Following the previous blog entry, I want to share some more detailed tricks to create vector objects, that have a variety of blur levels along their edges. This technique is useful when drawing realistic images in Affinity Designer, because hard edged objects rarely look realistic. This technique involves clipping of objects - placing one object inside an other - and gaussian blurring, which is applying a controlled blur ratio to objects. In addition gradient colour fills and gradient opacity can be applied to the objects to attain an appearance that makes objects look realistic. Apart from being useful in drawing realistic vector portraits, this technique has many other applications for artists who need to give their artwork a convincing realistic appearance.
First some basics. In the image below this paragraph I placed an image in which the basic principle is shown. After that I will go into more complex techniques that are nonetheless based on the same principle. To give an object a blurred edge, click the fx button at the bottom of the Layers panel. In the dialog that pops up after clicking it, at the top of the dialog, the blur ratio can be set in rounded numbers and can even be set as precisely in decimal values. That is rather basic, but in reality most objects have edges that have a level of blur in their circumference that varies in a non-linear way. This can not be achieved by just using the fx Gaussian blur function.
 |
| Basic clipping and blurring principle |
Circle 1 has a blur ratio of 10 (roughly medium blurry) and is clipped inside a grey curve that has a blur ratio of 5 (less blurry). Where the objects intersect, at the top of circle 1, the intersecting area inherits the blur ratio of the grey curve - blur ratio 5, while the rest of the circumference of circle 1 maintains its blur ratio of 10. So at the top of circle 1, the edge is less blurry than in the rest of the circle.
Circle 2 that has a blur ratio of 24, which is more blurry than that of circle 1. Where circle 2 and the grey curve intersect, circle 2 will inherit the blur ratio of the grey curve, so the blurriness at the bottom of circle 2 is far less than in the rest of the circumference of circle 2; there is a greater contrast of blurriness along the edge of circle 2 than there is in circle 1.
By moving the right edge of the grey curve, by selecting the nodes at the top and bottom of its edge with the Node tool, closer to or further away from circle 2, the level of blurredness at the right side of circle 2 can be set quite precisely. Basically, the further away the edge of the parent curve is from that of the clipped child curve, the less its blurriness affects the edge of the child curve. When the edge of the parent curve overlaps the edge of the clipped child curve, the latter inherits the level of blur of the parent curve in the area where the overlapping is defined.
Below this paragraph I placed an image with a transparent curve with a blue outline (for clarity) that has no blurred edge, in which a transparent curve with a green outline is clipped that has a blur ratio of 6, in which a red circle is clipped that has a blur ratio of 50.
 |
| Objects clipped inside other objects, with outlines drawn for clarity |
The crux is that the parent object in a clip always determines the blurriness of edges. Parent objects can both be given a higher or lower blur ratio than the objects clipped inside of them. If the clipping parent objects have no outline - which is the practical way to work with such a technique - the image shown above would look like the image below this paragraph, indicated by the yellow arrow. The resulting appearance of the clipped red circle becomes clear and this is how clipping and blurring would be used by graphic artists and designers that aim to create realistic objects or parts of objects.
 |
| Objects clipped in other objects, without the outlines, which is how the clipping and blurring would normally be used |
In the reality that most people experience, areas that have different ratios of blurriness along their edges, are the majority of what occurs / appears in this dimension. Also objects rarely have one equal colour throughout their shape, which is what in addition to the above, can be applied to the mix by giving objects a gradient colour fill or even clip differently coloured child objects clipped into parent objects to mimic reality even closer. Although what I described in this blog entry is not always an obvious intuitive mode of operation, Affinity Designer is very capable of coming a long way in 2D design. Programs of Designer's competition may have Mesh Fill options, but creating those requires much more time and editing them afterwards even more.
