Having a landscape is nice and all, but what’s the point if there isn’t anything on the landscape? In this article, we will populate the landscape with black objects containing bright neon silhouettes.
For this tutorial, we’ll place some silhouettes in our composition. I will assume you’ve read the previous two tutorials before this one, and have followed the instructions. If not, take your time, and go through the following articles:
Following these tutorials is important. You will best benefit if you follow along, making your own scene in the process; as this goes into topics that aren’t generally covered by more basic tutorials (at least, not the free ones).
You will also want to download the palm tree model at this link; part of this tutorial will include the addition of previously-existing meshes into a project.
Errata: Brightness, Contrast, and Boundaries
In the previous articles, I had made use of the brightness/contrast node to limit the sparkles of the background. However, this had a negative effect on the overlaid blur effect. It took me a while to pinpoint the source, but I will share it now before we begin.
First of all, look at this wireframe close-up:
The reason this is happening is because of the Brightness/Contrast nodes I placed in the starry sky composition node collection.Do you see the two “Brightness/Contrast” nodes on the left? These were placed to dim the sparkles, reduce the number of stars that sparkled, and make the stars themselves brighter. However, the end result was that the background’s Brightness (4:1) and Contrast (12:1) was having an effect on the effects of scenes that combine with the background; this means that all the effects that used alpha (partial transparency) were brighter and more contrasting.
The end result of this is that all translucent parts of an image, such as fades and blurs, will cut off like the first example, and some colors will actually change in the process. This is what I discovered when working on the new outlines (I used orange, but the outward glow was deep red).
Keep in mind that while this might be a useful effect for certain projects, the fact that it affects everything that overlays the background means that it is pretty hard to control.
The way to fix this is to use the “Map Value” node. Once you’ve used the Bright/Contrast node to make the adjustments, and made all appropriate adjustments, place a “Map Value” node just before the final “Add” or “Alpha Over” node.
Make sure that you check both “Use Minimum” and “Use Maximum;” these set the boundaries that the values will not go below or above.
Why is this happening?
In Blender, most values that exist in a scale use 0 to mean “no amount,” 1 to mean “full amount,” and any number beyond 1 is a multiplier. So, a color value of 0 means black, a color value of 1 is full color, and a color value of 2 is “double the value.”
By default, Blender doesn’t allow value to go higher than 1 or below 0; the slider blocks any further advance in either direction. However, when combined with “Brightness/Contrast,” negative values, as well as values above 1, are possible. And while they are not immediately obvious (black is black, and full color is full color to the human eye), they can have a profound effect when mixing with other colors (which is what happens when a partially-transparent color is placed over them). So, if you have a transparent and blurred blue on top of a black background that is made “Blacker” by Brightness/Contrast, it will be darkened as if it had been affected by that node. Which can become jarring if it is being used in an outline.
Map Value’s “Use Minimum” setting is a hard limit; if the incoming brightness is darker than 0, then Map Value will change the color to 0. The same applies to “Use Maximum;” if the ending brightness is greater than 1, the node will clamp it back to 1.
TL;DR: If you use Bright/Contrast to make changes to a background, make sure you use “Map Value” to limit the color boundaries to prevent unpredictable side-effects with transparent layers.
With that out of the way, let’s get to the topic of this article: Neon Outlines!
Making a New Scene
As before, the same process applies. Create a new scene in the Top menu, change the rendering engine from “Blender Render” to “Cycles Render,” and link the camera to this new scene, which we will name “Outlined Silhouettes.” Make sure that you add the “Map Value” nodes as needed before continuing; otherwise, your glows might take an unexpected direction. Also remember to set the “Film” settings (under the “Render” options panel) to enable transparency.
Importing a Tree Mesh
Now, we will import the tree. To do this, you need to first open a second instance of Blender, and load the palm tree mesh in that instance. Then, simply highlight all parts of the tree, and use Copy/Paste (Ctrl-C/Ctrl-V) to bring it over to our new scene. Once it’s imported, you can safely close the source instance; we no longer need it.Unlike the wireframe and the sky, the trees are a discrete part of our composition, so it’s necessary to have some means of seeing where things should be.
Before we go any further, let’s cover some theory.
When an artist is at work, their first consideration is where things are placed. Horizons, framing, and positioning are essential in making an image interesting. There are several rules to consider when making your pictures.
The Center is Boring
An image where the subject is dead-center is boring. Nothing is happening, and nothing is really important. There’s no movement, no exploration… nothing. In essence, a centered picture is like a single word; it can be useful for some things, but it’s a weak form of communication.
When making a sentence, you have a subject (someone or something performing an action) and an object (someone or something that is receiving an action). A person eats food. A child plays with their friend. You are reading an article. With a sentence, a lot more can be gained than just with a single word; a meaning can be much more complete.
A picture should also communicate with the viewer. Treat it like a sentence. Have a subject, have an object, and place them in such a way that the viewer can’t help but look at them in order, and actually understand what the message is; instead of getting a single word, a picture should convey one or more sentences (hence the oft-cited quote of a thousand words).
Note, the sentence rule is not necessarily the only reason, but it helps one get a perspective about art that can have a profound effect on their output in future projects. Besides, it’s how I think about it when I do anything visual.
The Rule of Thirds
In an image, the subject and the object should share the space, but neither should get full billing. That means each should have a place of importance. This is where “compositional guidelines” comes in. You need some way for two images to be on a picture at the same time, while ensuring that each has a place of importance over everything else except their opposite number. In this picture (collected from here; watch this video for more information), you see a woman sewing on one side of the image, and a child watching from the other side. Each of them is at a vertical boundary of “thirds,” meaning that the child is in the boundary between the center and right thirds, and the woman is on the boundary between the left and center thirds.
Also, notice that they both also occupy space where those boundaries meet with their horizontal associates; the woman is centered at both the top and bottom boundary crossings, while the child is at the bottom boundary crossing. Because of this, it is instinctive to assume that the woman, her sewing table, and the child are important elements in the picture, while the door, windows, and house corner are not.
As such, you can get the sentence “The child is watching the woman sew.” Subject. Object. Sentence. Meaning. The picture works. But it only starts there. Now that we know the actors in this scene, our eyes will continue to wander. The woman is concentrating; she is making something interesting. The child is looking at the woman, but the body is facing away; they were doing something else, and took a moment to look at the woman. What were they doing? Well, they were standing in front of a door, so maybe they want to go inside. Perhaps they need the woman to open the door?
See how such placement makes a scene more interesting? Instead of just centering on a sewing woman, it instead makes a relationship between her and the child behind her, one in which something is not just being, but actually happening.
Compositional Guides in Blender
Blender’s camera has a lot of features that would put all point-and-shoot, and many DSLRs, to shame. Among those features is a compositional guide tool. In this case, we’ll be using the compositional guide to insert guidelines for rule of thirds onto the camera. These guides will not render; they are simply there for the artist’s convenience.
To set them select the camera, and go to the camera tab in the properties panel. Under “Display,” you’ll see the “composition guide” drop-down box. Simply set this to “thirds” in order to get that composition guide placed on the screen. Your camera view should now look like this:
Now, in a single view, it can be rather tricky to position the tree while making sure it isn’t dropped beneath the wireframe (which would look rather odd if any of the wires covered the base of the tree). So, Blender has a feature you would normally see in CAD software: the Quad View.
This will split the 3D window into 4 views: top, front, right, and camera views. The top, front, and right views will all be orthographic; there will be no perspective applied, so what you see is the tree’s position, without any angling or size distortion due to perspective.
Once in the quad view, move the tree around in the three orthogonal views. That way, it can be resized and positioned in such a way that the base of the tree will never fall below the “ground” of the composition (at the Y level of 0). All you need to do is move it around and scale it, all while watching what happens in the camera view.
The end result should look something like this:
However, maybe this isn’t enough; many 80s art with palm trees will see them in a cluster. So, why don’t we duplicate this tree, and place the copies next to it, so we have a cluster?
So, now we have a cluster of three trees, congregated on the left side of the image. So, what are we going to put on the right? Well, that’s the subject for another article. Let’s just focus on these trees for now.
Now that we have the trees where we want them, how do we make them into a silhouette? By making them black. In this way, they will block whatever’s behind them, but they will not be seen directly. That is for the Freestyle tool to do later.
For now, let’s go to the compositor. Make sure that the scene is in the “Outlined Silhouettes” scene; the “Composite Scene” comes later. Once you’re in the compositor, change the node tree view to the material one, select one of the tree objects (fronds or trunk, doesn’t matter) and delete the current material, using the “X” button next to the material name.
That was easy! Now, just select each of the other objects, and change them to the “Black” material (use the material dropdown next to the name). Now, all the trees should be nothing but silhouettes, not actually visible, but blocking out everything behind them. The color and roughness should prevent any light from reflecting off of them.
If you render this scene, this should be what you see:
Outlining the Trees
Now that our trees are nice and black, and set against a transparent background, we want to outline them in a bright color. For this scene, let’s make them yellow (since the beach is green). I know it’s goofy, but hey, so were the 80s!
Freestyle is a tool built into Blender that allows it to draw lines, either as a border around objects, or as a boundary for edges and corners. People often think of it as a tool you use to make rendered scenes look more hand-drawn (or photoshop-designed). In this case, however, this will be used to draw the boundary in glowing lines, so the trees will have the same colored glowing outline as the ground’s wireframe.
Before it can be used, it must be enabled. If you enable it, it will only activate for this scene, so the starry sky and wireframe scenes will not be affected. You can find a checkbox in the “Freestyle” section of the “Render” options panel.
Now, there are other freestyle settings we need to change, but they’re not in the “Render” properties, but the “Render Layers” properties, so switch over there. You should see an empty list box in the “Freestyle Line Set” properties. Click the “+” and name the set “Outline.”
If you try to render the view now, however, you won’t see anything, because, by default, Freestyle uses black lines (to simulate pen drawings). To change that, We need to go further down, under “Freestyle Line Style” settings, click the “Color” button, and choose our base color (Yellow, in this case). Rename the line style, to keep track.
Because, by default, edges get rendered by Freestyle. To fix this (in case this isn’t the effect you want), we need to go back up to the “Freestyle Line Set” properties, and look a little lower, under “Edge Types”.
Despite the names, we don’t want “Border,” or “Silhouette,” and we obviously don’t want “Crease,” which would render all strong creases in our view. Instead, we want “External Contour” to be selected; External contour is essentially the outer edge of a shape.
At this point, our trees are rendered, and ready for compositing. They use the same compositing process as the wireframe (so they can glow), so we just repeat the process we performed for the wireframe. For the sake of demonstration, I’ve also added a black background, so you can see both inner and outer glows; the outer glow is hard to see on the checkerboard background.The result should look like this:
Now that we have our trees, we do, of course, want to add them to our existing composition project. To do this, let’s now switch to the “Composite Scene” scene, and add our new scene into the mix.
Adding the New Scene
There’s not a lot new here. Add a new “RenderLayers” node, and set it to the new “Outlined Silhouettes” scene. Then, repeat the same process used by the wireframe scene. Copying and pasting should make it pretty trivial.
Now, comes the new trick: adding it to the existing project. The “Alpha Over” node is obviously only able to take two images, so how do we add a third? By linking another Alpha Over node into that one, and connecting the second and third lines to that.
The silhouette is completely different in orientation, but it remains three trees surrounded by yellow. Everything I’m showing you in these articles can be used in animations; they are not limited to still images. So, if you want to create an animation, these will give you the tools you need to get as creative as you wish.
At this point, the main composition’s node view is getting pretty cluttered; there’s stuff all over the place!To fix this, one thing we can do is group up all the compositing for each of the scenes into their own single boxes for the main view.
First, select all the processing nodes for a specific scene:
Here’s what it looks like from outside now (press “Tab” to enter and exit group view):
Note that we did not include the “Render Layers,” “Alpha Over”, or “Composite” nodes to the group. Blender won’t allow “Render Layers” or “Composite” nodes to be included in groups, since they are main sources and the final destination. As for the “alpha over” nodes, we want to keep them separate, so we can keep adding scenes as we make them, since we will need to keep adding “Inputs” as we go.
So, now that the Starry Sky processing group is made, let’s give it a name. We’ll call it “Sparkly”… because it’s so sparkly! Then, we’ll repeat the process with the wireframe (Landscape Glow) and Tree (Tree Glow) processing trees.
One thing to note, now that the individual processing groups have been packed into their own nodes, the process of adding new scenes should be pretty self-explanatory, since the alpha over nodes now show a clear pattern. Once we have a fourth image, we can simply create a new “alpha over” node, and insert it into the second slot of the second “alpha over” node, making sure both the trees and the new scene are connected to the new node. And it can be repeated for as many scenes as we will be adding (and there will be quite a few more).
At this point, we can now add black, outlined objects to the scene, which are true 3-D objects (not black cutouts), but still silhouettes with glowing outlines. This can be repeated for trees, objects, or even (and especially) text. Go nuts. Have fun! And make something good!
And, of course, HAPPY NEW YEAR!