Table Editor

The table editor displays an animation table in a simple text editor so that the contents can be viewed and edited with the mouse and keyboard. In addition, the table editor provides many other functions useful for animation, such as interpolation, mathematical operations on columns, and various input/output options.

Each table editor is labeled with an identifying number, which is displayed at the top of the window. Below that label is a column bar which numbers the columns. This is useful for keeping track of which column is which during horizontal scrolling of the table.

Beneath the column bar is the main editing window. If there is a current animation curve when the table editor is created, then the contents of the curve are automatically loaded into the editing window. This creates a table with four columns for the time, x, y, and z parameters, and one row for each node in the curve.

After clicking on the editing window with the left mouse, the user can edit the text. The editor event bindings are the default bindings for a Tk text widget, described in detail in the Tk documentation under libtk/doc/text.n. Some of the more important bindings are listed below:

mouse 1, click			- move insertion cursor
mouse 1, drag			- select text
mouse 1, 3 clicks		- select line
mouse 1, 3 clicks and drag	- select lines
mouse 2, click			- insert selection
Insert				- insert selection
mouse 2, drag			- drag view
arrow keys			- move insertion cursor
shift-arrow keys		- select text
CTRL-/				- select entire contents
Meta-w 				- copy selection to clipboard
CTRL-w				- cut
CTRL-y				- paste
In addition, the editor supports many commonly used emacs key bindings (CTRL-f, CTRL-k, etc.).

The buttons which implement the rest of the table editor's capabilities are found in a menu bar at the bottom of the window.

Write
The contents of the table editor can be written to a file, curve, or view curve. (See the next section to learn about view curves). After specifying one of these three destinations, the user is asked for the name of the file, curve, or view curve to write to, and which columns of the table should be written. While a file can accept any number of columns, curves and view curves are picky about the number of columns they accept. A curve can accept either three or four columns; in the latter case, the columns are interpreted as "time x y z", and in the former case, columns are interpreted as "x y z" and the time of each node is set to the node index. A view curve may need anywhere from seven to nine columns, depending on its view format. If the number of columns in the table is incompatible with the destination, an error message is displayed.

To specify which columns of a table will be written to a given destination, the following syntax is used. The string all in this context represents all the columns. Comma-separated integers are used to identify individual columns: 0 is the first column and n-1 is the last of n columns. Dashes represent ranges. For example, 2-4 is identical in meaning to 2,3,4; 4-2 is the same as 4,3,2; and 4- represents columns 4 through n-1.

Read
The table editor can read data from a file, another text editor, a curve, or a view curve. When the Read menu is posted, it contains a from File option and one option for each of the existing editors, curves, and view curves.

If from File is selected, then the user is asked to specify which columns of which file to read. Buttons on the bottom of the window are used to select one of three possible insertion modes: Replace, Append, and Add Columns. Depending on the insertion mode, the new data will replace the previous table, will be appended on the bottom in new rows, or will be added to the right side of the table in new columns.

If any of the editor, curve, or view curve options is selected, the user is similarly asked which source columns should be read and which insertion mode should be used. The syntax for specifying columns is the same as that used with the Write option, above.

Clear
This button clears the contents of the table editor.

Interpolate
This button creates an interpolation widget for interpolating the data in an animation table. When using the interpolation widget, the first column of the animation table should always be the time column. Each of the other columns is interpolated with respect to the first column, using any of a variety of algorithms, including step, linear, and spline interpolation. The interpolator is implemented by a system call to tabinterp.

The interpolation widget contains three boxes which control the temporal limits and resolution of the interpolation. The output table begins at the time labeled Start Time and ends at the time labeled End Time. The number of output rows created for each second of time is controlled by the Frames Per Second parameter. The start and end times default to the initial and final times in the input table, and the frames per second parameter defaults to 30 fps, a common value for animations intended to be shown on video.

Above these time-control boxes are column-control boxes, labeled by the index of the output column to which they correspond. Column zero of the output table is always the time column. Then the box labeled 1: controls the contents of output column 1, and so on. For each desired output column, the user must provide an interpolation command in the corresponding box. To delete an output column, the user need only delete the interpolation command from the corresponding entry box. To add an output column, the user creates an new entry box with the Add Column button and enters an interpolation command.

The valid interpolation commands are described below. The notation i is an integer representing the index of a column of the input table, and j represents the index of a column of the output table.

step i
Interpolate column i of the input table with step functions. In other words, values only change in instantaneous jumps at the times specified in the input table.

linear i
Interpolate column i of the input table with a piecewise linear function.

spline i
Interpolate column i of the input table with a cubic spline curve with open end conditions.

cspline i
Interpolate column i of the input table with a cubic spline curve with periodic end conditions. Note: the row of the input table corresponding to End Time is ignored and assumed to be equal to the row corresponding to Start Time.

quat i
Register column i of the input table as a quaternion parameter. Four consecutive columns of the output table should contain quaternion commands, in which case the four columns will be interpolated jointly using quaternion interpolation. If the number of consecutive quaternion commands is not a multiple of four, an error message is displayed.

rate start increment
Create a rate-based column. The first row of the column will contain the start value, and then successive rows will be change linearly so that increment units are added for every second of time.

accel start increment
Create an acceleration-based column. The first row of the column will contain the start value, and then successive rows will change exponentially so that the value is multiplied by increment during every second of time.

next j offset
This command creates an output column which is a copy of the output column indexed by j. The column is shifted forward in time by offset rows. For example, the command next 2 3 creates a copy of output column 2 with the first three entries removed. The last entry in the column is repeated three times so that the resulting column is the same length as column 2. Negative values are permissible for the offset. It is an error for a next column to reference itself or a non-existent output column.

By default, the interpolation widget uses spline interpolation on each column of the input table, but all other interpolation types are available by changing the interpolation commands. The Active Command menu provides a shortcut for changing the commands for several columns at once. When the user presses on the label of a given column, the interpolation command for that column is replaced with the active command. For example, all of the columns could be changed to linear interpolation by selecting the entry labeled Linear (src) from the Active Command menu and then clicking on each of the column labels in turn.

The text on each entry of the Active Command menu reminds the user of the arguments each command requires. For example, the label Linear (src) is a reminder that the linear command requires an integer argument specifying the source input column.

Edit Columns
This function pops up a column editor which allows mathematical operations to be applied to the columns of an animation table. The editor consists of three parts: the column-control entry boxes, the button bar, and the Number of Rows field.

Each column-control entry box is labeled by the index of the output column to which it corresponds. For each desired output column, the user must place an expression defining its contents in the appropriate entry box. With the exception of some substitutions described below, these output column definition expressions are treated like normal Tcl expressions. This means that the whole gamut of Tcl operators (*,+,!, etc.) and functions (cos, sqrt, floor, etc.) can be used. Thus a column definition such as: sqrt(2.0)/2.0 results in a column in which the value of every row is set to 0.707107. The precision of the result, which defaults to six significant digits, can be increased by setting the global variable tcl_precision to the desired number of significant digits. For example, ten significant digits are obtained by typing:

mged> set tcl_precision 10
in the MGED interaction window. Note that this will effect the precision to which expressions are evaluated in the entire Tcl environment.

Much of the usefulness of the column editor comes from the substitutions which can be performed on a row-by-row basis. A @ character followed by a substitution keyword in the column definition expression is replaced according to the following rules before the expression is evaluated:

For example, the expression @2 simply copies column two from the input table, and the expression (80.0*@i/(@n-1)+20.0) creates a column which varies linearly from a value of 20.0 in the first row to a value of 100.0 in the last row.

An output column can be deleted by deleting the column definition from the corresponding entry box. To add an output column, a new entry box is created with the Add Column button, and the desired definition is entered. The specified output columns are created when the OK button is invoked, and the output table appears in the place of the input table.

The parameter to the Number of Rows field determines how many rows the output table will have. By default, the field contains the string all, indicating that the output table should have the same number of rows as the input table. Any other number of rows can be specified by entering an integer in the box. Note that if a table is being created out of nothing (i.e. the input table is empty) then the number of rows in the output table must be specified explicitly or the length of the output table will also be zero.

Estimate Time
When creating an animation curve, the animator usually knows where the object should move and how long the whole motion should take, but he or she doesn't necessarily know the best time parameters for the intermediate key points. The Estimate Time function calculates appropriate time values for an animation table containing an animation curve. The first column of the animation table must be the time column, and three other columns must specify a path through space. The Estimate Time function leaves the first and last values in the time column unchanged and adjusts the intermediate time values so that the resulting motion is smooth.

When the Estimate Time button is pressed, a window pops up asking for the desired start and end speed of the object. The user can specify an absolute speed, in millimeters per second, or a relative speed. A relative speed is specified as a percentage of the speed needed to traverse the path in the given time at a constant rate. For example, if it is intended that the object move at a constant speed along the entire path, the start and end speeds would both be set to 100%. (These are the default settings.) Or, if it is intended that the object start from rest and accelerate to a final speed of 2 mm/s, the start and end speeds would be 0.0 and 2.0, respectively. Starting and ending speeds must be positive and must not be greater than three times the rate corresponding to 100%, or an error message results. See the manual page of the underlying system call for more information.

The user is also asked which three columns of the input table specify the animation path. By default, it is assumed that these are columns one through three, but any three columns can be used.

Clone
This button creates a new copy of the current table editor It is good practice to create a clone before doing interpolations or complicated column editing, so that data is not accidentally lost.

Up
This button raises the parent of the table editor to the top of the stacking order.

Cancel
This button closes this table editor.

Next Section: View Curve Editor

Previous Section: Curve Editor

Index