Every time I make a colored region with bezier outline, Synfig makes two objects (layers), linked together, sharing the vertex list. One layer is the color fill, with very few properties.
the other layer is the outline, with many great features, like width points and dashed parameters.
Sometimes I make so many filled regions, that it would be nice to have only half the number of layers.
Is it possible to make a new kind of layer: starting with advanced outline, and just adding a color parameter for fill color?
And making both of them inside a Group Layer?
Well, it creates an additionnal layer, but when closed it looks like one ^^
The philosophy of Synfig is to have the outlline and the fill region separated, so it is possible to apply Blending Modes, Opacity and other effects to both of separately.
Even creating an “hybrid” would require a double management internally (not the same algo) and the properties of one of the layers will not be reachable.
I agree that having separate layers allows independent blending … and potentially independent z-order.
but I wasn’t talking about replacing advanced outline. I was talking about a new kind of layer with an additional color for fill, not nesting a “region layer” inside.
That’s why I told double management: at some point the outline overlaps the region, as they share the same vertices but it can have a different width.
So there are 2 steps implied (the region under, the outline above, overlapping the border of the region)
And if you apply opacity or blending mode, if you have only one set of parameters, it would be both the outline and the region that would be affected.
So the next step would be to expose both parameters (opacity outilne, opacity region), with another parameter to have them linked or not, and so on …
That’s also the reason why Synfig uses Group Layer so much, it is the most flexible solution to keep the things simple.
Even when you import an image, it is inside a Group Layer, by design for have the possibility to apply transformations independently from the content.
I had this issue as well back then. And at that time I did find some
workarounds (albeit somewhat less convenient), like exporting some spline lists and link them,
or meticulously link every single point one by one; vertex, tangents, and all of that
In the end, I eventually learn to just… deal with it.
I decided to not keep using those workarounds for stability’s sake.
With the export one I’m concerned that any changes made to synfig
might gonna break the shapes (or worse, the file itself),
and with the second one… no, can’t be bothered doing that.
Now in that post, I did discover another somewhat working alternative.
But considering the caveat that it brings,
(losing the ability to change tangent split/merge, losing ability to animate those shapes, etc.).
I don’t think it’ll worth the time that I’d take to just link those points one by one…
Though, one thing that I always keep in mind is to label my set of outlines
so I know which belongs to which.
I’ve used blank group layers to this before,
but as of now I’ve switched to “muted” plant layers
(muted, as in their properties are shrunk down to zero as close as possible,
including their opacities, And, prevent them from rendering.
It’s an overkill, but I need to make sure that they won’t lag synfig too much)
And the reason I don’t use groups to do this, is because they limit the scope of the blending.
(it has something to do with my composition workflow. And they’re quite handful)
(like… it’s a lot)
You can’t see the region layers here, since I put them in a separate group layer.
So those alpha layers (masks) have to take their places instead.
but Yeah, I do agree, the fact that I can tailor every single element of the spline objects;
lines, regions, and every single of their properties, to my own liking,
is really helpful and plain awesome!
… and potentially independent z-order.
