Hello everyone.
Just decided to take a break from the cartoony style and try to get some realistic animation. This exercise will consist of the animation of the following sequence:

The important thing here is that this kind of sequences are commonly used as reference for videogames, and as this medium is different from movies (here you have to make it right from any angle and not only to that of the camera), for some of the movements you just have to 'figure it out' how the body mechanics are working on. This is because as you can see, there are some takes where you don't get to see the character legs or you just get to see a body part like the head or fist for one of them.

After a couple of days I came with the following results. This is the animation based on the same camera movements from the reference:

While doing Pre-Viz work for movies, its a common thing to find issues like continuity in the footage you are provided with (like a punch repeated several times Van-Damme style), and you have to figure out the better way to create the animated sequence so it can be seen correct and from any angle.

At some points in the reference (usually on the camera cuts) the characters are in different positions and time vs the previous take. This is a common thing in filmaking for the sake of having the choreography a bit more spectacular, and also because its clear that it is very difficult to have the stunt men exactly at the same spot while changing the camera and retaking that last kick or punch from another angle. Thats also why Pre-Viz animation is so important in film making.

Here is a second take of the same sequence with different camera positioning and using the same animation.

The idea here is not necessarily to have a super polished material. Pre-viz animation allows the director to reposition the camera in the computer allowing him to choose the better angles for the sequence before shooting it on stage; which otherwise would be a very expensive way to figure out which camera takes works best for each scene.