Bezier curves for value interpolation over time

[caryoscelus]

Currently Synfig uses complicated waypoint interpolation system, which is both confusing to edit (that's admittedly UI issue (#267), but nobody seems to be interested in fixing it for a long time) and hard to maintain (code is pretty messed up there). So while working on #272 i decided to switch to interpolation using bezier curves, which should be easier to understand and to make GUI for, as well as to implement and maintain.

From technical point of view, this reduces waypoints to holding a link to valuenode which does all the interpolating work. TimeCurve node (better name suggestions are welcome) is introduced to replace user-editable interpolation. Many current waypoints will be replaced by TimeCurve knots.

Throwing out current waypoint interpolation modes (it's necessary not for TimeCurve introduction, but for #272) will introduce incompatibility. Converter can be made, but required accuracy of it should be further evaluated, depending on whether there is actually any demand for it.

Code: https://github.com/caryoscelus/synfig-playground

---

Want to back this issue? Post a bounty on it! We accept bounties via Bountysource.

[caryoscelus]

Moved to https://github.com/caryoscelus/synfig-playground so that there's no need to switch branches to build it.

[morevnaproject]

You can clone into different directory on your workstation and have tow instances of source code. ^__^

[caryoscelus]

I know, even though it's still cumbersome (still got to have two build envs set up and rebuild every time). And i already have a few working branches even without module ones.

But regardless of that, i don't like monolithic codebase approach. Splitting whole synfig repo requires some effort though and pretty much impossible without breaking autotools, so it'll have to wait; but separating a bunch of experimental modules can be done easily, so i did that.

[Jose-Moreno]

@morevnaproject @caryoscelus Thank you for this. I work with After Effects almost exclusively for motion graphics and animation effects plus some compositing, but I've always been looking for an open-source replacement and thought Synfig would fit that niche very well.

With that said I have to ask what do you think would be required to implement this? How much time? How much money? Any additional requirements? Is it possible to provide the community with an estimate? I'd love to help out coding but I'm not remotely proficient with C++ and I have no clue on how to build a proper computer app, although if there's a coding guide for Synfig, I'd like to review it in case I can find a developer that becomes interested in the project.

For now I can help to discuss the software logic, overview animation production use cases or testing the software 🙂

[caryoscelus]

@Jose-Moreno sorry for the late reply. As far as i'm concerned, at this point synfig is not much better than dead (unless you're already using it..). That said, i already made this feature work (although it's not and unlikely to appear in official builds), though without friendly interface (as in screenshot above, you edit the curve on the canvas). Adding proper interface wouldn't be complex per se, but with the way synfig is coded it would still take some non-trivial effort which i'm not interested to make.

[Jose-Moreno]

@caryoscelus Ah that's unfortunate. I really had high hopes for this to be honest. As mentioned before been searching for an After Effects FOSS replacement for a long time and I thought Synfig was it after understanding its use cases.

The application certainly has its strengths but along this particular issue, there are just too many other things that can undermine productivity in a studio pipeline when using it, and sadly producers and artists in general will rather use what they know / are familiar with / is proven, than to spend in R&D for industry's sake.

I think my best bet right now still is to learn software development and C++ to help with these FOSS projects. When and how I suppose that remains to be seen. Thanks again for your sincere answer! 🙂

[ice0]

Hello, guys!
I assumed that from the outside it would look like this. But I did not think that it would look so bad :)
First of all. Project is not dead. And in active development.

Yes, at the moment there are not so many changes as we would like. But it should be so.
So now what is in the priority:

• Stability
• Documentation
• User interface
• Rendering speed
• Codebase
• Simplify build

We have already discussed curves with Konstantin (@morevnaproject), but agreed that we will tolerate a little more and focus on the priority tasks.

We also publish a weekly report on our page at Patreon (https://www.patreon.com/synfig), where we write about the current status of the project.

Also a lot of research work goes beyond the Synfig codebase. Ivan (@blackwarthog) is working on optimization of hardware rendering and do almost daily commits in the repository (https://github.com/blackwarthog/lab/commits/master).
I'm working on improving UI codebase and experience. Also we are fixing all errors that leads to Synfig crash.

All of this is background work, and didn't adds new shiny features to Synfig, but it must be done before we move forward.

[Jose-Moreno]

@ice0 Thank you for the information! Well I was a bit surprised about the "dead project" commentary but since it's not that unusual for these type of FOSS projects (and I'm new to the development side of Synfig) I didn't think to question it, so I thank you for taking the time to explain the current project focus as well as pointing out who is doing what during the stabilization period.

I truly agree that is critical to stabilize and bug-fix any program before getting into feature requests, particularly if those have complex implementations like this one.

With that said, I'll try my best to use and test Synfig thoroughly for different types of animation projects to see If at least is possible to contribute to it by improving the external "image" that other communities and artists might have of the project. I believe it's important to have professional content made with the software to gather enough attention from the outside, and while curve editing is important for an animator, it's not impossible to animate without it so i think we can hold on for a while 🙂