“狂喜者”的制作

這張照片是我一直在做的一個(gè)動(dòng)畫的一部分，我讓它保持動(dòng)畫的動(dòng)力。很久以前，在我無聊的時(shí)候，我畫了一個(gè)奇怪的人物，像貓、老鼠和蝙蝠——全都在一起。我做了一些小動(dòng)畫，但是我從來沒有設(shè)置過，所以我創(chuàng)建了這個(gè)場(chǎng)景。它的概念是在另一個(gè)無聊的時(shí)刻完成的;我想象了一個(gè)復(fù)雜的房間，里面有很多東西，所以在做任何嚴(yán)肅的造型之前，我都試著做一個(gè)低的poly代理版本。我決定做一點(diǎn)實(shí)驗(yàn)，所以我把我想要的東西都畫在紙上，然后用它作為紋理。我把它投射到一個(gè)空心圓柱和擠壓部件上，就像桌子、窗戶、架子和洗衣機(jī)一樣——只是切割和擠壓。用Max的切割工具，所以我真的很喜歡做不規(guī)則的幾何剪裁。當(dāng)我從高聚體開始時(shí)，低聚模型是非常有用的。雖然我確實(shí)改變了一些事情，但我做的大部分事情都是低分辨率的。為了好玩，我也試著用燈光和色彩的幾何圖形來照亮低聚物。它非�？ㄍ�!我想我將來會(huì)做一個(gè)簡(jiǎn)短的場(chǎng)景和角色合并——根本沒有必要渲染，因?yàn)樗�在DirectX上看起來很好!(Fig01和02)

This image is a part of an animation that I have been working on from time to time, and I made it to keep my spirits up - to keep motivated for animating.  It all started a long time ago, at the moment of boredom, when I drew a strange character that resembled a cat, rat and a bat - all in one.  I did some small animations but I never had a setting for it, so I created this scene.  The concept for it was done in another moment of boredom; I imagined a complex room with lots of stuff there so I tried to make a low poly proxy version of it before doing any serious modelling.  I decided to experiment a little so I drew everything I wanted to have there on paper, and used that as a texture.  I projected it on a hollow cylinder and extruded parts, like the desks, windows, shelves and washing machine - just cutting and extruding.  Max's cut tool has a mind of its own so I really enjoyed the careless cutting that made a lot of illegal geometry!  The low poly model was very helpful when I started with the high poly; I knew what was left to model and where everything should be.  Most of the things I did according to low poly, although I did change some things...  For the fun of it, I also tried lighting the low poly, too, using lights and colouring the geometry with vertex colours.  It came out very cartoonish!  I think I will do a short using that technique for scenes and characters in future - there's no need to render at all, as it looks good in DirectX!  (Fig01 and 02)

圖 01

圖02

我試圖解決的主要問題是紋理和渲染。我想把重點(diǎn)放在這里，因?yàn)槠渌�的東西都是直接的。首先,我有很多對(duì)象,我不覺得我想要貼圖和質(zhì)地,(我猜我懶惰,或者只是似乎過度這么做!),所以我試圖想出一個(gè)方法,使組織的材料對(duì)于某些類型的對(duì)象。我做了大的，普通的金屬和木頭的紋理，上面有鮮明的特征，并把它們和我在每個(gè)物體上畫的頂點(diǎn)顏色結(jié)合起來(使用凹凸貼圖作為混合量的面具)。我第一次使用了頂點(diǎn)著色修飾符(3ds Max)和I lo-ooo -ved。我把它放在了subdiv下。修飾符主要使顏色看起來更平滑。我盡可能多地使用這種技術(shù)，無論是我在2d應(yīng)用中烘烤和使用的基本色彩規(guī)劃，還是混合使用一般的紋理來獲得更復(fù)雜的材料。我也用它來掩飾一些顏色對(duì)靜態(tài)幾何的影響。

The main problem that I was trying to solve was in the texturing and rendering.  I'd like to focus on that here since everything else was straight forward.  First of all, I had lots of objects that I didn't feel like I would want to map and texture, individually (I guess I'm lazy, or just seemed to be overkill doing all that!), so I tried to come up with a way to make groups of materials for certain types of objects.  I made large, generic textures of metal and wood with distinctive features on them, and combined them with vertex colours that I painted on every object (using a bump map as a mask for mix amount).  I used the Vertex Paint modifier for the first time (3ds Max) and I lo-oooo-ved it.  I placed it below the subdiv. modifier mostly; it makes the colours appear smoother.  I use this technique as often as I can, whether for basic colour planning that I bake and use in 2d apps, or mixing with generic textures to get more complex materials.  I also used it for faking some colour bleeding effects on static geometry, too.

你可以看到在最后渲染的窗口周圍的墻上有紅色和綠色的陰影。我在很短的時(shí)間內(nèi)就使用了一些紋理來給很多對(duì)象著色，并且玩得很開心!我這么做的另一個(gè)原因是為了保存一些記憶，因?yàn)槲冶仨氂涀。�我主要是在為�?dòng)畫做一個(gè)場(chǎng)景，我需要每一個(gè)速度的提升，我能做到!(Fig03)

You can see red and green shades on the wall around windows on the final render.  I managed to colour a lot of objects in a really small amount of time using just a few textures - and had fun doing it!   Another reason why I did this was to save some memory as I had to keep in mind that I was primarily doing a scene for animation and I would need every speed boost I could get! (Fig03)

圖 03

在試圖渲染整件事情時(shí)也發(fā)生了類似的問題……我試著避免渲染很多不同層次的幾何圖形(比如角色、前景、背景)，以及任何一種可以通過閃爍和失控來對(duì)抗動(dòng)畫的全局照明解決方案。我當(dāng)時(shí)沒有經(jīng)驗(yàn)，但我現(xiàn)在仍然盡量避免它。

A similar problem occurred whilst attempting to render the whole thing...  I tried to avoid rendering lots of separate layers of geometry (i.e. characters, foreground, background) and any kind of global illumination solutions that could fight with animation by flickering and getting out of hand.  I had no experience with it then and I'm still avoiding it as much as I can now by faking it..

在做了一些奇怪的實(shí)驗(yàn)之后，我決定用簡(jiǎn)單的掃描線和Max的陰影地圖的力量。我用了很多的燈光:2個(gè)主要的聚光燈通過窗戶照射進(jìn)來，并在場(chǎng)景中投射出最強(qiáng)烈的陰影(我希望我當(dāng)時(shí)有現(xiàn)在的viewport陰影預(yù)覽!)其他的燈光則是為了給我留下一個(gè)非常柔和的陰影——我放置了一些來反射房間周圍的陽光撒到地板上;其他的大部分都是泛光燈和點(diǎn)燈，如果用一個(gè)我認(rèn)為好看的角度來照明的話，它會(huì)看起來很漂亮。我也照亮了那些太暗、太無聊的部分，雖然沒有光線可以到達(dá)真實(shí)的效果(例如，冰箱上方天花板上的紫色燈)。

After a few weird experiments with mental-ray GI, I decided to go with simple scanline and the power of Max's shadow maps.  I used lots of lights: 2 main spot lights doing the sunlight through the windows and casting the sharpest shadows in the scene (I wish I had the present viewport shadow preview then!); other lights were meant to create an impression of GI by casting very soft shadows around the scene - I placed a few to reflect the sunlit floor around the room; others were mostly omni lights and spot lights directed at parts of geometry that would just look nice if lit by an angle that I decided would look nice.  I also lit parts of the scene that were too dark and boring, although no light could get there in real life (the purple light on the ceiling above the fridge for example).

所以，我猜，這是一種卡通式的、不切實(shí)際的照明方式吧?特別通過創(chuàng)造色彩對(duì)比溫暖和寒冷，淺色!我的第一個(gè)和唯一的目標(biāo)就是讓它感覺舒適和舒適，而不去想什么是可能的，什么在現(xiàn)實(shí)生活中不會(huì)發(fā)生……Max的半透明材料幫助很大;如果使用陰影貼圖，因?yàn)椴恍枰�的人工制品�?huì)使它幾乎毫無用處，這是很糟糕的，但我想，當(dāng)大量的暗影映射的光線擊中它時(shí)，人工制品會(huì)消失并相互抵消。植物，人物的皮膚，耳朵，牙齒和黃色的塑料鴨子都是很好的例子!我認(rèn)為這種材料應(yīng)該得到更多的關(guān)注:它是一個(gè)非常好的假貨，尤其是在頂部覆蓋了遮擋層的情況下。

So, a cartoonish, unrealistic approach to lighting, I guess?  Especially by creating colour contrasts with warm and cold, light colours!  My first and only goal was just making it feel nice and cosy without trying to think about what's possible and what would not happen in real life...   Max's translucent material helped a lot; it's bad if used with shadow maps due to unwanted artefacts which make it practically useless, but I guess the artefacts disappear and cancel each other out when lots of shadow-mapped lights are hitting it.  Plants, character's skin, ears, teeth and the yellow plastic ducky are good examples of it!  I think that material deserves a little more attention: it's a really good fake, especially with the occlusion layer overlaid on top!

我也經(jīng)常玩反射貼圖。我所做的每一個(gè)模型都有光滑的角邊和反射，使亮點(diǎn)看起來更可愛。最后，我只使用了一張貼圖。jpg有很好的光/暗部和一些綠色的陰影。每一種材料都有它，主要是混合了菲涅耳并且模糊了很多。這有助于定義你的體積，即使它很弱。我想我可以做一個(gè)房間的球形貼圖，然后用它代替……(Fig04)

I played with reflection maps a lot, too.  Every model I did had smooth fillet edges and reflections made the highlights look cuter.  In the end I used just one map, I think: a small .jpg that had good light/dark parts and some greenish shades.  Every material had it, mostly mixed with fresnel and blurred a lot.  This helps defining your volume, even if it's very weak.  I guess I could have made a spherical map of the room and used it instead...? (Fig04)

圖 04

后來，我把掃描線和周圍的封閉通道結(jié)合在一起，這真的產(chǎn)生了很大的不同。這是一項(xiàng)很酷的發(fā)明，它能幫我節(jié)省很多時(shí)間!我渲染了一些動(dòng)畫鏡頭，渲染時(shí)間是可以的，記住我每次拍攝都用了好幾層。掃描線(普通的強(qiáng)大的max掃描線與反射，反射和陰影圖)，mray遮擋，體積(用于太陽射燈)。加上一個(gè)假的DOF通道，這是一個(gè)簡(jiǎn)單的灰度衰減陰影應(yīng)用到整個(gè)場(chǎng)景，并鏈接到一個(gè)輔助對(duì)象的距離。這樣一來，我就可以通過在場(chǎng)景中激活助手的位置來控制焦點(diǎn)，并通過應(yīng)用曲線來進(jìn)行額外的調(diào)整。雖然它可以用來強(qiáng)調(diào)在很多方面的焦點(diǎn)幾何，但它最好作為一個(gè)復(fù)合模糊的面具來偽裝DOF。最后，這張照片經(jīng)過了大量的色彩校正，一些在Photoshop中畫出的亮點(diǎn)讓人感覺更卡通化(Fig05)。

Later I combined the scanline pass with an ambient occlusion pass and that really made a lot of difference.  Occ. is a really cool invention in rendering and it saves my neck often!    I rendered a few animation shots and the render time was OK, keeping in mind that I used several layers for each shot; scanline (regular mighty max scanline with speculars, reflections and shadow maps), mray occlusion, volumetric (for sun spotlights).   Plus a fake DOF channel, which is a simple greyscale falloff shader applied to the whole scene and linked to a helper object's distance.  That way I can control what is in focus by animating the helper's position in the scene, and doing extra tweaks in After Effects by applying curves to it.  Although it can be used to accentuate the geometry in focus in many ways, it's best as a mask for Compound Blur to fake DOF.  Finally, the image went thru a lot of colour corrections and some painted highlights in Photoshop to get the more cartoonish feel (Fig05).

圖 05

這個(gè)角色的頭發(fā)后來變長(zhǎng)了，所以我為他創(chuàng)造了另一層，并為他的牙齒留下了口水。一個(gè)動(dòng)畫是在小分辨率(1280x720)中渲染的，而不是我現(xiàn)在展示的圖像，但是渲染過程和后期工作基本上是一樣的。

The character's hair grew later, so I created another layer for that, and for the saliva hanging from his teeth.  The animation was rendered in smaller resolution (1280x720) than the image I'm presenting now, but the render process and post work was basically the same.

我希望這能對(duì)你們有所幫助!感謝您的閱讀。

I hope this helps a little!   Thank you for reading. 

最終圖像