The Flow Seeding settings determine the spatial location of points used to start the flow integration. These points are either determined by a Rake or as a list of points. A Rake is just a 3D box that determines the maximal extents of the generated seed points. The location of the Rake is specified by 6 sliders for the rake center and size in each dimension. The rake position can be controlled using the mouse in the 3D scene in Rake mode.
There are three different ways to specify the seed points, as indicated by a selector in the middle of the Flow Seeding settings:
- Random rake : The seed points are randomly selected from within the rake bounds. The number of such points is specified by the seed count.
- Nonrandom rake: The seed points are on an evenly spaced grid, bounded by the rake, and of length in each dimension as specified by the selectors indicated by "By dimension X: .. Y: .. Z: ..". The width of the rake in each dimension is divided into (k+1) equal intervals, where k is the length in that dimension. The seeds are positioned at the k endpoints of those intervals excluding the 2 exterior rake bounds.
- List of seeds: The seed points will be a list of (x,y,z,t) coordinates, unrelated to the rake bounds. These seed points may be specified by clicking "Edit Seed List", or may be read from a file by clicking "Load Seeds". Points in the seed list can also be specified using a cursor in the Probe, or in the 2D tab or in the Image tab. The points in the list that are used to seed the integration must have t coordinate equal to the current time step and the spatial coordinates must lie inside the current Region. If the t coordinate is "*" or -1, then that point is used as a seed at all time steps. The file format for a list of seeds consists of one line containing a 4-tuple of values for each seed.
- Seed Distribution Bias: When using the Random Rake, it is possible to bias the selection of random seeds to obtain seeds for which a field is large or small. The components of the field used to bias the seed distribution are specified by the three selectors indicated by "Seed Distrib. Field". The Bias slider and text box are set to 0 by default, resulting in a uniform seed distribution. When selecting N points with bias value B greater than 0, N*2^|B| points are randomly generated, then the N points with largest value of the seed distribution field are used as seed points. When B is negative, points are chosen with smallest values of the seed distribution field. B must be between -15 and 15.
When using Field Line Advection, there are several additional flow seeding parameters, controlling how flow lines at one time step are used to produce seed points at the next time step.
