GIMP Version: 2.8.14
Operating System: Linux
GIMP Experience: Experienced User

---

Greetings, everyone.

By way of experimenting to see if I could get my stretchy head around how an animated brush functions, I set myself the task of creating a brush which could be used to stroke a circular path with the result that 4 numbers would appear at the right locations and in the right rotations around the dial.

Initial results, with four numbers, XII, III, VI, and IX, have only created mystery, and I think I have a new stretch mark and something tore.

I admit I only skimmed the documentation covering GIH (animated brushes|image hoses), but figured I'd be able to figure out most of it by making a simple one and observing it in application. Um, no. The 4 numbers are printing backwards, and I can't quite figure out which way the layers print.

Not really helping my understanding is the apparent necessity to save the brush with the rank set to "angular" which however also requires that the stroke be instructed to "emulate brush dynamics." Also necessary is that the, or at least 'an', appropriate "Dynamics" be selected on the Brush options tab. The working of a dynamics matrix is not obvious, but I searched and found information on creating a new one, so I made a new one with, only, "angle" tied to "tilt."

The third screenshot is of 4 layers; the question is:
* How to save them as an animated brush which prints numbers (in whatever the default order is) at rotations of 0, 90, 180, and 270?

The second screenshot is of the choices I made.

The first is the result of Edit->StrokePathing the red and blue paths and painting manually in green starting at 12 o'clock.

Comments are welcome. Derision is fine. My burning question is: "Is it even possible to create a brush of, say, eight compass arrows such that 'stroked' items will show the arrows pointing in 'correct' directions?"

I thank you and I am happy to be down among the magnolias again. (From the Red Skelton movie, "A Southern Yankee.")
--
gramp

Wallace to the rescue... Thank you! and thank you for posting the relevant section of the manual.

I have to ask about the Brush Dynamics: which you used? what the settings were?

I note you used "incremental" when saving the brush, whereas I had chosen "angular." Thanks for showing that. I would have tried changing anything else but that because I interpreted "incremental" to mean "use the next rank, no matter what" and I just knew the choice had to be angle-based. I will give it another try.

I'll follow your lead and post (some of) the documentation regarding "the dynamic matrix", to fill out the post for a possible future visitor ("Greetings, Visitor from the Future.").

I'll start by observing that "a dynamic matrix 'connects' mouse or tablet stylus 'actions' to individual brush 'parameters.'" (Note: at least in 2.10, matrixes cannot be edited for "standard" offerings, which must first be copied to be editable.) Note too that some of both the actions and the parameters are irrelevant to the mouse. That said, the mouse/stylus 'actions' columns are in the first image, and a sampling (just two) of the 'brush parameters' rows are in the second. Note that the Angle one I need is missing... , but too that the parameters can be influnenced by a stylus or a mouse.

Furthermore, because the documentation involves references to "speed" and "direction" of "the mouse," it's beginning to dawn on me that Edit->Stroke Might Not Use The Mouse, but I will continue this thought later.

Shows only Size and Angle. Nine (!) more, incl. opacity, hardness, jitter...

Declaration: I do not have a graphics tablet, only a mouse.

Once again, I am mystified by the results of stroking the path with an animated brush, this time with Wallace's kindly supplied clock.gih. These two results, for different path diameters (visible) and with Dynamics Off on the brush options tab, show, not unexpectedly, that the path and the brush sizes must be compatible.

However, what still surprises is the starting location for the stroke AND the fact that the numbers run counter-clockwise...


In the case where I actually enabled my custom dynamics which "only" connects the Angle parameter to the Tilt action (which isn't even supposed to work for a mouse anyway as I understand it), it gets even weirder: The brush appears upside-down(huh?) so path-stroking results in upside-down numbers. My guess is that because it's not defined for a mouse, results are, ummm, un-defined.

EDIT: I meant to say that the paths were created with the circular selection tool, set at fixed 1:1 and expand-from-center.

Flip the path and it should change the direction.

To change starting point experiment with adding a new control point, delete and add back the segment or just make a little gap in the path.

Ofnuts paths tools are the best way overall:
e.g. Reverse strokes
http://gimp-path-tools.sourceforge.net/ ... se-strokes

@teapot
Thank you for those insights regarding reversing and breaking/reconstituting a path.
And on the advice about ofnuts' paths' tools'.

(On the subject of the "direction of a path" I wonder now if a circle expanded from the center by pulling "up" has a different direction than a similar one made by pulling "down.")

@Wallace
Ahhhh! Again, I would never have taken your circle for hand-drawn.

With your two's's advices I ought to be able to get off this little limb I have crawlt onto.

Reverse-strokes is no more, function os now part of ofn-path-edit.

The path from a selection is an approximation of an approximation and there is little control on where it starts and ends. If you want a circle path, there are scripts to create them, with a known start point and path direction (ofn-path-to-shape, in my scripts, but there are others).

Perhaps mundane to this thread, my ofn-brush-strokes-on-path script. Unlike stroking the path, it allows you to accurately put equally spaced brush strokes on a path with an exact integer number of intervals.

Thanks Ofnuts. I forgot that when posting. Am using the latest ofn-path-edit though.