Interface and method of interfacing between a parametric modelling unit and a polygon based rendering system
An interface for use in a 3-d graphics system comprising a parametric modelling unit for modelling objects as high order surfaces, and a polygon based rendering system for rendering polygon modelled objects for display. The interface comprises an input for receiving data and a subdivision unit coupled to the input for processing the data. The interface includes a converter coupled to the subdivision unit for determining from leaf patch data a first plurality of values representing vertices of tessellating polygons describing the leaf patch, and for determining from sub-leaf patch data a second plurality of values representing the vertices of tessellating polygons describing the sub-leaf patch. The interface also has a combiner, coupled to the converter, for combining the values to form leaf polygon data defining the polygon vertices at a first subdivision level, and an output coupled to the combiner for outputting the leaf polygon data.
1. A 3-d graphics system comprising an apparatus for determining a surface normal vector for a vertex of a surface patch used in modelling of an object to be shaded by the system, the apparatus comprising:
means for deriving the locations of control points required for derivation of a surface normal for a vertex;
means for deriving a set of three candidate tangent vectors at a vertex from the control point locations;
means for selecting two tangent vectors from the candidate tangent vectors; and
means for deriving a surface normal from the selected tangent vectors.
2. A 3-d graphics system according to claim 1 , wherein the means for deriving a set of three candidate tangent vectors further comprises:
means for determining a candidate tangent vector T S in the S direction;
means for determining a candidate tangent vector T T in the T direction; and
means for determining a candidate tangent vector T D in the diagonal direction.
3. A 3-d graphics system according to claim 2 , wherein the means for selecting two tangent vectors from the candidate tangent vectors is adapted to:
select tangent vectors T S and T T when tangent vectors T S and T T are non-zero and linearly independent;
select tangent vectors T S and T D when tangent vector T S is non-zero and tangent vector T T is zero;
select tangent vectors T T and T D when tangent vector T T is non-zero and tangent vector T S is zero.
4. A 3-d graphics system according to claim 3 , wherein the means for deriving a surface normal from the selected tangent vectors is adapted to:
calculate N=(−T T )×T D , when T S =0, and output N as the surface normal;
calculate N=T S ×T D , when T T =0, and output N as the surface normal; and
calculate N=T S ×T T , when T S and T T are nonzero, and output N as the surface normal when N is nonzero, otherwise when N=0 calculate N′=T S ×T D and output N′ as the surface normal.
5. A 3-d graphics system according to claim 2 , wherein the means for estimating a candidate tangent vector T S in the S direction is adapted to:
(a) estimate T S as being equal to T S1 , where
T S1 =P 10xyz P 00w −P 00xyz P 10w
if P 10xyz P 00w is not equal to P 00xyz P 10w ; or
(b) estimate T S as being equal to T S2 , where
T S2 =P 20xyz P 00w −P 00xyz P 20w
if P 20xyz P 00w is not equal to P 00xyz P 20w ; or
(c) estimate T S as being equal to T S3 , where
T S3 =P 30xyz P 00w −P 00xyz P 30w
if P 30xyz P 00w is not equal to P 00xyz P 30w .
6. A 3-d graphics system according to claim 2 , wherein the means for estimating a candidate tangent vector T T in the T direction is adapted to:
(a) estimate T T as being equal to T T1 , where
T T1 =P 01xyz P 00W −P 00xyz P 01w
if P 01xyz P 00w is not equal to P 00xyz P 01w ; or
(b) estimate T T as being equal to T T2 , where
T T2 =P 02xyz P 00w −P 00xyz P 02w
if P 02xyz P 00w is not equal to P 00xyz P 02w ; or
(C) estimate T T as being equal to T T3 , where
T T3 =P 03xyz P 00w −P 00xyz P 03w
if P 03xyz P 00w is not equal to P 00xyz P 03w .
7. A 3-d graphics system according to claim 2 , wherein the means for estimating a candidate tangent vector T D in the diagonal direction is adapted to estimate T D according to the formula T D =P 11xyz P 00w −P 00xyz P 11w .
8. A 3-d graphics system according to claim 1 , wherein the means for deriving a surface normal from the selected tangent vectors is adapted to test whether the derived surface normal is zero, and when the derived surface normal is zero to:
cause an alternative two tangent vectors to be selected; and
use the alternative two tangent vectors to derive an alternative surface normal, and output the alternative surface normal as the vertex surface normal.
9. A method for determining a surface normal vector for a vertex of a surface patch used in modelling of an object to be shaded in a 3-d graphics system, the method comprising the steps of:
deriving the locations of control points required for determining a surface normal for a vertex;
deriving a set of three candidate tangent vectors at a vertex from the control point data;
selecting two tangent vectors from the set of candidate tangent vectors; deriving a surface normal from the selected tangent vectors; and rasterizing the image for display.
10. A method for determining a surface normal for a vertex according to claim 9 , wherein:
the step of deriving the locations of control points comprises the sub-steps of:
(a) deriving a first subset of control points along one edge of the patch adjacent to a chosen corner,
(b) deriving a second subset of control points along the other edge of the patch adjacent to the chosen corner, and
(c) deriving a third subset of control points consisting of a control point off-set from the corner of the patch in each of a set of chosen parameter dimensions; and wherein
the step of deriving three candidate tangent vectors comprises deriving first, second and third candidate tangent vectors from each respective subset of control points and the corner points.
11. A method according to claim 10 wherein a candidate tangent vector at a corner control point is derived by repeatedly selecting control points from a given subset of control points in any given order; and
deriving a first weighted vector by multiplying a chosen control point's components by a weighting component of the corner control point;
deriving a second weighting vector by multiplying the components of the corner control point by a weighting vector of a chosen control point;
deriving a difference vector between a first and second weighted vectors; and
using the difference vector as a candidate tangent vector if it is a non-zero, and otherwise progressing to the next control point of the subset.
12. A method for determining a surface normal vector for a vertex according to claim 9 , wherein the step of deriving three candidate tangent vectors at a vertex further comprises the steps of:
estimating a candidate tangent vector T S in the S direction;
estimating a candidate tangent vector T T in the T direction; and
estimating a candidate tangent vector T D in the diagonal direction.
13. A method for determining a surface normal vector for a vertex according to claim 12 , wherein the step of selecting two tangent vectors from the candidate tangent vectors comprises:
selecting tangent vectors T S and T T when tangent vectors T S and T T are non-zero;
selecting tangent vectors T S and T D when tangent vector T S is non-zero and tangent vector T T is zero;
selecting tangent vectors T T and T D when tangent vector T T is non-zero and tangent vector T S is zero.
14. A method for determining a surface normal vector for a vertex according to claim 13 , wherein the step of deriving a surface normal from the selected tangent vectors comprises:
calculating N=(−T T )×T D , when T S =0, and outputting N as the surface normal;
calculating N=T S ×T D , when T T =0, and outputting N as the surface normal; and
calculating N=T S ×T T , when T S and T T are non-zero, and output N as the surface normal when N is non-zero, otherwise when N=0 calculating N′=T S ×T D and outputting N′ as the surface normal.
15. A method for determining a surface normal vector for a vertex according to claim 12 , wherein the step of estimating a candidate tangent vector T S in the S direction comprises:
(a) determining an estimate for T S , where
T S1 =P 10xyz P 00W −P 00xyz P 10w
if P 10xyz P 00w is not equal to P 00xyz P 10w ; or
(b) determining an estimate for T S , where
T S2 =P 20xyz P 00w P 00xyz P 20w
if P 20xyz P 00w is not equal to P 00xyz P 20w ; or
(c) determining an estimate for T S , where
T S3 =P 30xyz P 00W P 00xyz P 30w
if P 30xyz P 00w is not equal to P 00xyz P 30w .
16. A method for determining a surface normal vector for a vertex according to claim 12 , wherein the step of estimating a candidate tangent vector T T in the T direction comprises:
(a) determining an estimate for T T , where
T T1 =P 01xyz P 00w −P 00xyz P 01w
if P 01xyz P 00w is not equal to P 00xyz P 01w ; or
(b) determining an estimate for T T , where
T T2 =P 02xyz P 00w −P 00xyz P 02w
if P 02xyz P 00w is not equal to P 00xyz P 02w ; or
(c) determining an estimate for T T , where
T T3 =P 03xyz P 00w P 00xyz P 03w
if P 03xyz P 00w is not equal to P 00xyz P 03w .
17. A method for determining a surface normal vector for a vertex according to claim 12 , wherein the step of estimating a candidate tangent vector T D in the diagonal direction comprises determining the estimate T D =P 11xyz P 00w −P 00xyz P 11w .
18. A method of determining a surface normal for a vertex according to claim 9 , wherein when the derived surface normal is zero an alternative pair of tangent vectors is selected and an alternative surface normal is derived.
19. A method for determining surface normal vectors for “tessellation” vertices of a surface patch used in modelling of an object to be shaded in a 3-d graphic system, the method comprising the steps of:
deriving surface normals for four corner vertices of the patch, with the surface normal for each corner vertex being derived by the sub-steps of:
(a) deriving the locations of control points required for determining a surface normal for the vertex,
(b) deriving a set of three candidate tangent vectors at the vertex from the control point data,
(c) selecting two tangent vectors from the set of candidate tangent vectors, and
(d) deriving a surface normal from the selected tangent vectors;
generating additional control points required to derive surface normals for non-corner tessellation vertices by the sub-steps of:
(a) partially subdividing the patch in a first parameter dimension to derive a first subset of control points of a first sub-patch,
(b) partially subdividing the patch in a second parameter dimension to derive a second subset of control points of a second sub-patch, and
(c) performing a third partial subdivision to derive a third subset of control points from the first and second subsets of control points; and
using the first subset of control points to derive surface normals for two mid-point tessellation vertices, with the surface normal for each midpoint tessellation vertex being derived by the sub-steps of:
(a) deriving the locations of control points required for determining a surface normal for the vertex,
(b) deriving a set of three candidate tangent vectors at the vertex from the control point data,
(c) selecting two tangent vectors from the set of candidate tangent vectors, and
(d) deriving a surface normal from the selected tangent vectors;
using the second subset of control points to derive surface normals for the remaining two midpoint tessellation vertices, with the surface normal for each remaining mid-point tessellation vertex being derived by the sub-steps of:
(a) deriving the locations of control points required for determining a surface normal for the vertex,
(b) deriving a set of three candidate tangent vectors at the vertex from the control point data,
(c) selecting two tangent vectors from the set of candidate tangent vectors, and
(d) deriving a surface normal from the selected tangent vectors;
using the third subset of control points to derive a surface normal for a tessellation vertex at the centre of the patch, with the surface normal for the tessellation vertex at the center of the patch being derived by the sub-steps of:
(a) deriving the locations of control points required for determining a surface normal for the vertex,
(b) deriving a set of three candidate tangent vectors at the vertex from the control point data,
(c) selecting two tangent vectors from the set of candidate tangent vectors, and
(d) deriving a surface normal from the selected tangent vectors; and rasterizing the image for display.