2009/12/27 Krzysztof Kosiński
<tweenk.pl@gmail.com>
Let's take another approach. Instead of trying to assign meaningful
names to existing controls, let's try to figure out what the intuitive
controls could be (without thinking about any of their implementation
details for now) and then try to adapt the existing controls to this.
I have not seen the spray tool yet. Here are the controls I would
expect to find.
- Size: controls how big is the spray, like brush size. This tool
could actually use vector brushes to determine the shape of the spray
area, for example a Pick Brush button that would allow one to pick a
shape from the drawing, but this is probably not how it was
implemented.
- Spray rate or Flow or Thickness: how fast the objects appear or how
many of them there is. This should be relative to the area of the
original object, so spraying two different objects with different
areas should fill the same fraction of the spray area.
- Focus: how concentrated in the middle of the spray's area the
objects are. High means most objects appear near the center, zero
means that objects appear uniformly over the entire area, negative
means that objects are more likely to appear near the edges than the
center.
Those options would be useful in single path mode:
- Rotation variation: 0% means every object is sprayed exactly as it
is, 100% means that every rotation is equally likely. At 50% the
distribution of rotations looks like a sin^2(x): the probability is
highest at 0°, then drops to reach zero at 180°, the rises again to
reach the peak again at 360°. At 100% it is uniform. It affects the
distribution of rotations very much like Focus affects the
distribution of objects from the center.
- Maximum scale: 1.0 means no no scaling, 2.0 means objects can be up
to twice as big or small as the original, 4.0 means four times bigger
or four times smaller, etc.
- Scale variation: like rotation variation.
Color options are unnecessary, because a variation in coloring can be
easily obtained using the Tweak tool.
Graphical editing of the distribution might be unnecessary, because
the user is going to drag the spray over the canvas and I doubt
complex radial distributions would be much different from a simple
Gaussian in that situation.
Regards, Krzysztof