Mesh Fill Tool is the most powerful and multi-purpose tool for modeling of complex color transitions. Besides that, it can contain a variety of different colors, and gradients between them will have the most arbitrary shapes, any of the Blending modes can be applied to them. All this makes the objects of this group extremely useful and sometimes almost impossible to replace in creating complex realistic illustrations. It is not surprising that sooner or later the hands of anyone who tries to work seriously with CorelDRAW, start to drag themselves to the Mesh Fill Tool button.
Of course, I also could not ignore this topic. At first, frankly speaking, the Mesh Fill Tool seemed quite moody and inconvenient. But if you follow certain rules, the result can exceed the most ambitious expectations. Basically, pretty soon I familiarized myself with the topic and has now decided to share some of my observations. Maybe even for some beginners (and not only beginners) it will come in handy.
As an example, I took one of my artworks (at the time of creation it was only the third one of its kind, created using this technique)
Just letting you know that I did not set myself a goal to describe the work of the tool in all details, especially since there is enough information on this topic over the internet. I will just focus on some peculiarities that I myself have been very interested in the process of learning how to work with the Mesh Fill Tool.
Ways to create gradient meshes. Part 1
Gradient meshes can be obtained in two ways:
a) convert the mesh into an existing object;
b) create a gradient mesh of the desired shape, based on a simple geometric figure, which was converted into a mesh with the correct number of cells. Since the cells always have rectangular shape, the easiest way is to use a rectangle.
First, let us take a look at the first option. To do this, try to convert the right hand of the Flight attendant into a gradient mesh. Take the Mesh Fill Tool (M). Set on the Property bar the amount of rows and columns equal 4 and using the cursor click on the contour of the hand.
The result of automatic conversion does not inspire much optimism ...
It will be clearly inconvenient to work with an obtained design. The fact is that the mesh lines with automatic calculations on one hand follow the configurations of the adjacent path segments more or less, but on the other hand then always tend to reach out for the horizontal and the vertical. In this case, we can only rotate the original object in a way to bring the features of the program to our needs as best one can.
By the way, if you rotate the asymmetric objects with the pointer, the center of rotation of the object will be shifted along with the center of the coordinates of the object. This leads to the fact that with the subsequent rotation to the starting position the object will not return to its original place. To avoid this, set the center of the transformation before turning the object. To do this, either double-click to select an object and move the center of rotation to a new position (which is then stored for subsequent turns), or use the Transformation docker (Alt + F8). In this case, select the object, turn off the Relative Center option, then set the angle of rotation and click the Apply button. As a result, the center of rotation of the object is also fixed at one position.
Reapply the Mesh Fill Tool using the same settings.
Return the object to its original position. To do this, set the angle of rotation, but this time with an opposite sign in the Transformation docker.
At the next stage, it makes sense to simplify the mesh by removing unnecessary points. These usually include additional nodes, which in contrast to the primary nodes do not form lines crossing, and usually are not very important - they cannot be filled with the color and get in the way while editing mesh geometry. As an exception, additional nodes located on the external lines of the mesh. They define the shape of the object and therefore their removal is not always advisable. The nodes located on the outer lines of angular cells form their integral part (since each cell must contain 4 corners) and therefore cannot be deleted. They can also be colored.
While modeling the mesh new nodes can be added. Double clicking on the line or inside the cell immediately creates a new intersection. After a single click, insertion marker appears. Clicking right button of the mouse brings up context menu, you can select the type of the created node – primary or secondary. Newly created lines lie parallel to the existing ones.
All the intersections created by the program are corner. For convenience, usually you can immediately convert them into smooth. This is done in the usual way for Bezier curves.
After the initial setup of the object can be returned to its original position. To fix the properties of the created mesh, simply do something with it, for example move or remove the anchor node. If you rotate the object right after applying the mesh, then all of its properties will lost. By adding or removing anchor nodes and moving them to the desired position, as a result we obtain the required shape of the mesh.
You can now start its coloring.
Working with color
While familiarizing with the methods of applying color on the gradient meshes, I did not discover any peculiarities. The colors can be applied to single nodes as well as to the entire cells. Color samples can be taken either with the help of an eyedropper (in the Sample Color mode) from the original image (if there is one) or from the preset custom palettes, or getting it by mixing the components in the Color docker. Color assignment to the selected items is done in a common to all the other objects method. The procedure depends on the complexity of the mesh. If the amount of cells in it is relatively small, you can first build the mesh completely and then proceed to its coloring. If there are a lot of color transitions, it is much easier to outline the general shape of the mesh first and then assign the colors to its key nodes, and then paint the other nodes and cells in the process of adding them. The general rule is, the simpler grid is, the better. It is not necessary to create additional nodes where the setting of lines curvature using the controls of already existing nodes will do. During the editing process make sure that lines do not overlap and there are no sharp kinks.
In this way you can model simple objects.
...as well as the complex ones. (Lazy people will have to go through a lot).
By the way, the Mesh Fill Tool as a tool to represent fabric folds is simply beyond competition in convenience and flexibility.
A curious fact: all gradient meshes at the time of creation have CMYK color mode assigned to it, even if the original object had a fill in RGB color mode before it. An attempt to assign to individual elements of such mesh color in the RGB mode, causes in automatic substitution of the nearest equivalent in CMYK, as it can be easily seen after taking a sample of the color with an eyedropper and looking at the Color docker. It is noticeable with the naked eye. If, however, the artwork is done in RGB, then the only (known to me) way to bring the mesh to the desired color model is to apply the fill of one RGB color to all cells at once. To do this, select all the nodes with the Mesh Fill Tool and fill them with the necessary color in any of the above mentioned methods.
Ways to create gradient meshes. Part 2
Now let’s take a look at more complicated example of building the mesh. To do this, let’s try to convert to mesh a fairly complex object such as lips of the Flight Attendant. Mesh obtained through automatic conversion, looks inconvenient for further editing. Rotating an object in this case does not give the desired effect.
So let’s go in the other way, more familiar to those who knows about the three-dimensional modeling. If editing a failed geometry of object threatens to take away a lot of time and effort, then it will be easier for us to create a gradient mesh with the required parameters from the beginning. Take a regular rectangle. In the Property bar of Mesh Fill Tool set the necessary for us amount of rows and columns. As already mentioned, all the created with the program control points are corner by default – this is acceptable at this time, and in addition convert into lines all line segments (they are curved by default)
We can now proceed to placing the nodes.
After placement and addition of new nodes, mesh takes quite a harmonious and logical view. Direction of the lines corresponds with the main axis of the object.
Here again we may need the lip contour created earlier. First, it can be used as a guide for more accurate placement of anchor nodes. Second, it can be used as a container for placing inside the created mesh (Effects > Power Clip > Place Inside Container). This method has its advantages as well as disadvantages. On the one hand there is no need in special attention for placing mesh nodes along the object perimeter because edges of the container mask all its roughness, and if you need to change the outer shape of the object, it is enough to fix one or two nodes of the external outline without affecting the geometry of the cells. On the other hand, it is not very convenient to color the gradient mesh inside the container, because at the time of editing it is hard to predict how its colors blend in with the surrounding objects. In this case I decided not to use this method, preferring greater visibility.
Now, for smoothing the shape all the lines can be converted back into curves, and transform the nodes into smooth ones.
As a result of automatic smoothing we get just such a structure.
Adding new nodes to it and correcting the existing ones, obtain the final shape of the lower lip and prepare it for applying the color. Make some nodes pointed, in order to achieve the effect of folds.
Do the same thing with the upper lip. Then take a lipstick ...
The same method was used to do the legs of the flight attendant…
Combining gradient meshes
For all its versatility gradient meshes have a number of limitations. For example the maximum size of the mesh is 50x50 cells. Key anchor nodes, as opposed to vertices in the three-dimensional modeling programs, can connect a fixed amount of lines – either 3 (for nodes lying on the external edges of the mesh) or 4 (for the rest). Nodes cannot be arbitrarily split, merged, etc. However, when creating realistic illustrations sometimes it is necessary to design complex surfaces with variable level of detail. Such as a human face, where there are relatively flat areas (cheeks and forehead) located next to relief (eye sockets, eyelids, nose, ears). Gradient meshes are suitable for such tasks in the best way possible. Just not everywhere their use is equally convenient. Technically, nothing prevents from split any area on any amount of cells. But the problem lies in the fact that there are a lot of added new lines, which because of overall linearity of its structure are spread across all the entire area (including, even where they are the least needed). This eventually leads to excessive complexity of the grid and makes it bad-managed. The goal to portray the face by using a single gradient mesh itself looks attractive, but does not always work. In this work I have tried another method, which can be called "modeling with fragments." In contrast, for example, to 3ds max here cannot be combined into one unit, but their direct task of creating mesh areas with different structures, is done very well. While working on a portrait, I decided to divide the head of the flight attendant into 6 pieces, each of them should contain only the necessary amount of cells if possible. I would not say that everything was exactly as it was expected to be, but the overall results of the experiment were found to be satisfactory.
"Fragments" were located to partially or almost completely overlap each other. The key part, of course, was the masking of seams of separate fragments. In general, the task was not very difficult – we just needed to choose and distribute the right colors on the edges of the meshes (which, however, in some areas have led to appearance of unplanned nodes). Eyes, eyebrows and mouth, of course, were made as separate objects and are placed on a layer lying above. Flight attendant’s suit, by the way, was also created on this principle. But there the problem of putting together individual fragments was facilitated by the presence of edging and other natural seams, masking seam lines.
Note from the editor: this tutorial was written using the CorelDraw X3. Today's releases of the program have the ability to assign transparency to individual nodes and cells of the gradient mesh. This feature allows you to mask the seams of the gradient meshes easier.
Vlad Sokolovich is a freelance designer from Russia. You can contact him here