transitions dilemma

I’m making an animated demonstration video like this style where objects enter the screen, animated then they exit. I can’t divide it into multiple scenes (with multiple sifz’s) as long as in some transitions two objects of two different scenes are on the screen in the same time, so I have different options:
1- divide them into different scenes and create special scenes for the transitions which is effort wasting and time consuming thing.
2- make the whole animation in one file, but it means that a lot of objects will keep animating till the end of the video although they are no longer shown on the screen (I guess this is memory consuming and I’m afraid that the program would crash).

so what do you recommend? I prefer the second option, is it dangerous?

You have only to worry about memory consumption if the objects are imported image layers. If they are vector, just disable them and work with the rest for each scene. At the final render, enable them all and render the whole animation.

Alternatively you can do this:
Create a master file with all the objects involved in the whole animation.
Disable all the objects that doesn’t appear from time t=0 to time t1.
Create the animation over the master document. Then save the master document as scene 1.
Continue working with the master document the same way but from t=t1 to t=t2. Then save the master document as scene 2.
Repeat the above as much as you need.
If it crash or something is wrong, just rewind one scene.
Finally render all the scenes and compose them in a video editor.

Good luck!

I’ll make it anyway, although I’m using vectors, it’s a safer workflow, thank you.

I’ve been thinking about an alternate workflow when composing many scenes (objects, planes, characters, etc) into a complete animation.

Assuming each individual component is rendered into an image sequence with transparency, then it should be possible to compose them all into the final product.

So what I’m proposing is to have a GUI tool (separate from Synfigstudio!!!) which can:

  • Load image sequences (ie. PNG file list, PNG spritesheet, video files, etc)
  • Compose them together (ie. layers, blend methods, etc)
  • Time shifting sequences in order to synchronize them
  • Maybe apply some effects
  • Encode the result in some format

Is there already a tool which can do the above effectively? I know it’s possible with some scripting and imagemagick. But having a GUI tool for this would make it much easier.

Blender? :wink:

Isn’t that a bit overkill? :wink:

Actually Blender is a great compositor, I use it for my projects. I wouldn’t say overkill, you can do what you want but still you have the power of a full fledged compositor if you want to do anything more complicated than combining render layers.

I saw there was just now a new rebirth announced for Jashaka (the last rebirth was announced a few years ago but nothing happened then… Perphaps this time?) which is an open source compositor. Don’t get your hopes up too high on that one.
There also was a promising floss compositor started called Ramen but I think that project died.