这是一篇高级教程，其中蕴含了很多的设计理念和设计技巧，是一篇不可多得的好文章。在这个教程中，作者将结合了各种方法技巧，完成了一个时尚而富有视觉冲击力的作品——“骑乘在音乐之上”。

最终效果：

译者按：

这是一篇高级教程，不仅最终效果时尚绚丽，负有强大的冲击力。而且，设计思路阐述明了，讲解详细而系统，技法繁多而实用。其中，创建透视、纵深感、视觉冲击力的技巧无不有所提及，还涉及了一些作者的个人技巧 。另外，多个软件平台的配合使用，更是体现了作为高级设计师的实力！当然，正如作者自己所言，很多在其他软件中的效果在Photoshop中一样可以实现，只是可能麻烦些。所以，不要被教程中的Illustrate和Modo所吓倒。

我知道，这篇教程字数众多，可能会让你心生厌倦，但是，我保证耐心终究会换来不菲的成果。请你一定要像作者那样，不怕困难与麻烦，反复尝试，坚持到底。这样一定会做出好的作品！

如果，教程中有翻译不当之处，欢迎添加批注。

如果，教程中你有什么不明白的地方，请留言，我一定尽力帮助解决。

如果，你认为这篇教程很有价值，请“顶”一下，这是对我的鼓励！

谢谢！

Step 1 – Preparing the canvas

步骤 1 - 准备

Before we begin there are a few things worth mentioning. All the photos in this tutorial belong to the author, Pirosca Marcel. They were not gathered from stock sites around the net. The main character was shot in my photo studio.

在我们开始之前，有些事情是需要注意的。这篇教程中所用的全部照片，其版权属于作者Pirosca Marcel。它们并不是从网上的图片网站收集而来，其主要人物的照片都拍摄于我的照片工作室。

First off I opened the original image resulted from the shooting with the main purpose being isolating the character and the shadows on the floor. First thing I noticed was that his face was not expressive enough for what I had in mind.

首先，我打开拍摄的源照片，这张照片主要是为分离人物和地面上影子而拍摄的（这里是说，作者需要一张易于扣取人物及其影子的照片——译者注）。对于这张照片，我注意到的第一件事是，他的脸不怎么符合我的期望。

Before we begin cutting out the character I needed to replace his head with one from another picture where he looked more optimistic. I always shoot a lot of pictures every session, so when situations like this arise I have plenty of other images to choose from. The most important thing is that the light sources must be in the same place so it integrates seamlessly and believably. The light needs to come from the same directions. I have found the image below to be a good match for what I needed.

在我们将人物扣取出来之前，我需要将这个脑袋换成另一张照片中看起来更为理想的脑袋。我总是在每一次的拍摄中，拍摄大量地照片，这样，当情况更适合让脑袋抬起来时，我就会有许多选择的余地。还有，最重要的是，两张照片的光源必须一致，这样脑袋和身体才会真实可信地合成在一起。

Using the Polygonal Lasso Tool create a selection around his head. The background contains only a flat color, so the selection should be very easy to make. I selected his head just as in the image below, then copy/pasted it into the original image, on a new layer.

在脑袋的周围，使用多边形套索工具建立一个选区（使用这种方法需要连续点击鼠标，以确保选区的平滑与准确——译者注）。因为照片的背景颜色单一，所以选区十分容易建立。如下图一样，我将脑袋用选区选出来，然后将它拷贝粘贴到原图的新图层上。

I scaled and rotated it so that it matched the old head.

我将新的脑袋进行了缩放和旋转，以使它与旧脑袋一样和身体相匹配。

（图释：移动到旧脑袋上）

As you can see some integration is needed to be done for the head to blend perfectly. First off, cut out the background head so we can work on our new one. Use the Polygonal Lasso Tool to select and delete it. We should now have something similar to what we see below.

就像你所看见的，这儿需要做一些合成，来使脑袋完美地与身体融合在一起。首先，删除背景中的旧脑袋，这样我们能更好地在新脑袋上作调整。使用多边形套索工具选择并删除旧脑袋。现在，我们得到效果应该与下图相似。

Notice that the most important region to work on is his neck. In order to obtain a nice transition from one to the other we need a transparency mask on the new head, then with a relatively soft (50%) Eraser Brush start deleting some of the harsh edges on the neck until you reach a good result. This part is open to a lot of trial and error, so using a mask will help us recover some of the parts we may have accidentally deleted. Finally, you should obtain something similar to the next image.

注意了，脖子是最重要的修饰区域。为了在新脑袋和身体间得到一个漂亮的过渡，我们需要在新脑袋上做一个透明蒙版，然后用比较柔和的画笔（硬度50%）开始将脖子上一些粗糙的边缘擦除掉，直至你得到了一个满意的结果。在这一阶段，我们需要反复尝试修改，而使用蒙版会帮助我们恢复我们意外擦除的部分。最终，你得到的效果应当与下图类似。

（图释：①左上：因为旧脑袋被删除，所以他这里的汗衫被截断了；②考虑到他身后有光，这里就显得太暗了）

As we can see we need to work on some of the details. We need to fix the shirt and neck lighting in order for him to be believable. For the neck I used a combination of Dodge/Sponge tool as illustrated below.

如我们所见，我们需要做一些细节修饰。我们需要修正汗衫和脖子上的光，以让他显得真实可信。如下图所示，在脖子上，我混合使用了减淡工具和海绵工具。

（图释：用减淡工具给这边的脖子增加高光。然而不幸的是，这会使颜色过渡饱和）

（图释：使用海绵工具降低这个区域的饱和度，让它融合得更好）

In order to fix the shirt I copy/pasted a part of his shirt and then I blended it (using curves and transparency mask).

为了修正汗衫，我复制粘贴了他汗衫的一部分，然后将它与修正部分融合（使用曲线和透明蒙版）。

（图释：复制粘贴他的部分汗衫）

（图释：移动到适当位置，以使线条连续）

（图释：用曲线使这个区域变得更亮些，因为这里之前有些暗）

（图释：用柔和的画笔涂抹，使其边缘融合）

（图释：最终效果）

We are now ready to isolate our guy and the shadows on the floor. As he is a hard edged surface against a flat background it should be fairly easy to select him using the Polygonal Lasso Tool. I personally prefer this tool over the paths approach to selection because I don’t have to adjust any control points and tangents, even though it means a lot more clicks. Copy/paste him into a new layer.

现在，我们将从照片上分离出我们的人物和地上影子。因为相对于单一的背景，他有一个边缘分明的外观，所以，使用多边形套索工具选取他是十分容易的。我个人比较喜欢用这个工具，而不是将路径转换为选区的方法，因为这样我就不需要去调整控制点和切线，尽管这中方法意味着更多地敲击鼠标。接着，复制粘贴他到一个新的图层。

Next thing we want is having the shadow on a different layer. As you can see the background of the photo studio was kind of dirty, so in order to remove the stains I applied some Surface Blur. I found that this type of blur is best for this operation as it removes noise while preserving some detail.

下一步，我们希望在一个不同的图层中提取影子。就像你看到的，照片工作室的地面有点脏，所以为了移除污渍，我使用了一些表面模糊。我发现这种模糊滤镜十分适用于这种只移除噪点而保留细节的操作。

Once we got to this stage we can select the shadow using Color Range and the following settings.

一旦到了这一步，我们就能使用颜色范围来选取影子了。请参照如下参数。

Of course there will be some unwanted regions in the selection. Hit OK, and after the selection was made enter Quick Mask mode. Proceed to erase the unnecessary parts, as depicted in the image below.

当然，在选区中会出现不想要的区域。点击确认，出现选取后进入快速蒙版模式。像下图显示的那样，开始擦除不需要的部分。

Now in order to have the exact shadow we need to subtract the contour of our guy from the remaining selection. While in Quick Mask mode Control-click the thumbnail of the main character layer in the layers palette. This will load the selection for that layer. Press Delete to erase the quick mask inside the selection.

为了得到多余阴影的选区，现在我们要从现有选区中减去人物的轮廓。保持快速蒙版模式，同时在图层面板中，按住CTRL键点击那个包含了提取人物的图层的缩略图。这会读取那个图层的选区。按Delete键删除选区中的快速蒙版。

Exit Quick Mask mode and we have the selection of the shadow. Just copy/paste it onto a new layer.

退出快速蒙版模式，我们得到了影子的选区。将影子复制粘贴到一个新的图层。

（图释：一旦影子被分离出来，背景色就会将影子突显得更为清晰）

We now have the character and his shadow isolated and on separate layers. We are ready to begin the final image.

现在，我们已经将人物和他的影子在不同的图层上分离出来了，准备开始制作最终图像吧！

Step 2 – Composition

步骤 2 - 合成

This was a HUGE piece of work, size related. In order to achieve the final image we need to create an image 3613 by 5000 pixels. To get an idea of what I actually wanted, first I took the main character and moved it around the canvas; finally deciding to place him on the lower half of the image following the flow of his body. Being on a separate layer meant I could add all sorts of details behind him. The first thing that needed to be done was to further differentiate him from the gray background.

这是一个庞大的作品，我是说尺寸庞大的作品。为了得到最终的图像，我们要建一张3613×5000的画布。为了弄明白我到底想要得到什么，我将人物在画布上到处移动。最终，我决定顺从他身体的趋势将其放在图像中间偏下的位置。使用分离的图层，意味着我能够将所有种类的细节放在他身后。我们首先要做的事情是，将他从灰色的背景中区别出来。

Use a white, soft brush with 20% Opacity to create a glow behind the guy.

使用不透明度20%的白色软画笔，在人物身后创建一些光芒。

Next it’s time to determine how the lines in the image would flow. For this I used my tablet with the size controlled by the pen pressure and Opacity set to about 30%. I proceeded in drawing the lines on which our man was surfing by holding Shift (so that the brush draws a straight line) and pressing harder on the tablet (for the line to be thicker) at his feet and then pressing less as the lines faded away.

接着是决定图像中线条该如何流动的时候了。为此，我使用上了我的数位板，以便能使用压感笔的压感来控制线条的粗细。另外，要将不透明度设置在大约30% 。我开始绘制人物冲所骑乘的线条，绘制时要按住SHIFT键（这样可以绘制出直线），并且绘制到人物脚部时加大压力（这样绘制的线条更厚重），绘制到末端时减轻压力。

This gave the lines the feeling of perspective and also helped me to give a sense of depth. I have to add that this is a process prone to a lot of trial and error, so it might take a while until you are satisfied with the results.

这种绘制方法给线条以透视感，并且帮助我制造纵深感。对此，我不得不多说一句，这是一个需要反复尝试修正的过程，所以，在你满意之前，恐怕要花上些时间。

Now it gets interesting. In order to be able to replicate all of the planes fading away and also maintain proper perspective I used the Vanishing Point Filter to eyeball the plane our man is surfing on. There is an entire discussion here, because you can’t define that plane exactly how it is in reality since you have no reference. After several attempts I settled for the one below, thinking it works. To do this go to Filter > Vanishing Point.

现在，制作开始变得有趣了。为了能够复制出所有的消失平面，也为了保持正确的透视关系，我使用了消失点滤镜来观察人物所骑乘的平面。在这一步上我们需要充分的考虑，因为你没有参照物来确定这个平面在现实中究竟是怎样的。经过几次尝试后，我如下图那样做了设置，觉得这应该可行。做这一步设置，要执行 滤镜>消失点 。

The deal with Vanishing Point is that these planes can be extended at 90 degrees or arbitrarily. Using this method I further constructed additional planes to help me keep track of perspective, while trying to match my sketch lines as close as I could.

用消失点来处理这些平面的话，可以将平面以90°或任意角度延展。以这种方法，我构建了更多的平面，以帮助我在尽可能地匹配线稿的同时，保持透视关系。

With this setup in position I was now able to add elements that will be flowing along these planes. But first I had a point to settle.

使用消失点来配置各平面的位置，可以让我能够顺着平面变化的趋势来增加元素。但是，在这之前，我还有一点需要准备。

Somewhere along the line I experimented with color variations since the gray seemed a little dull. I love color and decided to use its power to give it a little punch.

因为灰色看起来有些死板，所以我在线条所在的一些地方增加了色彩变化。我喜欢色彩，并且要用它给画面带来些冲击力。

Step 3 – Color Scheme

步骤 3 - 配色

In order to obtain the proper colors for this piece I used three adjustment layers and placed them on top of the image. Two of them affecting only the skin (using masks), and a third one to set the general mood. For the skin, a Levels and a Color Balance to add subtle color enhancement.

为了在画面上得到理想的颜色，我在图像顶端使用了三个调节图层。其中两个只作用于人物皮肤（使用蒙版），而第三个用于调节整体色调。在皮肤上，用色阶和色彩平衡微微提升些色调。

And then a Color Balance affecting the whole image to give an overall greenish mood. The settings are shown below.

然后，用一个色彩平衡调节图层给整个画面蒙上一些绿色调。设置如下。

We are now ready to add in the details.

现在，我们可以开始添加细节了。

Step 4 – Digital Volume

步骤 4 - 添加数码音量图

Open Illustrator and by copy/pasting and then scaling vertical rectangles you can obtain something similar to the first digital volume in the image below. Make a copy and move it below.

打开 Illustrator ，建立一些垂直的长条，复制粘贴它们，并调整长度，然后，你会得到和下图中类似的数码音量初稿。复制并向下移动。

By adding horizontal rectangles and subtracting them, it’s easy to obtain the second one. Just create one rectangle, move it vertically (by holding down the Alt key to create a copy) and press Command + D repeatedly to multiply it.

创建一些水平的长条，用它们减去初稿，我们就很容易地得到了第二稿。这只需创建一个长方形，水平复制一份（按住ALT键复制），然后按COMMAND+D重复复制。

Copy the second volume graphic and go to Photoshop.

拷贝第二稿，进入Photoshop中。

Now paste it into the Vanishing Point filter. For some reason Photoshop doesn’t allow directly pasting paths from Illustrator there. So we must paste it on a new layer as pixels first. Rotate it and scale it vertically just as in the image below in order to have as much resolution as we can.

现在将它粘贴到消失点滤镜中。因为一些原因，Photoshop的消失点滤镜并不支持直接从Illustrator粘贴路径。所以，我们必须先将它作为像素粘贴到一个新的图层中。像下图一样旋转并垂直缩放到尽可能大的分辨率。

Then Command-click the layer thumbnail in the layers palette. This will bring up the selection of that layer. Press Command + C to copy it into the cache, then delete the layer. We can now move to the Vanishing Point filter.

在图层面板中，按住COMMAND键点击这个图层的缩略图，提取其中的选区。按COMMAND+C将其复制到缓存中，然后删除图层。现在，我们要转而使用消失点滤镜了。

Go to Filter > Vanishing Point and paste. You should have something like that shown below.

执行 滤镜>消失点 ，粘贴。你应该能得到下图效果。

Now if you drag that selection onto the floor plane, Photoshop will automatically calculate the right perspective for you.

现在，如果你将选区拖入表示地面的平面中，Photoshop会自动为你计算出正确的透视关系。

Do it again to extend the volume further back.

重复同一步骤，在背后延长音量图。

Go to blending options and add the styles shown below.

进入混合选项，添加下图中的图层样式。

You should now have something looking like this.

你现在应该得到了与这类似的效果。

In order to to have some of the squares transparent, like in the final image, create a Transparency Mask for this layer, and then use a 50% gray (#7a7a7a) to paint over the squares that you want transparent. This can be a tedious process and it is up to you to decide which ones you leave opaque and which not. I will illustrate.

为了像最终效果那样有一些透明的方块，要在这个图层上创建一个透明蒙版，然后使用50%的灰（#7a7a7a）去涂抹那些你想使其透明的方块。这可能会是个无聊的过程，

（图释：①中部：白色不透明区 ②灰色透明区）

Now to give it a little punch, duplicate that layer, set its Opacity to 68%, and apply the following Curves (Command + M) to its transparency mask and you get this. The gray in the mask should have turned to black so this layer affects only the opaque squares.

现在，给数码音量图添加一些视觉刺激。复制图层，设置不透明度为68%，并且对透明图层应用下图曲线（COMMAND+M）。然后，你得到了这个。蒙版中的灰色应当变为黑色，这样图层效果就只会影响不透明区。

It should look something like this.

效果应该看起来是这个样子。

We have reached a delicate part which takes a lot of tweaking to get right, but the main idea is to use a copy of the volume, move it into place and then use the Warp Tool to simulate the curvature. The tricky part is that some of the lines may get curved the wrong way and I had to personally correct them. Fairly time consuming I must say, but no other idea to create that corner came to mind.

我们到了一个需要优雅地处理多处弯曲的步骤，但是，其核心方法不过是拷贝、移动和使用变形工具模仿曲线。然而，这里令人头痛的是，有些线条会被错误地弯曲，我不得不手工修正它们。我必须声明，在这里会消耗大量的时间，但是我的确没有想到其他制作曲线的好方法。

Using the same technique as before, create a transparency mask and block out some transparent squares using a 50% gray.

用与之前相同的方法做一个透明蒙版，用50%的灰遮盖透明方块。

（图释：用蒙版创建一些透明方块，同时保持中间不透明。）

Again, use the Vanishing Point Filter, Warp and a Transparency Mask to create the next section. It will take some work, but I already illustrated the necessary techniques.

在下一个阶段中再次使用消失点滤镜、变形和透明蒙版。这要花费些功夫，但是我已在下图标记出要点。

（图释：①左部“参考消失点处理对这些部分 ②中部：用变形工具稍微添加些弯曲 ③右部：用蒙版遮盖这些区域）

We proceed to add the rest of the sections until they vanish.

继续为余下的部分添加效果，直到它们消失在远处。

Back to Illustrator. Using the Pen Tool to create a path along the sketch lines. Then round the corners. The easiest way for rounding vertices is to use a script called Round Any Corner found in this archive. Read more here.

返回Illustrator，用钢笔工具沿着线稿周围创建路径。然后，圆滑那些转角。圆滑这些转角最容易的方法是使用一种叫做“圆滑所有转角"的脚本，点击这里获取脚本（这个脚本的效果可以简单地用圆角滤镜代替，执行 效果>风格化>圆角 —— 译者注）。点击这里了解更多内容。

（图释：①上部：用钢笔工具沿着线稿周围创建路径 ②左部：圆滑路径中尖锐的转角）

Fill the final path with white, remove the stroke, and copy/paste it into Photoshop underneath the “volume” layer.

用白色填充最终路径，取消描边，并将其复制粘贴到Photoshop中音量图下面的一层中。

In order to enhance the look I decided to add another volume, this time white, underneath the one we created. Use the Vanishing Point filter once again, copy the first volume created in Illustrator, and extend it into perspective, as shown below.

为了加强视觉效果，我决定再添加一个音量图。这次使用的是白色，并将其置于第一个音量图下。再次使用消失点滤镜，复制在Illustrator中创建的第一个音量图，像下图一样将其扩展以适合透视。

As we can see, the trail is overlapping the white volume, so we need a mask to hide part of it. Leave only the top-left section visible.

就像我们看见的，白色得条带与音量图重叠了，所以我们需要用蒙版将其部分隐藏，只留下左上端可见。

We just need to add the shadow underneath our guy and we’re done with this section. We isolated it earlier in another document so just copy/paste in a new layer and align it. Set it to Color Burn with 50% Opacity. Now he looks like he’s actually there.

现在我们该在人物下添加影子了，并且影子已经做好。早些时候我们已将它从另一个文件中分离出来，这时只需将它复制粘贴到新图层中并对齐就可以了。将混合模式设为颜色加深，不透明度为50% 。现在来看看它的真实效果。

Step 5 – Background and Foreground Images

步骤 5 - 添加前景和背景图像

There are multiple ways to create the flying images, but I found that the quickest and easiest was the 3D approach. Basically, I took the planes generated by Vanishing Point and tried to recreate them in 3D. Considering that they were flying on a vertical plane on the surfer’s sides meant I had to create only one plane. I could extend everything else from there. Hang on, you’ll see what I mean.

有许多方法可以创建飞行的方块，但是，我发现最便捷的方法是使用3D软件。基本上，我都是使用消失点来制作平面，并试图将它们在创建在三维空间中。介于这些飞行的方块是存在于人物骑乘面旁的一个垂直平面中，所以我只需要创建一个平面。我还可以在这个平面里添加其他东西。耐心点，你会明白我是什么意思的。

The application of choice was Modo, but this can be done just as easily in others too. Again, I did this in 3D because it was much faster, I could move the planes into perspective much easier, which means more experimenting hence better results. Arguably I could have done this with Vanishing Point too, but it would have been time consuming.

我所选择的应用程序是Modo，但是这个效果也可以很轻松地在其他程序中创建。再次声明，我使用3D软件的原因是它更加快速，我可以更轻松地在透视关系中创建平面，这是我反复验证的结果。可以这样说，我同样能在消失点中制作这种效果，但是，这会很浪费时间。

Besides on my computer, Photoshop kept crashing if I worked more than 30 seconds with Vanishing Point, which happened about 100 times when doing work on the music volume. After that, I knew 3D was the only choice. Besides you’ll learn a new technique.

另外，在我的电脑上，只要我使用消失点超过30秒钟，Photoshop就会停止响应。这种事情在我制作音量图的时候发生了百余次。之后，我就明白了，3D软件是我唯一的选择。另外，这样做也会让你学习到新的技术。

First and foremost we need to prepare a backplate for the 3D environment with the plane we need to match. I used the following image. Save it as a separate JPG.

首先，我们必须先准备一个后平面，来为在3D软件的环境中创建效果时提供匹配参考。我使用的是下图。将其另存为一个单独的JPG文件。

（图释：①右上部：三维后平面 ②左部：我们将匹配这个平面）

In Modo create a new scene.

在Modo中创建一个新场景。

Go to Item > Create Backdrop. This creates a plane which we can add the image above as reference.

执行 项目>创建背景幕 。这会创建一个平面，来添加我们刚才保存的图像。

On the right you will see the following menu. From the Items list choose the backdrop item you just created, then in Properties, go to Image > Load Image.

在窗口右边你会看到下图的菜单。在项目面板中选择刚才创建的背景幕，然后在属性面板中执行 图像>置入图像 。

Choose the “backplate” we saved earlier.

选择我们刚才保存的”后平面“文件。

As you can see, the backdrop rotates as we rotate the view. We don’t want that, we want it to remain still and only move the geometry. In order to do that go to Projection Type > Camera. This will make the “backdrop” invisible for the current perspective view, but it will act as a background for the camera.

就像你所看到的，在我们转动视角的同时，背景幕也随之转动。我们不希望这样，我们希望背景幕保持静止，而只移动方块。为了达到这种效果，执行 显示方式>摄像机 。这会让”背景幕“在当前的透视角度中不可见，但让它成为摄影机中的背景。（这里是说，图像会以正视图的方式固定显示——译者注）

Press Command + Space. This will bring up a pop-up menu where you can choose what your viewport will display. Select Camera of course.

按COMMAND+SPACE，这会激活弹出菜单，在弹出菜单上你可以选择你的显示视角。当然，选择摄像机。

Now if we rotate the viewport the background image will remain still. To make sure the rendering will have the same dimensions as our piece, on the right side go to Shader Tree > Render and set the width/height of the scene to 3613/5000.

现在，我们如果旋转视角，背景图像仍会保持静止。为了确保渲染的效果图和我们的图像尺寸相同，执行窗口右侧的 渲染树>渲染 ，并设置高度和宽度为3613和5000 。

We are ready now to match some geometry.

现在，我们开始匹配一些几何形。

（图释：用这个工具创建一个零高度的立方体平面）

（图释：高级OpenGL菜单）

Make sure that Trackball Rotation is active in your camera Advanced OpenGL menu. This will ensure that the camera will be able to rotate on all axis not only on XZ.

确保高级OpenGL菜单中的球形旋转选项是被激活的。这能保障你的摄像机会在全部的坐标轴上旋转，而不仅仅是X轴和Z轴。

Now the tricky part. By rotating the camera slowly try to match our 3D plane with the one in the reference. Also scale the plane and make it longer.

又到了一个让人头痛的地方。我们得慢慢旋转摄像机以将我们的3D平面匹配到参考图中，另外还要将它拉长些。

One thing is obvious, the camera perspective is not dramatic enough to match the two planes. When I say dramatic I mean that the parallel lines do not converge fast enough. So we need to increase it.

很明显，摄像机的透视不够灵活，无法匹配图像中两个平面。我说的灵活是指，平行线很难很快汇集到消失点上。所以我得想办法加速汇集。

The perspective of a camera depends on how wide the entering cone of light is. The wider the cone, the greater the perspective. The narrower the cone, the lower. That’s why wide camera lenses tend to distort images and emphasize perspective. That cone angle is directly controlled by the focal length. Fortunately Modo’s camera model is very similar to the real world one, so we can edit that parameter.

摄像机的透视是根据圆锥入射线的宽度来决定的（这是对”一点透视“的原理图的一种比喻——译者注）。圆锥越宽，透视范围越大（平行线就越容易汇集——译者注）。圆锥越窄，透视范围越小。这就是为什么广角镜头的摄像机容易让图像变形，并容易加强透视的原因。圆锥角度直接受到焦距的控制。幸运的是，Modo的摄像机模型很接近真实世界中的摄像机，我们可以编辑它的参数。

I have found by trial and error that 10mm focal length is close to what we need.

我反复不断尝试，发现10mm的焦距最为理想。

Of course the objects are now smaller, because the wide angle covers more space. No worries there. We just have to close in with the camera, and rotate it. Suddenly, the plane fits much better.

因为广角涵盖了更大的范围，所以对象现在理所应当地缩小了。你不用担心这里，我们只需将它拉近摄像机，并旋转就可以了。冷不防地，平面就会变得更加合适了。

Not the perfect match but with a little bit of tweaking the camera we are there.

我们不需要完美匹配，在摄像机上有一些小差距也是允许的。

This can be a tedious process because the first time you create the plane, you have to rotate the camera and change the focal length at the same time to get a good match. Which can be tricky and time consuming. But once you have the focal length determined, everything else is a piece of cake. Below are the camera settings to get the exact match.

这个过程可能很乏味，因为你要先建立平面，还不得不去旋转摄像机，并调节焦距以得到好的匹配效果。这让人头痛，而且会非常浪费时间。但是，一旦你确定了焦距，其他事情就不值一提了。设置下面的摄像机参数，以达到更精准的匹配。

Once we have that plane into place, we can duplicate, scale and move it to create different copies. You can move them parallel to each other to give the feeling of depth.

一旦平面能够准确匹配，我们就可以开始复制、缩放和移动不同的拷贝（刚才所建的零高度立方体的拷贝——译者注）。你可以移动它们，使它们相互平行，以制造纵深感。

Observe the wide angle of the camera and the final geometry.

观察广角摄像机以及最终的几何图形。

Press F9 to render the Camera view. It shouldn’t take long because it is a simple scene. This is what you get. Switch to Alpha Output and save a JPEG.

按F8渲染摄像机视图。因为这是一个简单场景，所以渲染不会太耗时。下面就是你要得到的。转换到Alpha输出，并保存为一个JPEG文件。

Back to Photoshop. Let’s see what we got so far.

回到Photoshop中。先让我们看看我们现在的成果如何。

Open the Alpha Image we saved from Modo in Photoshop. Copy it, then switch to our surfer, in the Channels tab create a new one and paste the alpha inside.

在Photoshop中打开刚才我们在Modo中保存的Alpha图像。复制一份，然后转回到原来的工作窗口中，在通道面板中新建一个通道，将Alpha图像粘贴到里面。

Now if you Command-click the channel thumbnail you get a selection.

如果你现在按住COMMAND键点击通道缩略图，你会得到一个选区。

Create a new layer and fill it with orange (#ff8f00).

新建一个图层，并填充橘黄色（#ff8f00）

Cut and paste the images that are supposed to be behind him on a new layer and move it underneath the “digital surfer” layer in the Layers Palette. This way they will seem to be behind him in space.

将图像中假定在人物后面的部分，剪切并粘贴到一个新图层中，并在图层面板中将它置于人物图层的下方。用这种方法，在空间中会让它们看起来像是在人物身后。

There are too many of them flying right now, they overcrowd the scene so proceed to erase some of them in order to maintain balance. Most are from the lower-left corner.

现在，图上的飞行物太多了，它们让场景过渡拥挤。所以，为了保持画面平衡，要删除一些。大多数被删除的方块位于在图像的右下角。

Now we will round off the corners of the images, so they look a little less pointy. To do that load the layer selection by Command-clicking on its thumbnail in the Layers Palette. We will first round off the small rectangles, as they need a smaller radius, and then progressively do the others. Subtract parts of the selection to obtain the one below.

现在，我们要圆滑图像中的转角，因为它们看起来有些尖锐。为此，按住COMMAND键点击图层面板中的缩略图读取图层选区。我们首先要圆滑小方形，它们需要更小的圆角半径，之后我们再处理其他的。需要被减去部分的选区如下图。

Go to Select > Modify > Smooth and set the radius to 10 pixels. Then press Command + Shift + I to inverse the selection and delete. There you go. You now have rounded corners.

执行 选择>修改>平滑 ，并设定半径为10像素。然后，按COMMAND+SHIFT+I 反向选区，并删除。看下面，你现在得到了圆角。

（图释：半径10像素的圆角）

We need to apply the same technique to the larger images. This time with a radius of 20 pixels since they are closer to the camera. Do this for the next selection.

我们需要将同样的方法运用在更大的方块上。因为它们更接近摄像机，所以这次要使用的半径为20像素。这样做能得到我们想要的下一个选区。

You will get this result.

你将会得到这样的效果。

（图释：半径20像素的圆角）

And now the final step of this process is to round the corners of the foreground images. Repeat the above steps with using the radius in the image.

现在到了本阶段的最后一步，圆滑背景中图像的转角。使用下图参数重复上文步骤。

（图释：①上部：半径25像素的圆角 ②中部：半径25像素的圆角 ③下部：半径40像素的圆角）

Proceed to apply the following style to the image layers. We need to give it a little depth.

我们需要赋予这些图形一些纵深感，继续操作，在这些图形的图层上应用下列图层样式。

The difference is very subtle around the edges but its still there.

边缘上的变化十分微妙，但是效果显著。

Now we will assign different opacities to various images depending on how far they are from the camera. I need you to copy and paste the following onto a new layer and assign it 25% Opacity.

现在我们要根据与摄像机的远近，对不同的图形赋予不同的不透明度。我需要你复制粘贴下图所示的图形到一个新的图层，并分配其25%的不透明度。

（图释：复制粘贴下面的图形到一个不透明度为25%的图层）

Copy and paste into a new layer the following selection and assign it a 65% Opacity.

复制粘贴下图所选择的图形到一个新的图层，并分配其65%的不透明度。

（图释：复制粘贴下面的图形到一个不透明度为65%的图层）

Copy and paste the following selection into a new layer and assign it a 44% Opacity.

复制粘贴下图所选择的图形到一个新的图层，并分配其44%的不透明度。

（图释：复制粘贴下面的图形到一个不透明度为65%的图层）（看图像，这个参数的确应该是44%，作者出错了——译者注）

Copy and paste the following selection into a new layer and assign it a 75% Opacity.

复制粘贴下图所选择的图形到一个新的图层，并分配其75%的不透明度。

（图释：复制粘贴下面的图形到一个不透明度为65%的图层）（哎，还是参照文本吧，75%——译者注）

After all these steps you should be looking at something like this.

经过以上操作，你应得到与下面类似的图像。

The next thing to add are some highlights over the flying rectangles. Select all of them, then with a 20% Opacity, soft brush, paint some highlights on them. Take a look at the second picture below to get an idea.

下一步要添加的是飞行方块的高光。选择全部方块，然后使用不透明度20%的软画笔在上面涂抹一些高光。请参照下面第二个样图。

（图释：高光）

The only thing left to do now is add some textures on top of them. I will demonstrate the technique used for one of the images, then by repeating it a lot all the other rectangles can be filled too. We’ll use one of my photos to do this. You can find the rest of the images used images here: xn3ctz.deviantart.com/gallery/. This is my portfolio of stock photography. Feel free to use them.

剩下唯一要做的事情是，在这些方块上添加一些肌理。我会在其中一个图形上做演示，然后你可以在许多其他的方块上重复填充肌理。我们会用到我照片中的其中一张做演示，而你可以在xn3ctz.deviantart.com/gallery/中找到其他这里用到的照片。那里是我主要照片的文件夹，请随便用。

Anyway, copy and paste this image into a new layer in Photoshop

总之，复制粘贴这张照片到Photoshop的新图层中。