Not sure if this has been mentioned before. If not, I’d like to see what you guys think.
I’ve recently been trying to see how Typography, Kinetic Typography and Motion Graphics could work in Synfig (because apparently, my artistic skills are SO BEHIND THE TIMES and Ireallyshouldbrushuponmotiondesignohshit!).
Theoretically, it’s all possible with Synfig, because it has every algorithm we need under-the-hood. Practically… weeeell… Let’s just say I couldn’t go any further, because doing something like those videos (I mean, “precisely” like those videos, with every little detail) was too frustrating. And this is obviously because, so far, Synfig has only focused on being studio-animation driven, rather than motion-design and it has some significant priority checklist to complete.
Even so, I thought I’d bring to your attention that tools for typography and motion-design may also greatly improve the quality of studio-animation works, and that in some future versions we may need focus on that bit of a niche. Simply put, the only good applications that help complete such projects currently are all proprietary tools, such as After Effects. If we could create toolsets in Synfig to make it easier for artists to create motion-design and typography, we wouldn’t need to rely on proprietary tools at all!
I know that Synfig doesn’t lack in ways to animate objects. XD (Which is why I menat that Synfig is, theoretically, pretty capable because it already has all the needed features under the hood.) But it certainly lacks in efficient and quick means to create beautifully designed typography (and Kinetic Typography) swiftly enough. By which I mean, if something can be done in AE in just few minutes, it usually takes an hour or more in Synfig – even though it’s very much possible – such as typed words “morphing” into different fonts, then each letter falling / breaking apart, changing colors, disappearing, etc. all the while still being properly kerned, etc.
Which is why I wondered if Synfig could eventually focus on typography toolsets to make it easier for designers and illustrators to make beautiful visuals (as seen in the links provided in the first post)
Unfortunately for me (and people like me), programming doesn’t make much sense to me. I can master everything, from physics to poetry, to animation… but if there’s one thing that’s been eluding me, it’s programming…
There’s still some limitations though. First is the lack of a valuenode to extract individual letters from a string. Second is the lack of a way to query information about how each letter should be placed. Above I simply assumed uniform offsets between each character, but that’s not how real text works. There could be a valuenode which could extract the relative positioning of each letter so that it somehow could be used together with duplicate layer to properly layout text.
Morphing one font type into another is a tricky problem. First there would need to be a way to extract letters as a Spline (note that a single synfig spline cannot reproduce the shape of letters on some cases). Then there needs to be a way to interpolate from one spline to another spline, which is not easy to do unless both splines are “compatible”. And making splines “compatible” takes a lot of manual work using the usual methods.
It’s amzing what you do there, with the transformation of the type of values and the linking of them. I recently used the type transformation, and I must recognize it’s a powerfull tool. But sometimes you get lost, with the manipulation of all this values and linking. It seems that it’s unavoidable. I mean: when using Blender, or Inkscape’s filters, or Excel Solver or dynamictables, or AutoCad layers …etc, sometimes there is a moment when the complexity of the tasks, and the quantity of variables is too much.
Thanks for the videos. I think you could add a voice narration. In my case, I just don’t follow the videos images, as follows the voice.
And yes, I see what you mean. I’m kinda used to this sort of things because I’m a programmer. But other people might have a harder time.
I try to always minimize the complexity. Using keyframes for animation whenever possible. But with duplicate layer there’s just no way around it, because it’s mode of operation is thru linking and valuenode transformations.
I was hoping that by keeping the “recipes” short and to the point would be enough to help others deal with the complexity.
Yeah, I was afraid of that. I can type in English just fine, but my spoken english isn’t as good.
On the same topic there’s now a Synfig Import Labels plugin that allows to import text lines and their respective timings. The plugin can generate keyframes for you, and optionally a few animations of the provided texts or words (appearance, descend, flight, progressive reveal). New effects can be added (rather) easily.
Any contribution is welcome, such as feedback from tests, new effects templates, a GUI in Scribus… other ideas, check out the open issues!