Maya - Matchmoving #2 - 4

Continuing our matchmoving workshops, we've begun to track some live scenes using Autodesk's Matchmover. Essentially, the camera tracking data is translated from Matchmover to Maya, allowing one to construct proxy environment and place new objects into the scene with the confidence that it will appear correctly in shot.

We experimented with a number of methods from:

Autotracking (added sphere)


to manual tracking method known as Target Tracking (added sphere)


to a Masked Auto Tracking method, requiring one to maintain the mask throughout the scene (added wind-up Key).

Maya - Matchmoving

In this workshop, we're beginning to look into the process of matchmoving with Maya - effectively blending live-action scenes with CG elements.

Here, we took a static image as a starting point. Without any prior information to the actual image or focal length settings, aligning the camera is a tricky yet integral part of the process.



Modelled Set
Once the camera is aligned, elements in the scene had to be modelled in order to achieve correct information with lighting, shadows and reflections. We then proceeded to add our CG element to the scene, an orrery, as well as some random spheres to test elements within our set such as the shadows and reflections.



Final Composite
After rendering out several passes (shown below), the shot is composited together to form the final scene. Note the integration of the orrery to the scene, the shadows on the sphere under the walkway, as well as the reflection to the far right sphere towards the back.



Various render passes
By rendering out different elements of the shot separately, these can then be effectively controlled and composited within the scene by adjusting any individual values. These were rendered out with render layers, and later composited in After Effects.







This ends the beginning introduction to matchmoving, where I believe we will continue by dealing with moving elements in the next workshop.