< Previous | Contents | Manuals Home | Boris FX | Next >
Sometimes you may be supplied with exact 3-D coordinates for a number of features in the shot, as a result of hand measurements, laser (lidar) scans, or GPS data for large outdoor scenes. You may also be supplied with a few ruler measurements, which you can apply as size constraints; we won’t discuss that further here, but will focus on some aspects of handling 3-D coordinates. The full details continue in following sections.
Important: If the shot contains a number of unordered frames, rather than a video shot, be sure to use SynthEyes's Survey Shot capability. SynthEyes normally assumes that each frame is only slightly changed from the previous shot, as in movies. If that assumption is not true, ie for a collection of random stills, Survey Shot mode is required. This is different than the setup of the XYZ coordinate data discussed here.
First, given a lot of 3-D coordinates, it can be convenient to read them in automatically from a text file, see the manual’s section on importing points, which will help you specify tracker locations.
Tip: You can export point locations from Google Earth as a JSON file, then use “File/Import/Trackers/Google Earth XYZ tracker positions” to set the coordinates of (existing) trackers with those names.
You can also use File/Import/Mesh to read in tables of XYZ data as lidar files (.XYZ or .PTS) to produce mesh objects; they can include per-vertex RGB information as well. You can snap trackers to those vertices, or add facets using the perspective view mesh edit capabilities. See the Lidar Meshes section for important information about reading lidar data.
You can use the Place mode of the Perspective view to lock trackers onto an imported mesh or lidar file, see Placing Seed Points and Objects.
Tip: See the “Animorphic Solve Using a Lidar Scan” tutorial in the Lens Distortion tutorial section and the SynthEyes 2304 Youtube playlist .
SynthEyes gives you several options for how seriously the coordinate data is going to be believed. Any 3-D data taken by hand with a measuring tape for an entire room should be taken as a suggestion at best. At the other end of the spectrum, coordinates from a 3-D model used to mill the object being tracked, or laser-surveyed highway coordinates, ought be interpreted literally.
Trackers with 3-D coordinates, entered manually or electronically, will be set up as Lock Points on the Coordinate System panel, so that X, Y, and Z will be matched. Trackers with very exact data will also be configured as Pegs, as described later.
If the 3-D coordinates are measured from a 2-D map (for a highway or architectural project), elevation data may not be available. You should configure such
trackers as Any Z (Z-up coordinates) or Any Y (Y-up coordinates), so that the XY or XZ coordinates will be matched, and the elevation allowed to float.
If most of your trackers have 3-D coordinates available to start (six or more per frame), you can use Seed Points solving mode on the Solver control panel. Turn on the coordinate system panel’s Seed button for the trackers with 3-D coordinates. This will give a quick and reliable start to solving. You must use Seed Points and Constrain modes on the solver panel if you are matching a single frame from survey data.
You can use the Nudge Tool (spinner) on the Coordinate System panel to adjust seed locations along the depth axis, towards and away from the camera. Right-clicking the Nudge tool will snap the seed point onto the line of sight of the 2D tracker position on the current frame. (This works for Far trackers also.) If you have visions of doing this from multiple frames, check out ZWTs instead, which will be faster and more accurate.
For more information on how to configure SynthEyes for your survey data, be sure to check the section on Alignment vs Constraints.
©2024 Boris FX, Inc. — UNOFFICIAL — Converted from original PDF.