Now that we have a raw render, it is time to learn how to improve it using the compositing tools of Blender.
Blender is a complete tool that also allows compositing. This is the ability to edit an image or a sequence after the rendering phase. You probably have already tried compositing, maybe unknowingly. For example, Adobe Photoshop© is a software that allows us to composite a single image. Unlike Adobe Photoshop©, Blender uses a nodal system that provides a great flexibility. We can make changes at any point without the loss of information. Let's try this:
We have now two nodes, a Render Layer node and a Composite node.
Now the render image appears behind the nodes. It will be pretty useful to do compositing in full screen. If you want to move the render image, use Alt and MMB. Two other interesting short keys are V to zoom in and Alt + V to zoom out.
The possibilities of compositing in Blender are enormous. For instance, you can easily use keying techniques that are often needed in the movie industry. It consists of replacing a green or a blue screen behind an actor by a virtual set. Compositing is an art and it take too long to explain everything, but we will see some of its basic concepts so that we can improve the shots of our sequence.
Now, we are going to work on the first shot of the sequence:
The following is the combined output socket of the Image node, and there is a Depth output socket. This pass will allow us to simulate an atmospheric depth. It is an effect that can be observed when we look at a distant landscape and a kind of haze is formed. The Depth pass is a visual representation of the Z-Buffer on a grayscale. The objects near the camera will have a gray value close to black, unlike the distant elements, which will have a value close to white. This pass could serve for other things such as masking or blurring the focal depth. It all depends on the context. A controlled atmospheric effect may bring realism to the image.
Now that we have added the nodes, we are going to connect them as follows:
A render before and after the ZDepth pass
Color correction of the shot
One of the most important aspects of compositing is color calibration. Fortunately, there are easy-to-use tools in Blender to do that.
There are two correction formulas: Lift/Gamma/Gain and Offset/Power/Slope. These are the two ways to get the same result. For each one, there are three color wheels and a value controller (Fac). You can modify the darker, the mid-tone, and the highlight values separately.
If you want more information about the color balance node in Blender, visit this link :
https://www.blender.org/manual/composite_nodes/types/color/color_balance.html
A render image with an adjustment of the Color Balance node will look like this:
Before finishing the compositing, let's add a few effects.
We will add a Filter node (press Shift + A and select Filter | Filter). We must connect the Image output socket of the Mix node to the Image input socket of the Soften node, and we will connect the image output socket of the Soften node to the Image input socket of the color balance node. We will keep the filter type to Soften with Fac of 0.500. This blends the pixels so that the image is less sharp. A photo is never perfectly sharp.
We will add then a Lens distortion node (press Shift + A and select Distort | Lens Distortion). We must connect the Image output socket of the Color Balance to the Image input socket of the Lens Distortion node and the Image output socket of the Lens Distortion node to the Image input socket of the Viewer node. We will check the Projector button. The distort option is at 0.000, and the dispersion option at 0.100. This node usually allows us to make a distortion effect such as a fish eye, but in our case, it will allow us to make a chromatic aberration. This adds a soft and nice effect.
The nodes of compositing
We have now completed the appearance of the shot.
A render with final compositing
We are now ready to make the render of our compositing: