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 秋令6689 于 2022-5-9 15:46 编辑
The Launch – Environment Breakdown – Pawel Ptaszynski 发射 – 环境崩溃 – 帕维尔·普塔辛斯基
Introduction Hello, my name is Pawel Ptaszynski. I’m from Poland but currently living in the UK. I’m a self-taught 3D artist, currently working as a 3D Prop Artist at Rocksteady on Suicide Squad. Previously, I was at Frontier Developments for over 5 years working on Elite Dangerous and Planet Coaster. 介绍 你好,我的名字是Pawel Ptaszynski。我来自波兰,但目前居住在英国。我是一名自学成才的3D艺术家,目前在Rocksteady的Suicide Squad担任3D道具艺术家。此前,我在Frontier Developments工作了5年多,负责Elite Dangerous和Planet Coaster。 What goals did you have in mind when making the Environment? I wanted to make something in Unreal Engine 5 using Lumen and play around with Megascans. It was very important for me to have fun while working on a personal piece, and being a big fan of near future and post-apocalyptic aesthetics, I was looking for very specific concept art. I found an amazing artist, , and one of his checked all the boxes. 在制作环境时,您有什么目标? 我想在虚幻引擎5中使用Lumen制作一些东西,并玩Megascans。对我来说,在创作个人作品的同时玩得开心是非常重要的, 作为近未来和后启示录美学的忠实粉丝, 我正在寻找非常具体的概念艺术。 What software did you use? My main 3D modeling software of choice is Blender. For texturing I used Substance Painter and Photoshop for creating more specific masks etc. I used Zbrush for my boolean-to-dynamesh technique (explained ). I used Megascans for foliage, Metahuman for characters, and finally brought everything together in the UE5 Engine. 您使用了什么软件? 我选择的主要3D建模软件是Blender。对于纹理,我使用Substance Painter和Photoshop来创建更具体的蒙版等。我使用Zbrush作为我的布尔到dynamesh技术(解释)。我用Megascans来制作树叶,用Metahuman来制作角色,最后在UE5引擎中把所有东西都整合在一起。 Planning 规划 For organization and better progress tracking I used Trello. It really helped me to focus on individual work and remember what there’s left to do. I started with the concept on my secondary screen and broke it down into each element that I would need to work on, creating cards in Trello for each task. They were not set in stone and I modified them as the project was progressing. 为了组织和更好的进度跟踪,我使用了Trello。这真的帮助我专注于个人工作,并记住还有什么要做的。我从辅助屏幕上的概念开始,并将其分解为我需要处理的每个元素,在Trello中为每个任务创建卡片。它们不是一成不变的,随着项目的进展,我修改了它们。 I started by creating UE5 project with a preset landscape. I placed the sun in the correct position and added any other necessary components like Exponential Height Fog, Post Processing etc. 我首先创建了带有预设景观的UE5项目。我将太阳放置在正确的位置,并添加了任何其他必要的组件,如指数高度雾,后处理等。 References After finishing my Trello list and setting up my UE5 scene, I started working on gathering references for everything I thought I would need. Some things were added later, for example, wind turbines, which weren’t on the concept, but felt like a nice addition. 引用 在完成 Trello 列表并设置了 UE5 场景后,我开始为我认为需要的所有内容收集参考资料。后来添加了一些东西,例如风力涡轮机,它们不在概念上,但感觉像是一个不错的补充。 I planned it out so everything except the city buildings would be Highpoly modeled and bespoke textured. Some assets, like the food truck, were “force perspective” which meant they are only supposed to be seen from certain angles. For more efficient workflow, I decided to focus on the areas included in the camera shot, such as the front of the food truck. The fact that the scene was at night with a lot of shadows meant I was able to hide stuff that was never meant to be seen. 我计划好了,所以除了城市建筑之外,所有东西都会被高聚建模和定制纹理。一些资产,如食品卡车,是“力量视角”,这意味着它们只能从某些角度看到。为了提高工作流程效率,我决定将重点放在相机镜头中包含的区域,例如食品卡车的前部。事实上,场景是在晚上有很多阴影,这意味着我能够隐藏一些本来就不应该被看到的东西。 Blockout phase 封锁阶段 I started modeling by blocking out the buildings for the background of my scene. The buildings were split to 4 categories: wo massive buildings that were the key focus of the city, large buildings that were placed by hand in specific places, small buildings which I used with foliage paint in UE to populate the city, medium which I used to support and break the silhouette of the small ones. 我开始建模,将建筑物作为我的场景背景。这些建筑分为4类: 巨大的建筑是城市的重点, 手工放置在特定位置的大型建筑物, 我在UE中使用树叶涂料来填充城市的小建筑物, 用来支撑和打破小的轮廓的中。 At the very end of production, I unwrapped all of them and used a tileable texture to give them emissive lights and roughness response where it was visible. 在制作的最后,我打开了所有它们的包装,并使用可平铺的纹理为它们在可见的地方提供发光和粗糙的响应。 Over the next couple of days, I spent my time populating the scene with foreground details. I created all the necessary blockouts from my list. My goal was to spend little time on each one, but position them accurately and with correct scale. 在接下来的几天里,我花时间用前景细节填充场景。我从我的列表中创建了所有必要的阻止。我的目标是花很少的时间在每一个上,但要准确定位它们,并具有正确的比例。 For the rocket launch smoke I wanted to experiment with smoke physics in blender and converting it into a solid mesh, which in hindsight didn’t work out. I was not satisfied with the result and I could just sculpt it from scratch but oh well… It worked for a blockout ��� 对于火箭发射烟雾,我想在搅拌机中尝试烟雾物理学,并将其转换为固体网格,事后看来这并没有成功。我对结果不满意,我可以从头开始雕刻它,但是哦,好吧......它为封锁工作 In addition to using the smoke physics as a blockout, I was able to re-use it as a 2d transparent texture for some of the smoke in the scene. 除了使用烟雾物理作为遮挡之外,我还能够将其重新用作场景中某些烟雾的2d透明纹理。 Megascans 巨型扫描 It was time to start gathering Megascans for the scene. I got different foliage, small and big rocks, materials for the ground, and puddles. It was a very fun step of the project and it was evolving throughout. 是时候开始为现场收集Megascans了。我有不同的树叶,大大小小的岩石,地面的材料和水坑。这是该项目中非常有趣的一步,并且在整个过程中都在不断发展。 megascans Progressing through the scene 在场景中前进 Through the next couple days I slowly built various things simultaneously. I spent some time on lighting – setting up different colored lights on the left side, adding some more lights inside the city, and lighting up areas which were too dark with artificial light. For example: lower right part of the scene. At the same time, I was also adding more detail to blockout models one by one. It was time to start doing some highpoly modeling. 在接下来的几天里,我慢慢地同时构建了各种东西。我花了一些时间在照明上 - 在左侧设置不同颜色的灯,在城市内部添加更多灯,并用人造光照亮太暗的区域。例如:场景的右下角。同时,我还为逐个遮挡模型添加了更多细节。是时候开始做一些高多的建模了。 Food truck 食品运输车 The food truck became the biggest asset in this scene, not because it had to be, but because I had the most fun making it. I spent a lot of time looking for references because I didn’t want it to look like any modern food truck, but like something from the ’50s. After some time I found a Russian company that was making and renting these amazing food vans. 食品卡车成为这个场景中最大的资产,不是因为它必须如此,而是因为我最有趣地制作了它。我花了很多时间寻找参考资料,因为我不想让它看起来像任何现代食品卡车,而像50年代的东西。过了一段时间,我找到了一家俄罗斯公司,该公司正在制造和租用这些令人惊叹的食品车。 Following the main concept and implementing my own ideas, I wanted it to look run down but still operational; flat tires indicating it was placed in this spot at some point a long time ago, but but it is slowly breaking down. I wanted it to look maintained with old and new technologies together, like the post-modern AC unit, contrasted with a futuristic power generator and waste unit. It now serves as stationary street food point. When it eventually breaks it will be left alone, with no one to clean it up. 遵循主要概念并实现我自己的想法,我希望它看起来破旧但仍然可操作;轮胎漏气表明它在很久以前的某个时候被放置在这个地方,但它正在慢慢分解。我希望它看起来与新技术一起保持,就像后现代的空调机组一样,与未来派的发电机和废物机组形成鲜明对比。它现在用作固定的街头食品点。当它最终破裂时,它将被单独留下,没有人来清理它。 I also wanted to add some bullet holes as this is in the middle of some old road in the wastelands, so I did quick sculpts and rendered them in blender using the Z information using 我还想添加一些弹孔,因为这是在荒原上的一些老路中间,所以我做了快速雕刻,并使用Z信息在搅拌机中渲染它们。 AC unit, power generator, waste unit etc. each module needed some extra love so I gathered references and modeled them one by one. 交流装置,发电机,废物装置等每个模块都需要一些额外的爱,所以我收集了参考资料并逐个建模。 For the cloth pipes I’ve used SimplyCloth in Blender. 对于布管,我在搅拌机中使用了SimplyCloth。 Overall the whole asset was a usual hardsurface modeling process with retopology, Uvs and texturing in Substance Painter. 总体而言,整个资产是一个通常的硬表面建模过程,具有物质画家中的重新拓扑,Uvs和纹理。 I made two LCD screens in photoshop, one showing live news about the rocket launch, and the other further in the background was just an advert. Both screens were old with dead pixels. 我在photoshop中制作了两个LCD屏幕,一个显示有关火箭发射的实时新闻,另一个在背景中只是一个广告。两个屏幕都是旧的坏像素。 Finishing up rest of the models 完成其余模型 I finished up rest of the models with typical hardsurface workflow, making highpoly for each one, baking them on lowpoly Uvs and texturing in painter. 我用典型的硬表面工作流程完成了其余的模型,为每个模型制作高额,在低聚Uvs上烘烤它们,并在painter中纹理化。 I took the old smoke blockout and sculpted more details to make it more “explosive” 我拿了旧的烟雾块,雕刻了更多的细节,让它更“爆炸性” The City 城市 When making the city I wanted an effect of pollution and fog in the center of it. Simple point lights and procedural fog wouldn’t work because of the scale of the city (it was the size of an actual city). I did some quick tricks with gradient textures, and used a simple alpha circle texture for fog and pollution and a second texture for beam lights. 在建造城市时,我希望在城市中心产生污染和雾的效果。简单的点光灯和程序雾是行不通的,因为城市的规模(它是实际城市的大小)。我用渐变纹理做了一些快速的技巧,并使用了一个简单的阿尔法圆圈纹理来处理雾和污染,并使用第二个纹理来表示光束光。 I modeled and textured a quick and simple tileable city wall texture, as I wanted to avoid plain looking blocks. 我建模并纹理化了一个快速而简单的可平铺城墙纹理,因为我想避免看起来很普通的块。 Characters 字符 Characters weren’t the main focus of this scene, and because I’m not an character artist I was on the fence if I should include them in the scene or not. In the end I’m happy I did because they add a bit of variety. My first plan was to photoscan myself into the scene which was very exciting because I’ve never done any photoscaning before, and the idea of putting myself in a 3d scene was very cool. 角色不是这个场景的主要焦点,因为我不是角色艺术家,所以我是否应该把他们包括在场景中。最后,我很高兴我做到了,因为他们增加了一点多样性。我的第一个计划是将自己的照片扫描到场景中,这非常令人兴奋,因为我以前从未做过任何照片扫描,并且将自己置于3D场景中的想法非常酷。 After a couple of attempts I quickly figured out it was a lot harder than I thought to scan a whole human outside of a studio setting… Many different artifacts, missing geo and very long wait time for everything to calculate made me give up on the idea. I’m still happy that I tried it out because I’ve learned something new – what to do and mostly what not to do… I want to try more photoscanning in the future with smaller objects but that was the end of it for this scene. 经过几次尝试,我很快发现,在工作室环境之外扫描整个人类比我想象的要困难得多......许多不同的工件,缺少的地理和很长的等待时间让我放弃了这个想法。我仍然很高兴我尝试了一下,因为我学到了一些新东西 - 做什么,主要是不该做什么......我想在未来尝试使用较小的物体进行更多的照片扫描,但这就是这个场景的结束。 My plan B was to use Metahumans, which came with it’s own challenges like rigging them into position. But because of Unreal Engine implementation it was quite easy to put them in the scene. Metahumans are well documented on Epic website. 我的B计划是使用超人类,这带来了它自己的挑战,比如操纵他们到位。但是由于虚幻引擎的实现,将它们放在场景中非常容易。超人类在Epic网站上有很好的记录。 One of the things that was missing for me was a road next to the food vans. After all, how did they get there? Just sitting on a desert in the middle of nowhere didn’t make a lot of sense. I grabbed a material from Quixel Bridge of an old road, created a simple mesh and used UE spline to put it in the desired space. 对我来说缺少的一件事是食品车旁边的一条路。毕竟,他们是怎么到达那里的?只是坐在偏僻的沙漠上并没有多大意义。我从一条旧路的Quixel Bridge中抓取了一个材料,创建了一个简单的网格,并使用UE样条曲线将其放置在所需的空间中。 Using a spline I created and added some wires running on the ground through the food vans somewhere out of camera. 我使用样条线创建并添加了一些电线,这些电线穿过摄像机外的食品车在地面上运行。 All that was left to do is to place cameras for different shots for Artstation and tweak Post Processing and minor placement of assets. Foliage and other elements was an ongoing process throughout. I was often taking screenshots of daily progress and self-reviewing them by looking at each element. 剩下要做的就是为Artstation的不同镜头放置摄像机,并调整后期处理和资产的次要放置。树叶和其他元素是一个持续的过程。我经常截取每日进度的屏幕截图,并通过查看每个元素进行自我审查。
|