinguish what can be ‘generative’.
I’ve heard the news that the poem written by AI. But it was actually modified by human and the part of AI remained only 30%. I was so confused about the evidence why they can argue such a statement. The cut-up technique is much more generative, isn’t it? Even though it is irrelevant from digital technology.

I’m still wondering whether is it allowed for generative art to modify the output of a program by hand. How far can artists interfere the generative process intentionally to keep their work called as a generative art.

I’ve often deployed very laborious ways for my works. I’d composited more than 5000 street view images with stop-motion footages frame by frame. I also made clay-animation with my friends. But since I’ve used programming and Kinect to make in-house tools, shooting system or custom visual effects, some of reporters introduced me as the artist who’s trying to integrate technology and art and has made his artwork by programming. On each time I’ve wanted to say “No, it’s not automatically generated indeed but mostly made by my hand!”. My only concern is how to improve my works, even if it can be a cheat in usual generative manner.

From my standpoint, the only difference between things called ‘generative’ and others is whether it is created using programming or commoditized tools. Suppose there’s a video that bunch of beautiful particles is flowing. If it is made by After Effects using Particular, the popular plugin for particle simulation, I assume it won’t be called a generative art. But if was a screen capture of openFrameworks sketch, most people will regard it as generative one.

As I’ve described, I think generative art is mostly context-dependent. The term is not mentioning its art style. Anyway, my concern is the possibilities that to try to make it generative might distract artists from improving their arts. Sometimes it’s better to make mesh and textures using Cinema4D and Photoshop rather than to use built-in primitives of OpenGL. When it comes to post-effects, it can be much cooler to use a real film grain footage than shader noise. Although that is so niche topic, I’d like to write down my thought and tips for generative arts as like ‘The Book of Shader’.

SFPC Homework: Plan my week without Google

Conclusion: it’s impossible after all.

Let’s see, suppose I’ll use Firefox instead of Google Chrome, and Bing instead of Google Search, OpenStreetMap for Google Map, Facebook Messenger instead of Gmail, and so on. However, there’s still the shadow of Google. Most of the news website is embedding ads and it’s usually served with Google AdSense. And some of them are using Google Analytics to collect the visitors’ information. Okay, I can use some Ad Blocker and an add-on to prevent from tracking. But it doesn’t mean that I’m free from Google.

In this July, Google and some corporations began the operation of the submarine cable called FASTER, which connect East Asia and North America. So if I visited Japanese website, or just using FaceTime with my girlfriend, some of the packets might go through it and I cannot deny the possibilities they steal it. In addition to Japanese website, there’s certain possibilities my packets is sent via it because of DNS or Data Center.

In short, we are always connected with Google in some meaning, even if we stop using Google services immediately. I think this is one of natural the Internet, that huge of decentralized system has intrinsically. Whether I can be free from Google for a week depends on the definition what does “without Google” mean. It could be just like:

  1. without Google services
  2. without Google infrastructures
  3. without any interference by Google

We only realize (1).

Although it’s just my standpoint, I take the service for granted and give up my privacy willingly. It’s not cynicism, but rather we already have a lot of issues relating our privacy when it comes to that topic. There’s an interesting comparison.


Hacking vrs Defaults

I know I could have such a optimistic thought thanks to the people who living in the Hacking side. If all of us are Defaults side and there’s no Snowden, the world could be getting much worse because of the profit of the huge cooperations.

The conclusion is: there’s no choice to take Google granted. But we always have to have an awareness about what could happen possibly by giving up our personal information.

SFPC Oct 11th: Three Questions

What you want SFPC to be

I would not like to say so much conceptual and abstract things. What I hope SFPC is to stay cool simply. I think the method of “sharing process and code” doesn’t come from a kind of moral or conscience, but just for a kind of efficience. I think so many young students interested in new media art want to be a cool artist, but some of them who don’t have a background in hacker culture doesn’t consider it as important at least in Japan. So I’d like SFPC to be an icon of new media arts on the earth, and spread the sense of value that “sharing some stuff is not only kind but also cool as an artist”.

What do you want to make at SFPC

I might be saying the same things again, however, I would like to make a video at the end. It’s just an idea of this moment, I’m planning to build a sort of interactive website, which has stunningly beautiful color and impressive texture. I already have an image for it, however, it’s too difficult to explain on Medium. If I was to say, it might look like chromatography or something.

On the other side, I’ve been hooked on the problem: “How to integrate a generative approach and hand-made approach”. In this case generative means using creative-coding framework, procedural, realtime. Hand-made means timeline-based, designing frame-by-frame, or some laborious work. I always like generative approach as an engineer. However, as motion graphics designer, I’ve thought such a kind of method add a look of works some certain similarities. In short, so many digital works looks like Tron:Legacy, Minority Report. Using a method which has developed in video production field in generative approach could resolve this problem. So I’d like to explore and expand possibilities of it. And then, I’m planning to archive the website as video before it becomes unavailable.

(This statement might be too abstract. I think below my work is good example for it. Which combined with stop-motion, kinect, and openFrameworks.)

What can you teach and share

(Japanese, interesting terrain I’ve found at Google Maps, Anime, basic use of terminal and git)

As I mentioned, I have a background in video production. Although I’m not a good engineer and much less hacker, I can teach my experience using both ways of creative-coding and timeline-based video production. For example, how to export keyframes from Cinema4D and play it on a browser with Three.js. Or how to control DMX using Cinema4D. It is still so niche though…

Anyway, I’m planning to share an idea and process of my ongoing graduation work as much as possible while the school term.

SFPC Oct 21st: The Smartest Computer on the Planet

Today it was the first class of Ramsey Nasser, a computer artist and designer. In this few year, I’ve tried to consider and relativize programming as one of painting tools such like brush, pencil, and plaster. However, through this class, I came to think it might be a bit unreasonable, at least with the present.

Programming is still depending on a text information, unlike other painting tools. In addition, there’s significant latency between ‘implementing’ and ‘executing’. If I put an oil color to canvas, I can see its result momentarily. However, Programming cannot behave such like this. So I inspired his thought that ‘Processing is not enough for instinctive creative coding since it has “Run” button.”

How to make programming more instinctive tool? Now I assume programming itself could not be instinctive forever. And what we can do with programming will be able to be achieved by GUI with the times. I don’t mean humankind will not need programming near future. However, coding as implementing variables, functions following to very struct syntax will be not essential except for few people, like even professional creative coder doesn’t need to write an assembly language anymore.

On the other hand, highly encapsulated system and software tend to put a limit on artists’ imagination unconsciously. Many graphic designers who using Illustrator barely come up with procedurally created logo because it doesn’t allow them to draw with a more generative approach. So no matter how the technology develops, we as new media artists still have a role to explore lower-level layers of computation, hack them, and inspire the other artists whose idea are limited by the design concepts of their software.