您需要 登录 才可以下载或查看，没有账号？注册
x
本帖最后由 执此一念 于 2021-12-29 13:06 编辑
水晶船： 创造一个受坏事启发的场景
环境艺术家 Bondok Max 谈到了受 Breaking Bad 启发的 Crystal Ship 项目，分享了工作流程，并解释了如何在没有概念艺术的情况下工作。
Introduction
介绍
Hi everyone, my name is Othman El Bahraoui, I am from Morocco. I’ve been working in the Video Games industry as a 3D Environment Artist since 2007. My first experience was at Ubisoft, where I worked for 9 years and contributed to the creation of several projects such as Prince of Persia, Rayman Origins/Legend, and Child of Light.
大家好，我叫 Othman El Bahraoui，来自摩洛哥。自 2007 年以来，我一直在视频游戏行业工作，担任 3D 环境艺术家。我的第一次经历是在 Ubisoft，在那里我工作了 9 年，并为多个项目的创作做出了贡献，例如波斯王子、雷曼起源/传奇、和光之子。
With the desire to create my own video game, I decided to leave Ubisoft and create a game development studio where we launched our first game on Steam called “The Dark Occult”. 出于创建自己的视频游戏的愿望，我决定离开 Ubisoft 并创建一个游戏开发工作室，在那里我们在 Steam 上推出了我们的第一款名为“The Dark Occult”的游戏。
I'm a self-taught artist and work primarily in Unreal Engine. Everything I've learned has been through the experience of building games and cinematics from talented mentors and colleagues and of course online tutorials. I’m constantly trying to improve and learn more about environments, testing new workflows and techniques. 我是一名自学成才的艺术家，主要在虚幻引擎中工作。我所学到的一切都是通过有才华的导师和同事以及在线教程构建游戏和过场动画的经验获得的。我一直在努力改进和了解有关环境的更多信息，测试新的工作流程和技术。
全部消失的项目All Gone scene is the continuation of the series of works I made lately ( Into the Mountains and Turn Around) in which my goal is to bring art concepts to 3D using mainly Quixel Megascans/Mixer assets and to contribute to the idea of these concepts to improve them even more. All Gone 场景是我最近制作的系列作品的延续（ 进入山脉和 转身），其中我的目标是主要使用Quixel Megascans/Mixer资产将艺术概念带入 3D，并为这些概念的想法做出贡献进一步改善它们。
First, I searched through ArtStation’s feed to find inspiring concepts to work with which contain an interesting theme/story. I found a great piece by Markus Lovadina which caught my attention. This concept covered my desires right on the spot, but what struck me the most was the amount of storytelling you could feel by only looking at it. 首先，我搜索了 ArtStation 的提要，以找到包含有趣主题/故事的鼓舞人心的概念。我发现了 Markus Lovadina 的一篇很棒的作品，引起了我的注意。这个概念当场满足了我的愿望，但最让我印象深刻的是你只看它就能感受到的讲故事的数量。
Before jumping into 3ds Max/Unreal, I always start with a preparation phase to get a clear overview of what is needed and to break down the concept. During this phase, I take time to analyze the concept for any key information that requires extra attention when carrying it over into my scene, taking into account the usage of the Quixel library. 在开始使用 3ds Max/Unreal 之前，我总是从准备阶段开始，以清楚地了解需要什么并分解概念。在这个阶段，考虑到 Quixel 库的使用，我会花时间分析任何在将其转移到我的场景时需要额外注意的关键信息的概念。
This is generally the way I start each project so that I can immerse myself and have a clear overview of all the process stages. It’s very helpful to go into the 3D scene building process with a good vision and plan. 这通常是我开始每个项目的方式，这样我就可以沉浸其中并对所有流程阶段有一个清晰的概述。以良好的愿景和计划进入3D场景构建过程非常有帮助
超级扫描With all the previously mentioned points in mind, I began with checking which assets were available in the Megascans library that would fit the concept. 考虑到前面提到的所有要点，我开始检查 Megascans 库中哪些资产适合该概念。
After that, I started out by making the initial blockout pass in 3ds Max, taking account of the human model for scale reference. 在那之后，我开始在 3ds Max 中制作初始块状通道，考虑到人体模型作为比例参考。
Then I brought it to UE5 to complete the blocking and set up the camera with a very basic lighting setup. It’s mostly about getting an initial feel of the composition and getting an idea of where everything will be. 然后我将它带到 UE5 以完成阻挡并使用非常基本的照明设置来设置相机。这主要是关于获得构图的初步感觉并了解一切将在哪里。
Right after planning all of that and doing the blockout, I jumped straight into modeling, I used 3ds Max for modeling and UV Unwrapping. I’d typically block out the props and use that mesh to create the high poly, and since I would use Nanite, which converts the high-poly mesh into a virtual mesh (meaning there’s no wireframe or polygon count), I didn't bake the low poly for those assets. All of the props were textured with Quixel Mixer. 在规划完所有这些并完成封锁之后，我直接进入建模，我使用 3ds Max 进行建模和 UV 展开。我通常会挡住道具并使用该网格来创建高多边形，因为我会使用 Nanite，它将高多边形网格转换为虚拟网格（意味着没有线框或多边形数量），所以我没有为这些资产烘焙低多边形。所有道具都使用 Quixel Mixer 进行了纹理处理。
I also utilized UE5's built-in Modeling Tools to add other objects like the wood planks on the floor and the piece of paper on the wall. This way was incredibly fast, which enables you to never have to leave the zone by leaning on other programs and apps for some assets such as those. 我还利用 UE5 的内置建模工具添加了其他对象，例如地板上的木板和墙上的纸。这种方式非常快，让您永远不必通过依赖其他程序和应用程序来获取某些资产（例如那些资产）而离开该区域。
I also used this modeling tool to back some of the Quixel Displacement Maps for some decals like the old rag into a mesh displacement. It's very simple and fast to do it. It goes like this: 我还使用这个建模工具将一些 Quixel Displacement Maps 的一些贴花（如旧抹布）支持到网格置换中。这样做非常简单快捷。它是这样的：
- First, enable the editor modeling tools in the plugins settings.
- 首先，在插件设置中启用编辑器建模工具。
- Put the mesh into the level, select it, and then select the modeling
- 将网格放入关卡中，选择它，然后选择建模
- Subdivide the mesh by clicking subdivide on the modeling panel.
- 通过单击建模面板上的细分来细分网格。
- Select "displace" on the modeling panel, and then add the displacement. Set the displacement texture, and press accept again
- 在建模面板上选择“置换”，然后添加置换。设置置换贴图，再按接受
Storytelling in Composition I believe that the way props are placed in a scene plays an important role in the narrative. It gives a story to the environment, and the way the objects are placed and arranged is how this is achieved. 我相信道具在场景中的放置方式在叙事中起着重要作用。它为环境提供了一个故事，而物体的放置和排列方式就是如何实现的。
The first rule is to create a hierarchy. It’s best not to draw attention to all objects at once. Instead, it’s advisable to group them together and arrange them in a way that creates a visual hierarchy. The (Big, Medium, Small) principle is a good method to establish this kind of hierarchy and make it visually interesting at the same time. 第一条规则是创建层次结构。最好不要同时引起对所有对象的注意。相反，建议将它们组合在一起并以创建视觉层次结构的方式排列它们。（大、中、小）原则是建立这种层次结构并同时使其具有视觉趣味性的好方法。
Another approach is to combine different types of objects and have them interact with each other. In this way, none of the objects will look isolated, since everything is visually connected in a harmonized way. 另一种方法是组合不同类型的对象并让它们相互交互。通过这种方式，所有对象都不会看起来是孤立的，因为一切都以协调的方式在视觉上连接起来。
纹理场景Now that most of the assets were added, my point of this environment was to make something where the materials and especially the decals do all the heavy lifting. First, I added the principal materials for the floor and walls, all using the basic material that we get from Bridge. I only adjusted some parameters such as the Albedo intensity and color. 现在添加了大部分资产，我对这个环境的观点是制作一些材料，尤其是贴花，完成所有繁重的工作。首先，我添加了地板和墙壁的主要材料，所有材料都使用我们从 Bridge 获得的基本材料。我只调整了一些参数，例如反照率强度和颜色。
From there I started adding decals in a detailed layering way and using multiple decals on top of each other. The key to doing this is to be subtle. My approach was as if I'm sculpting something: blocking out the form and then continually refining it until I get the fine details. 从那里我开始以详细的分层方式添加贴花，并在彼此之上使用多个贴花。做到这一点的关键是要微妙。我的方法就好像我在雕刻一些东西：将形式分块，然后不断完善它，直到获得精细的细节。
This stage truly enables you to add plenty of storytelling elements to your scene. 这个阶段真正使您能够在场景中添加大量讲故事的元素。
动画I wanted to give the scene a dynamic breeze and make it feel alive. So I created the cockroach and the rat. I made a quick and simple model in 3ds Max for both. 我想让这个场景充满活力，让它充满活力。所以我创造了蟑螂和老鼠。我在 3ds Max 中为两者制作了一个快速简单的模型。
The cockroach is rigged and animated in Unreal Engine directly. 蟑螂是直接在虚幻引擎中操纵和动画的。
The rat was rigged and animated in 3ds Max. I gave it simple animations like idle, walk, run and used the blending animation in the sequencer. 在 3ds Max 中对老鼠进行了操纵和动画处理。我给了它简单的动画，比如空闲、步行、跑步，并在音序器中使用了混合动画。
I kept the process of creating the fur for the rat rather simple as well. First I gathered some references to have a good idea of how the shape of the fur strands looks. Then I checked some online tutorials to learn about the basics of Groom Hair and how to bring it into the Unreal Engine. After that, I started testing what I have learned and was successful in creating a nice-looking rat. 我也保持了为老鼠制作毛皮的过程相当简单。首先，我收集了一些参考资料，以便对毛线的形状有一个很好的了解。然后我查看了一些在线教程，了解 Groom Hair 的基础知识以及如何将其引入虚幻引擎。之后，我开始测试我学到的东西，并成功地创造了一只漂亮的老鼠。
The tutorial below really helped to achieve these results. 下面的教程确实有助于实现这些结果。
I also added a very simple animation to the windows to simulate the movement of the wind. 我还在窗户上添加了一个非常简单的动画来模拟风的运动。
Finally, I added some dust particles to add extra movement to the scene. 最后，我添加了一些灰尘颗粒来为场景添加额外的运动。
Creating a lighting setup for the scene is always an important part. As all the lighting is completely dynamic and using Lumen, the entire process was very intuitive and iterative. I was very impressed when I used Lumen for the first time, as it brings robust dynamic global illumination. 为场景创建照明设置始终是一个重要部分。由于所有照明都是完全动态的并使用 Lumen，因此整个过程非常直观和迭代。第一次使用 Lumen 给我留下了深刻的印象，因为它带来了强大的动态全局照明。
To start with, I added a Directional Light, a Skylight, and Sky Atmosphere. These three did most of the heavy lifting in this scene, especially to achieve the bright light coming from the outside. Moreover, I added a Spotlight to get those shadows cast by the window store on the side walls. 首先，我添加了定向光、天光和天空氛围。这三个人在这个场景中承担了大部分的重任，尤其是为了实现来自外部的明亮光线。此外，我添加了一个聚光灯来获得侧墙上橱窗商店投射的阴影。
Afterward, I placed additional lights (Reclights) to enhance my focal point, make the materials pop, have the colors match the concept, and add some Rim Lights effect to separate the object from the background and make it really stand out and catch the eye. 之后，我放置了额外的灯光（Reclights）来增强我的焦点，使材料流行，使颜色与概念相匹配，并添加一些边缘灯效果将对象与背景分开，使其真正脱颖而出并吸引眼球.
This process stage required a lot of testing and trial and error, but in the end, the scene looked more cinematic which was my main goal. 这个过程阶段需要大量的测试和反复试验，但最终，场景看起来更像电影，这是我的主要目标。
Once I felt the lighting had hit a satisfying level, I started tweaking values in post-processing and enhanced a few things, making some really minor changes but keeping it minimal. I also adjusted the temperature a little bit and did some minor color grading. In addition to that, I upped the Vignette in Image Effects to frame the piece better. 一旦我觉得照明达到了令人满意的水平，我就开始在后期处理中调整值并增强一些东西，进行一些非常小的更改但保持最小。我还稍微调整了温度，并做了一些小的颜色分级。除此之外，我在 Image Effects 中提高了 Vignette 以更好地构图。
Since the reveal of Unreal Engine 5’s impressive demo, I was waiting to discover the technology behind it and the new features they have added and to see how this new system will improve the next-gen games and upcoming movies. 自从虚幻引擎 5 令人印象深刻的演示曝光以来，我一直在等待发现它背后的技术和它们添加的新功能，并看看这个新系统将如何改进下一代游戏和即将上映的电影。
Nanite really saves you time when designing massive amounts of detail and allows you to import film-quality art while maintaining performance. With this new special feature, you can just scan and go. As of Lumen, it allows you to make changes quickly and to work in a seamless way. 在设计大量细节时，Nanite 确实为您节省了时间，并允许您在保持性能的同时导入电影品质的艺术作品。有了这个新的特殊功能，您只需扫描即可。从 Lumen 开始，它允许您快速进行更改并以无缝方式工作。
So far, and even though it is in early access, the engine is much more stable. Another different aspect I like is the new interface: it’s more organized and gives you more space for the viewport. 到目前为止，即使它处于抢先体验阶段，引擎也更加稳定。我喜欢的另一个不同方面是新界面：它更有条理并为您提供更多的视口空间。
| 
