Enumeration of precision fixes for feeding 64-bit floating-point position coordinate values into the GPU pipeline.
enum
|
PositionPrecision
|
Float
|
= 0
|
Coordinate values are cast to 32-bit floating-point values (loosing precision in the process) and written to the vertex buffer as 32-bit floating-point values. Remarks: Assuming
that
32-bit
floating
point
precision
is
preserved
in
the
GPU
pipeline,
the
Float
coordinate
precision
can
be
used
for
maps
of
a
Size
up
to
|
---|---|---|
FloatDelta
|
= 1
|
Coordinate values are cast to 32-bit floating-point values and subtracted from the original values. The resulting delta values are written to the vertex buffer as 32-bit floating-point values. Remarks: Assuming that the GPU pipeline is capable of preserving 23-bits of floating-point mantissa precision, maps up to MaxSize can be used (when used in conjunction with Float). |