2D animation - Pencil, Inkscape and Synfig

Hi folks,

Sorry about the late reply, but I had some problems here in these months. I am building the project again, but I have some comments and suggestions for Synfig ( can I do it? ) .

First, The animated scenes ( I am editing the scenes in Blender ):

Well, as you can see, I’ve changed some styles…the boy is different and without outlines; I’ve added some brush strokes in inkscape using tablet with pressure ( nice result, I think! ).

Here, the boy prepared for animation in Synfig:

I had some problems with the boy because I created it in Inkscape and converted to synfig using the SVG2Sif; but the file is too huge and when I imported it inside of the scene file, the boy was huge, and inverted horizontally and vertically. My solutions was: I used the rotate layer and the stretch layer ( with negative value in the horizontal axis ) to use the boy in the scene. Well, it’s done, but with some work. Question 1: Is there another way to configure the file created with the svg2sif to fit the synfig world values? Did I make myself clear with this question?

Problem 2: the rotation layer show a very small blue dot with the file created with svg2sif; I mean, the blue dot is too close to the green one, so it’s difficult to grab it for rotation. Question : Is there a way to increase this distance between the blue dot and the green one?

Another question: Is there a way to see only the outlines to improve performance? The performance issue is the negative point in synfig.
Sorry about all of these questions, but I am trying to collaborate with synfig too.

I have 3 more scenes not animated yet, and I am working hard on then this weekend; next week I’ll be back…it’s a promise!
Please, answer me. I am trying to do a good animation with synfig.

P.S.: I am using the 1920x1080 resolution ( HDV full ).

Nice style man. I love the hair!

svg2sif is an external application. I don’t know if it is optimized for small amount of points and layers. Maybe you can contact with the creator.

This is because the rotation duck has a fixed distance (in units) form the rotation center. It would be good that the rotation has a canvas independent dimension. When you import the file form SVG it sets the units so huge not being needed that really.

Disable the layers in the layers panel. One by one, yes a pain. Group feature needs to do that but it is broken. Ikes!

no problem you’re welcome!

Hi Genete!

Thanks for the answers! Very helpful. I am trying to solve all of these problems, but they are not a brake to the project. It’s a simple project, I have 3 more scenarios to compose with characters, a just 2 complex animations to build. These complex will not be built with the cut-out animation. They will be frame by frame. The simple ones are cut-out.

I have a suggestion for the controls: I think ( just an idea! ), that when we create a encapsulated object, all the controls ( rotate, scale and move ) should appear as default; and the controls ( move, rotate ) should become more easy to use. The toonboon controls or the creatoon controls are very easy to use. I do not like the anime studio controls or the “bone” solution very much; the cut-out with the rotation control is good, I just don’t like the “blue dot” and the “green dot” linked as the control.

I have read in the wiki page a complaint about the “zoon” in the timeline, the person ask for it’s removal…please keep it there! The “zoon” feature in the timeline is a great option, because we ( the animators ) can “zoon in” to see frame by frame and manipulate the timing with precision.

I really think that the playback performance is the great problem of Synfig, the other things are less problems.

This project is a good opportunity for me to study Synfig. At the end of the project ( soon! ), I’ll send all the files to the community.

Regards…and thanks again!

One guy was doing a branch to incorporate a opengl render to synfigstudio and synfig. Unfortunately he is busy with real work now so it is not being developed at the moment. I hope he can come back soon.

here it is what he went so far:
synfig.git.sourceforge.net/git/g … mae_opengl

That’s good! Let’s see the final result. OpenGL is a good solution, I see it in Blender and it’s great! Blender is a 3D software using bones and mesh morphing, but the viewport shows a good framerate to test animation.

By now I am trying to use the scenarios as bitmaps, the characters and morphing objects as vector ( as simplest as I can ); I tested the “uncheck” layers solution too…it’s the best way to improve performance; you are right.

Now a question Genete: have you tested creatoon and toonboom as a reference? I know you used anime studio in the past and it’s a good reference( as flash is ).


Uh! Wouldn’t you suggest to take any of their GUI features to be included in Synfig? I’ve seem some toonboom videos and they have a lot of interesting GUI features but I want to be realistic, we have none working on code by now, so I don’t tend to be so ambitious in terms of feature requests.

Anyway, are you thinking on any of them in particular?

Well, I intend to keep going with Synfig ( nowadays I only use free softs or open source ones ); but in my past I have used all of then; and of course they have some positive features. I am not talking about the Gui ( but of course it’s a point for study ), but the way things are used, for exemple: the rotation, scale and move control.

About the Gui, what I see is that the proprietary softs have a more “user friendly” gui; they are more intuitive ( it’s a good answer for this; they have “probably” a design sector for tests ), and I know that in the open source world we have more programmers than designers.

Do you Know Ktoon? They started with a good GUI ( a toomboom reference ), but the project is freeze now.

I don’t know how can I contribute with this “design”, but if you want I can try to create a better user interface and send to you ( I can build this in Inkscape, so the objects should be vectors ); I mean…as a suggestion. I did a workshop here in Brazil to show the use of open source software to build animations ( the focus was 3D animation, so I talked a lot about Blender ); but when asked about 2D animation, I talked about Pencil, Synfig and Creatoon ( creatoon is not open source, but a nice freeware program ).

Now, with this project, I am learning a lot about Synfig and at the end I’ll write a “Making of tutorial about all the tools and features I used”; yes Synfig is powerful! We just need some adjustments to show it to the animators and prove to then that it really works fine.

When I showed Synfig to a animation studio, they complained a lot about the things that you already know:

  • User Interface
  • Performance
  • A frame-by-frame drawing option ( as pencil and toonboon; like a light table )
  • The controls not as layers, but attached to each object or group objects

These are the main complaints; but with this project I am trying to prove that is possible to built even with this issues. The problem is that with this issues, the work is harder and slowly.

But I’ll keep going and I will build the GUI in Inkscape as a suggestion for a user friendly interface.

Thanks again!


Just out of curiousity. If you had to name one aspect you found annoying/limiting as a first time Synfig user, and a second as a more experienced Synfig user, what will these two be? What simple things (in your estimations) can be done to make people more efficient and effective with Synfig? A complete redesign of the GUI is unlikely, just as new/other forms of animation, but smaller more specific problems might be solvable…


Hi Gerco,

Well, as a first time Synfig user; I got scared about the low performance playback ( when we test it with basic shapes it’s ok, but when we work with characters we find a big trouble because we can’t see our animation running well to refine it ); the second scare was how the controls work; we need to create then with layer options ( zoom, rotate and translate ), this is not common, we usually see these controls with the objects when we import or create then. It’s time consuming.

Now that I am more experienced with Synfig, I think that the playback performance in the workwindow is the big problem. It’s difficult to test the “timing” of the animation. The character animation is not so accurate as it should be with a short time of work. I need to construct animation, test it with the preview ( too slow! ), and then rebuild, and then again the “preview”, rebuild, “preview”… It’s time consuming!

Some complex animations I am building with Pencil 2D and then vectorizing it with Inkscape, and then importing to Synfig a Image list of the frame-by-frame animation. And this is one more idea to synfig, to have the possibility to create frame-by-frame drawings. If you know anime studio, it has a good solutions for this: we could have a kind of layer ( just an idea as it is in anime studio ), that let us draw frame by frame drawings, pointing the timeline to the time we want the new frame; and all of this drawings are stored inside of this layer.

And sure, we need the audio track! Without the audio track we can’t compose an animation with dialogs or a final edit of a scene.
I am building the scenes with Synfig and editing then inside of Blender ( using the sequence editor ). Remembering that, this animation has no dialogs.

Another feature I forgot to mention to Genete is the “Animation Field Guides” , I Have a Bitmap file here that I use as a referende, but I think that the soft could come with then: take a look at these sample: http://www.animationpost.co.uk/novice_notes/field-sizes.htm

I agree with you that the user interface is not the problem for me now, I know how to work with Synfig very well; I have tested all the features and discovered all the limitations to work well with it, but we must agree that when we have our first look at it , we get scared.
Let’s wait for the OpenGL feature to see if we have a better performance.

I’l keep going here. Today I am in the middle of the first frame-by-frame character animation…it’s time consuming and too hard, but the result is going fine. ( Pencil 2D and Inkscape ).

Thanks for the interest!

I “finished” ( some adjustments needed ) the first frame-by-frame animation yesterday late at night:

And here the avi file:

I just posted here to show you a frame-by-frame sequence. It’s a different effect from cut-out animation; that’s why I use both techniques.

Just a simple question: Is there a way to create drop shadows with Synfig? I am creating the drop shadow effect by duplicating the encapsulated layer of the animation and then applying the color adjustment layer + rotate layer + stretch layer; it works, but… could be better.
Any help?


There isn’t automatic drop shadow. That’s basically the way it is.

BTW, Ricardo, maybe this thread is interesting for you:


There is shade with behind blend mode

Yeah! but at the moment you cannot rotate or shear that shadow. It would be a good feature though!

Hi Genete,

This is a nice solution for image exchange in a animation; I like it, but I really did not understand how you did it. When I whant to exchange a hand pose as you did, I just use the “constant” option with keyframe and change the “amount” of two imagens, one to 1 and the other to 0. But the way you did seems faster. I´ll try to digg it, but if you can tell us how you did it… will be fine.

I would do a small tutorial when I have a little of more time. Oh man, lots of things TODO!


Ok, Genete, after our chat in the IRC channel… I´ll try and if I really get it, with your permissions; giving you the credits, I can post here the step-by-step process. I didn´t have the time to try it yet, but after our chat, I copied and pasted the text to a note file.

I think that the process is not dificult ( I really understood all the steps you wrote ), but it is very powerfull and the animation process seems to be faster. Congratulations to you!!! Nice solution!

I´ll try this with the new animation this week and keep you informed about it.

Thanks again!

For drop shadows, you can experiment with a Shade layer composited ‘behind’. You’ll find it under Stylize in the layer menu. It’s not very flexible, but it may help in some cases??


Ok pixelgeek, I´ve already tested; yes it works, but with limitations. When I need a perspective shadow ( as you can see in the image of the boy ) I need to duplicate the layer containing all the character animation, put it under the first one and then apply:color adjustment layer + rotate layer + stretch layer…but thanks anyway!!