In the studio, gift effects are indispensable to promote the platform revenue, add a means of the studio atmosphere, currently in the industry, more mature programs include the use of sequence frame, SVGA, Lotties, MP4 and other programs. Each solution has its own specific usage scenario. It is generally divided into the following categories:

  1. Pure process-oriented animation recording method

    For example, Airbnb launched Lotties

  2. Half – process – and – result – oriented animation recording

    For example, SVGA scheme launched by YY

  3. Pure results-oriented animation recording

    For example, MP4

Among them, the process-oriented animation recording mode needs to restore the recorded process when rendering at the player end. If complex dynamic effects are involved, the performance consumption of rendering will be relatively large, which is suitable for some simple special effects. The semi-process-oriented and semi-result-oriented animation recording method makes up for the above shortcomings to a certain extent. For some complex special effects, the way of embedding pictures can be used, but it will increase the volume of resources. Pure results oriented way, such as MP4, at the time of rendering, the client only needs to decode the every frame image, and then draw to the screen, this way to restore all the animation designer, what you see is what you get, at the same time can support 3 d dynamic effect, is currently live in some more big gift scenarios, advocate a kind of way.

Let’s use another real life example to understand the difference between process-oriented and results-oriented animation

Take the winter Olympics, figure skater Guo’s movement process as an example:

  1. The process-oriented recording mode is

    Kuo floats y meters to the left, spins around and a half, falls a crab step, and then reverses around and a half. In order to restore this animation, the process-oriented calculation process is quite complicated. It takes several weeks of rotation to calculate the characteristics (shape, size, position, etc.) of each frame of the player and then render.

  2. Results-oriented animation recording

    Athlete Xiao Guo, Frame 1 features (shape, size, position, etc.), Frame 2 features…… Frame n features. Because the features (shape, size, position, etc.) of each frame of athletes are recorded, it is not more complicated than the first example. By eliminating these complex calculations, complex animations can be played more smoothly