Train Pavilion Aerial

由专筑网韩平，刘庆新编译

Portfolio Vol. 4中还有几个地方空缺。恰巧我最近完成了一个火车站鸟瞰图制作，这张图很好地展示了建筑与城市标志性的水塔之间的关系。我制作了几个分析图来表达它们之间的关系，但是没有实际的图片展示这座水塔真实的样子。我制作了一个鸟瞰渲染图来联系之前作品集中项目的各个不同部分。

There are still some areas of Portfolio Vol. 4 that are missing information. One image that I just finished creating is an aerial of the train pavilion project that better shows the relationship of the new architecture to the iconic town water tower.  I have several diagrams describing the relationship but no realistic imagery showing what the actual tower looks like. I figured an aerial rendering could help tie together the many different pieces of the project described in earlier pages of the portfolio.


上图：SU线稿/Above: Sketchup Line Work

像这样的图片制作比较困难，因为涉及许多不同的元素：众多的建筑、汽车、地面铺装等。幸运地是在我之前的渲染图制作中，我已经基本完成了城市模型的建模工作。但是，我还是需要在项目附近建造一些住宅模型，增加一些树和汽车的代理。如果你对代理不熟悉，有一个利用V-Ray渲染大型配景的方法，比如汽车、树木，用这个方法我们不需要在SU中放置详细的多边形模型，只需要放置一个简化的低聚线框模型。这意味着你可以放置成千上万的树和汽车，但不会增加你整体模型的大小，同时在你渲染的过程中还能充分利用电脑内存。

Images like this are difficult because there are so many different elements involved in large quantities: Lots of buildings, trees, cars, ground plane textures, etc. Luckily, I had much of the town already modeled from previous renderings. However, I still needed to model some houses in the distance and add proxy trees and cars. If you are not familiar with proxies, there are a way for V-Ray to render heavy meshes like cars and trees without actually placing these high poly objects in the Sketchup model. Instead, only a simplified low poly wire frame placeholder of the object is used in the Sketchup model. This means you can load hundreds of trees and cars without ballooning up your model size or using up all of your ram during the render process.


上图：SU贴图后的SU模型/Above: Sketchup Model with Textures

该效果图的另一个难点是地面层的制作。地面上的贴图发生了很多变化，但我没有时间去建模。因此，我使用了一个航拍地图作为地面层。我对图片的局部进行了修改，去除了汽车和建筑的阴影。航拍地图的分辨率不是很高，到那时因为镜头设置在空中比较高，因此有些细节也并不重要。在上面的图片中，我们可以看到航拍地图应用到地面层的情况。

Another difficult aspect of this image was the ground plane. There are lots of changes in textures happening and this was not something that I had time to model. Instead, I ended up using the aerial map as the ground plane. I modified parts of the image to remove cars and building shadows. The aerial map was not at a high resolution, but since the camera was so high in the air, this lack of detail didn’t matter much. In the image above, you can see the aerial map applied to the ground plane.


上图：V-Ray 基础渲染图/Above: V-Ray Base Rendering

虽然这张V-Ray 基础渲染图中还有很多地方需要在PS中进行清理，但是至少一开始我就获得了一个不错的树木和背景建筑图。在PS中处理所有的事物当然是可行的，但是这也将耗费大量的时间，也会不够精确和清楚。

The V-Ray base rendering still has a lot of things that need to be cleaned up in Photoshop, but I at least have a good start in terms of trees and context buildings. Trying to Photoshop in all of this stuff would have been doable, but would have take a ton of time and would not have been as accurate or clean.


基础渲染图上的初步PS处理/Initial Photoshop over base rendering

在PS中，我首先在景观部分，增加了一些树木、沿着基地增加野草，调整水池的色调和反光。接着，我进行了背景建筑的处理，在主干道的建筑立面上增加一些细节，比如为火车站增加一些纹理、建筑灯光等。最后，我增加一些细节，比如人、交通灯、水塔的名字以及其它一些调整等。

Once in Photoshop, I first worked on the landscape adding some more trees, wild grasses around the site, and adjusting the tone and reflections of the reservoirs. I then moved to working on the context buildings, bringing in facade details to the main street buildings, adding textures to the train pavilion, and lighting up the buildings. Finally, I added details like people, crossing lights, water tower text, and other misc. adjustments.


上图：最后的色彩调整/Above: Final Color Adjustments

我进行了简单的色调调整，在远处利用轻微的大气雾霾，加一些黄色和蓝色的色彩层，还利用Topaz Labs调整了一些色调和细节。

I kept the color adjustments simple with a slight atmospheric haze in the distance, some yellow and blue color overlays, and some toning and detail adjustments made in Topaz Labs.








  
最后需要注意的几点，我是根据折叠后对图片影响最小来选择视角的。尽管这张图片是鸟瞰图，而且所有的细节都非常小，但是我还是需要确保我图中设计的所有重要的信息不会在折叠后消失。另外，我渲染的所有东西都采用了高分辨率（高于平常的6,500 px宽），所以读者可以近距离地看到这些细节。因而，我们的PS文件有3.5GB之大，所以在我的笔记本上不太容易操作。

Some final notes, I chose the view based on where the fold of the spread would have minimal impact. Since the image is an aerial and all of the detail is so small, I had to make sure the view was designed in such a way that none of the important information was getting lost in the fold. I also rendered everything at a higher resolution than normal (6,500 px wide) so that the detail remains sharp as viewers get close to the page. This meant a very large Photoshop file of over 3.5 GBs though still easily manageable on my laptop.


出处：本文译自visualizingarchitecture.com/，转载请注明出处。