Previzion records complete camera metadata into a Collada file.
- To specify the name of the take file to be recorded, enter it in the Take box.
- To start recording, click Roll.
- To stop recording, click Cut. A Collada file is automatically saved to the specified directory.
To change the directory in which takes are saved, click the Browse button and enter the new directory.
Other options include:
- Using the Screen Text data fields to automatically generate the name of the take
- Saving editable transform values at the time of the take’s capture into a separate file
- Capturing incremental thumbnail images
Either 1/3 size or full resolution thumbnail images can be captured. 1/3 size are recommended for most general use, due to the processing requirements to capture full resolution images.
Full resolution images are occasionally useful when troubleshooting a specific visual artifact.
Changing Take Name
The take name and take data (company, show, comments, etc.) can be changed during the take, up to the point where ‘cut’ is clicked. Previzion will then change the names of the following files to the new name:
- DAE take data file
- Take directory
- Saved transform text file
- JPEG thumbnail images
Things that are not changed:
- The screen text that is ‘burned in’ to the image of the JPEG files
Take Data Contents
The recorded Collada file includes:
- Camera position and orientation in a 4×4 matrix to avoid Euler angle errors
- Camera focal length and focus distance
- Subject distance (video plane setting)
- Camera aspect ratio
- Camera sensor width
- Camera frame rate
An example of a test Collada file can be found here.
This has been tested with Maya 2009 and the ColladaMaya 3.05C Collada plug-in. The link to the ColladaMaya 3.05C plug in is here.
When imported into Maya, the take data must be set to be used without interpolation in the following manner:
The Previzion motion captured camera data must be highlighted, and the red outlined “step” button clicked to tell Maya to use the data without interpolation. Otherwise, incorrect results will happen such as the camera flipping up and down near Euler angle limits.
Each application has a slightly different flavor of Collada in terms of camera handling. The default Previzion-generated Collada file has a number of custom tags in it to handle all of the metadata that Previzion records. However, each application handles camera field of view and imager size somewhat differently. To convert from the Previzion Collada files to the application-specific data files, Lightcraft provides Take Converter, a conversion tool capable of both GUI and command line operation.