NAME

       TDINIT - Initialization routine for TDPACK, called to define the position of the eye, the
       position of the point looked at, which way is up, and whether or not a stereo view is to
       be done.

SYNOPSIS

       CALL TDINIT (UMID, VMID, WMID, UORI, VORI, WORI, UTHI, VTHI, WTHI, OTEP)

C-BINDING SYNOPSIS

       #include <ncarg/ncargC.h>

       void c_tdinit(float umid, float vmid, float wmid, float uori, float vori, float wori,
       float uthi, float vthi, float wthi, float otep)

DESCRIPTION

       This initialization routine is called to define the position of the eye, the position of
       the point looked at, which way is up, and whether or not a stereo view is to be done.
       TDINIT precomputes some quantities in TDPACK common blocks that will subsequently be used
       in projecting points from 3-space (U, V, and W coordinates) to 2-space (X and Y
       coordinates).

       By default (that is to say, if the internal parameter 'SET' has its default value), TDINIT
       also calls the SPPS routine SET to define the mapping from the "user" coordinate system
       (the X/Y system) to the "fractional" coordinate system (in GKS terms, NDC space). This is
       done in such a way as to show in the projection plane a field of view of 'FOV' degrees.
       The viewport to be used in the fractional coordinate system is that defined by the values
       of the internal parameters 'VPL', 'VPR', 'VPB', and 'VPT'. In some situations, it may be
       desirable, after calling TDINIT, to call the SPPS routine GETSET to retrieve the arguments
       with which TDINIT called SET and then recall SET with a different set of arguments.

       Normally, the creation of an image with TDPACK starts with a call to TDINIT (perhaps
       preceded by calls to TDSETI, TDSETR, and/or TDSTRS to reset internal parameters of the
       package) and continues with calls to draw objects. (Of course, if all of the arguments in
       a call to TDINIT have the same values as in the last call and the SET call that was done
       as a result is still in effect, then it's not necessary to repeat the call to TDINIT.)

       For stereo views, one calls TDINIT with a negative OTEP, executes the object-drawing
       calls, calls TDINIT again with a positive OTEP, and then repeats all of the object-drawing
       calls. The exact way in which stereo views are drawn is also affected (slightly) by the
       value of the internal parameter 'STE'.

       The arguments of TDINIT are as follows:

       UMID, VMID, and WMID
               (input expressions of type REAL) - the coordinates of a point, E, at the eye
               position (if a single view is being drawn), or of a point midway between the two
               eyes (if a stereo view is being drawn).

       UORI, VORI, and WORI
               (input expressions of type REAL) - the coordinates of a point, O, that the eye is
               looking at.  That point defines the origin of the XY projection plane.  The line
               of sight is the line from E to O.  The projection plane passes through O and is
               perpendicular to the line of sight.

       UTHI, VTHI, and WTHI
               (input expressions of type REAL) - the coordinates of a third point, T, needed to
               completely specify the orientation of the X and Y axes in the projection plane.
               The Y axis of the projection plane is its intersection with the plane passing
               though the points E, O, and T.  The X axis of the projection plane passes through
               the point O and is perpendicular to the Y axis.

       OTEP    (an input expression of type REAL) - set non-zero if and only if a stereo view is
               to be drawn.  Make the value negative to draw a left-eye view, positive to draw a
               right-eye view.  The magnitude of OTEP is the distance from either eye to the
               point midway between the eyes; if R represents the approximate distance from the
               eye to the objects being drawn, then OTEP may be set using a statement like

                 OTEP = (+ or -) R *
                        TAN(.017453292519943*ANGD/2.)

               where the constant (.01745...) is just pi/180 and ANGD is the desired difference
               in the angle between the two views, in degrees; use a value of about 1 or 2
               degrees for ANGD.

               Note that, if the origin is approximately in the middle of the objects being
               viewed, then it's probably appropriate to use a value of R computed as follows:

                 R = SQRT((UMID-UORI)**2
                          (VMID-VORI)**2+
                          (WMID-WORI)**2)

C-BINDING DESCRIPTION

       The C-binding argument descriptions are the same as the FORTRAN argument descriptions.

ACCESS

       To use TDINIT or c_tdinit, load the NCAR Graphics libraries ncarg, ncarg_gks, and ncarg_c,
       preferably in that order.

SEE ALSO

       Online: tdclrs, tdctri, tddtri, tdgeti, tdgetr, tdgrds, tdgrid, tdgtrs, tditri, tdlbla,
       tdlbls, tdline, tdlnpa, tdmtri, tdotri, tdpack, tdpack_params, tdpara, tdplch, tdprpa,
       tdprpi, tdprpt, tdseti, tdsetr, tdsort, tdstri, tdstrs

COPYRIGHT

       Copyright (C) 1987-2009
       University Corporation for Atmospheric Research
       The use of this Software is governed by a License Agreement.