What is Movimento?

Movimento is REALVIZ’ new video-based motion capture software. Movimento offers the unique capability to capture the motion of any non-rigid object (human body, human face, animal…) in post-production, from synchronized image sequences with total flexibility:
o From 2 cameras to as many as needed.
o Cameras can fixed or moving.
o Cameras can be of different frame rates and different definitions.
o No need for a motion capture studio.

REALVIZ Movimento is an affordable, accurate and very flexible software to cover the needs for facial, hands, arms and full body motion capture. Movimento can also combine the matchmoving (3D camera tracking) and the motion capture processes for on-set mocap.

What is “on-set mocap” ?

On set mocap is the process of capturing the actor’s motion while shooting him (her). This process has been extensively used by REALVIZ customer FIDO (Sweden, http://www.fido.se) for a TV commercial for Nokia in 2006, in which some moving tattoos needed to move on the actor's body. This process may mix matchmoving (i.e.
computing the 3D camera path) and motion capture at the same time. This process also requires several cameras to be used. A typical set up is to use the main camera (35
mm for example), as well as 2 DV cameras on the side.

Where does Movimento technology come from ?

Movimento is derived from REALVIZ expertise in 3D tracking and from the successful sale of MatchMover Pro in the visual effects industry for more than 8 years. A first “mocap-oriented” version of MatchMover, called MatchMover Mocap, has been on the market for 3 years, and many high-end post-production houses have been using it for several projects (films and TV commercials) extensively using on-set mocap. Movimento is “Powered by SMART” (Scalable Matching Architecture for Tracking) as is all other 3D tracking software from REALVIZ.

How does Movimento differ from other motion capture systems ?

Other motion capture systems are either optical (such as the ones manufactured by Vicon or Motion Analysis) or magnetic (such as the ones developed by Polhemus). Optical systems use several custom-designed cameras and retro-reflective markers, and they must be used in dedicated studios. Magnetic systems use sensors placed in a magnetic environment, and, all the same, it is necessary to get a specific, iron-free, location to use such a system. In both cases, the exported data are tracked 3D points versus time. In both cases, the computation of 3D points can be done in real time, although there is typically a large amount of cleaning to be done on the computed data before they can be used in a 3D animation software.

Movimento is based on tracking features in synchronized video sequences. Whatever information can be tracked in a video can be used to compute 3D points. This is a post-processing technique, which means that capturing the videos and computing the 3D data are done in sequence. With video-based motion capture, there is no need for a specific and dedicated studio, and the cameras can be fixed or moving. So you can track the motion of a sports star in his real environment, or an animal in a zoo. No need to bring them to the studio for this purpose. For biomechanical or ergonomic studies, you can also capture the motion of any subject in his real working environment.

What are the advantages / drawbacks of video-based motion capture versus optical systems ?

Optical systems are widely used in the animation, scientific and medical industries. They offer robust and accurate data. But their set up is complex, and their cost is very high. Cleaning of data can be very time consuming. When it comes to facial motion capture, the complexity of the process makes it less attractive compared to key frame animation. The recorded data are the resulting 3D tracks, but you cannot save the original images corresponding to your capture for further processing or understanding of the actor’s performance if needed.

A video-based solution such as Movimento offers greater flexibility and costs much less. The recorded videos can be stored and re-used several times. The video playback
of the performance can prove to be very useful, particularly for facial motion capture when you need to post-synchronise with the voice. You can also mix motion capture and matchmoving for on-set mocap, or you can use moving cameras to track a running character.

How many cameras are needed with Movimento ?

A point needs to be visible in at least two cameras to be reconstructed in 3D, through triangulation. Therefore, a minimum of 2 cameras is needed in Movimento. However,
because of occlusions (i.e. the fact that some tracked points can be temporarily or constantly hidden to some of the cameras during the capture) it is often necessary to use more cameras to make sure that points are always visible from at least two different viewpoints. The number of cameras depends on the kind of subject tracked, and the type of the motion to be captured.

Which export format does Movimento export ?

Movimento exports the following formats:
- FBX files (*.fbx)
- Motion Analysis (*.trc)
- Maya files (*.ma)
- MaxScript files (*.ms)
- XSI files (*.xsi)
- LightWave 3D files (*.lws) - REALVIZ XML (*.rzml)

- C3D files (*.c3d) (coming fall 2006)

Pour en connaitre plus sur les produits Realviz et Movimento, contactez nous : info@3dvf.fr