2017-08-17 03:46:04 +00:00
|
|
|
/* GStreamer
|
|
|
|
*
|
|
|
|
* Copyright (C) 2014 Matthew Waters <ystreet00@gmail.com>
|
|
|
|
*
|
|
|
|
* This library is free software; you can redistribute it and/or
|
|
|
|
* modify it under the terms of the GNU Library General Public
|
|
|
|
* License as published by the Free Software Foundation; either
|
|
|
|
* version 2 of the License, or (at your option) any later version.
|
|
|
|
*
|
|
|
|
* This library is distributed in the hope that it will be useful,
|
|
|
|
* but WITHOUT ANY WARRANTY; without even the implied warranty of
|
|
|
|
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
|
|
|
|
* Library General Public License for more details.
|
|
|
|
*
|
|
|
|
* You should have received a copy of the GNU Library General Public
|
|
|
|
* License along with this library; if not, write to the
|
|
|
|
* Free Software Foundation, Inc., 51 Franklin St, Fifth Floor,
|
|
|
|
* Boston, MA 02110-1301, USA.
|
|
|
|
*/
|
|
|
|
|
|
|
|
#ifdef HAVE_CONFIG_H
|
|
|
|
# include "config.h"
|
|
|
|
#endif
|
|
|
|
|
2022-03-29 15:51:13 +00:00
|
|
|
#include <gst/gl/gl.h>
|
2017-08-17 03:46:04 +00:00
|
|
|
#undef GST_CAT_DEFAULT
|
|
|
|
#include <gst/check/gstcheck.h>
|
|
|
|
|
gl/utils: Fix NDC conversion matrices for column-majorness
The matrices were converting the wrong values with non-diagonal-only matrices.
e.g. a typical yflip matrix in [-1,1]^3 such as
1 0 0 0
0 -1 0 0
0 0 1 0
0 0 0 1
Would have actually required a matrix like this in [0,1]^3
1 0 0 0
0 -1 0 0
0 0 1 0
0 -2 0 1
Which is
1. not consistent with our multiplication convention and would require
transposing matrices or changing our multiplication order (from what is
generally used on opengl matrix guides/tutorials).
2. Produces incorrect values when input with actual vertices accounting for
the difference in multiplication order. e.g. some vertices multiplied by
the yflip matrix using vertex * yflip(== transpose(yflip) * vertex):
vertex: -> result: expected:
vec4(1,0,1,1) -> vec4(1,-2,1,1) vec4(1,1,1,1)
vec4(1,1,1,1) -> vec4(1,-3,1,1) vec4(1,0,1,1)
With the updated values, we now get the expected values.
Includes a test for this behaviour and the example above
2018-05-23 16:49:54 +00:00
|
|
|
#define VEC4_FORMAT "10.4f %10.4f %10.4f %10.4f"
|
|
|
|
#define VEC4_ARGS(v) (v)[0], (v)[1], (v)[2], (v)[3]
|
2017-08-17 03:46:04 +00:00
|
|
|
#define EPSILON 0.0001f
|
|
|
|
#define FEQ(a,b) (fabs(a-b) < EPSILON)
|
|
|
|
|
|
|
|
static void
|
|
|
|
debug_matrix (const float *m)
|
|
|
|
{
|
|
|
|
int i;
|
|
|
|
for (i = 0; i < 4; i++) {
|
gl/utils: Fix NDC conversion matrices for column-majorness
The matrices were converting the wrong values with non-diagonal-only matrices.
e.g. a typical yflip matrix in [-1,1]^3 such as
1 0 0 0
0 -1 0 0
0 0 1 0
0 0 0 1
Would have actually required a matrix like this in [0,1]^3
1 0 0 0
0 -1 0 0
0 0 1 0
0 -2 0 1
Which is
1. not consistent with our multiplication convention and would require
transposing matrices or changing our multiplication order (from what is
generally used on opengl matrix guides/tutorials).
2. Produces incorrect values when input with actual vertices accounting for
the difference in multiplication order. e.g. some vertices multiplied by
the yflip matrix using vertex * yflip(== transpose(yflip) * vertex):
vertex: -> result: expected:
vec4(1,0,1,1) -> vec4(1,-2,1,1) vec4(1,1,1,1)
vec4(1,1,1,1) -> vec4(1,-3,1,1) vec4(1,0,1,1)
With the updated values, we now get the expected values.
Includes a test for this behaviour and the example above
2018-05-23 16:49:54 +00:00
|
|
|
GST_DEBUG ("%" VEC4_FORMAT, VEC4_ARGS (&m[i * 4]));
|
2017-08-17 03:46:04 +00:00
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
GST_START_TEST (test_matrix_multiply)
|
|
|
|
{
|
|
|
|
/* A * B == C */
|
|
|
|
const float A[] = {
|
|
|
|
1., 1., 2., 5.,
|
|
|
|
0., 3., 0., 1.,
|
|
|
|
2., 0., 3., 1.,
|
|
|
|
3., 2., 1., 0.,
|
|
|
|
};
|
|
|
|
|
|
|
|
const float B[] = {
|
|
|
|
3., 1., 0., 2.,
|
|
|
|
1., 0., 3., 2.,
|
|
|
|
0., 1., 2., 3.,
|
|
|
|
3., 2., 1., 0.,
|
|
|
|
};
|
|
|
|
|
|
|
|
const float C[] = {
|
|
|
|
19., 13., 12., 10.,
|
|
|
|
6., 2., 10., 6.,
|
|
|
|
9., 7., 7., 13.,
|
|
|
|
11., 4., 8., 13.,
|
|
|
|
};
|
|
|
|
|
|
|
|
float res[16];
|
|
|
|
int i;
|
|
|
|
|
|
|
|
gst_gl_multiply_matrix4 (A, B, res);
|
gl/utils: Fix NDC conversion matrices for column-majorness
The matrices were converting the wrong values with non-diagonal-only matrices.
e.g. a typical yflip matrix in [-1,1]^3 such as
1 0 0 0
0 -1 0 0
0 0 1 0
0 0 0 1
Would have actually required a matrix like this in [0,1]^3
1 0 0 0
0 -1 0 0
0 0 1 0
0 -2 0 1
Which is
1. not consistent with our multiplication convention and would require
transposing matrices or changing our multiplication order (from what is
generally used on opengl matrix guides/tutorials).
2. Produces incorrect values when input with actual vertices accounting for
the difference in multiplication order. e.g. some vertices multiplied by
the yflip matrix using vertex * yflip(== transpose(yflip) * vertex):
vertex: -> result: expected:
vec4(1,0,1,1) -> vec4(1,-2,1,1) vec4(1,1,1,1)
vec4(1,1,1,1) -> vec4(1,-3,1,1) vec4(1,0,1,1)
With the updated values, we now get the expected values.
Includes a test for this behaviour and the example above
2018-05-23 16:49:54 +00:00
|
|
|
GST_DEBUG ("Matrix A:");
|
|
|
|
debug_matrix (A);
|
|
|
|
GST_DEBUG ("Matrix B:");
|
|
|
|
debug_matrix (B);
|
|
|
|
GST_DEBUG ("Matrix C:");
|
|
|
|
debug_matrix (C);
|
|
|
|
GST_DEBUG ("Multiplication Result == C == A * B:");
|
2017-08-17 03:46:04 +00:00
|
|
|
debug_matrix (res);
|
|
|
|
|
|
|
|
for (i = 0; i < G_N_ELEMENTS (res); i++) {
|
|
|
|
fail_unless (FEQ (res[i], C[i]), "value %f at index %u does not match "
|
|
|
|
"expected value %f", res[i], i, C[i]);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
GST_END_TEST;
|
|
|
|
|
|
|
|
GST_START_TEST (test_matrix_ndc)
|
|
|
|
{
|
|
|
|
GstBuffer *buffer = gst_buffer_new ();
|
|
|
|
GstVideoAffineTransformationMeta *aff_meta;
|
|
|
|
float res[16];
|
|
|
|
int i;
|
|
|
|
|
|
|
|
const float m[] = {
|
|
|
|
1., 0., 0., 0.,
|
|
|
|
0., 1., 0., 0.,
|
|
|
|
0., 0., 1., 0.,
|
|
|
|
0., 0., 0., 1.,
|
|
|
|
};
|
|
|
|
|
|
|
|
const float n[] = {
|
|
|
|
4., 6., 4., 9.,
|
|
|
|
1., 5., 8., 2.,
|
|
|
|
9., 3., 5., 8.,
|
|
|
|
3., 7., 9., 1.,
|
|
|
|
};
|
|
|
|
|
|
|
|
aff_meta = gst_buffer_add_video_affine_transformation_meta (buffer);
|
|
|
|
|
|
|
|
/* test default identity matrix */
|
|
|
|
gst_gl_get_affine_transformation_meta_as_ndc (aff_meta, res);
|
gl/utils: Fix NDC conversion matrices for column-majorness
The matrices were converting the wrong values with non-diagonal-only matrices.
e.g. a typical yflip matrix in [-1,1]^3 such as
1 0 0 0
0 -1 0 0
0 0 1 0
0 0 0 1
Would have actually required a matrix like this in [0,1]^3
1 0 0 0
0 -1 0 0
0 0 1 0
0 -2 0 1
Which is
1. not consistent with our multiplication convention and would require
transposing matrices or changing our multiplication order (from what is
generally used on opengl matrix guides/tutorials).
2. Produces incorrect values when input with actual vertices accounting for
the difference in multiplication order. e.g. some vertices multiplied by
the yflip matrix using vertex * yflip(== transpose(yflip) * vertex):
vertex: -> result: expected:
vec4(1,0,1,1) -> vec4(1,-2,1,1) vec4(1,1,1,1)
vec4(1,1,1,1) -> vec4(1,-3,1,1) vec4(1,0,1,1)
With the updated values, we now get the expected values.
Includes a test for this behaviour and the example above
2018-05-23 16:49:54 +00:00
|
|
|
GST_DEBUG ("Default matrix in the affine meta:");
|
2017-08-17 03:46:04 +00:00
|
|
|
debug_matrix (res);
|
|
|
|
|
|
|
|
for (i = 0; i < G_N_ELEMENTS (res); i++) {
|
|
|
|
fail_unless (FEQ (res[i], m[i]), "value %f at index %u does not match "
|
|
|
|
"expected value %f", res[i], i, m[i]);
|
|
|
|
}
|
|
|
|
|
|
|
|
/* test setting and receiving the same values */
|
gl/utils: Fix NDC conversion matrices for column-majorness
The matrices were converting the wrong values with non-diagonal-only matrices.
e.g. a typical yflip matrix in [-1,1]^3 such as
1 0 0 0
0 -1 0 0
0 0 1 0
0 0 0 1
Would have actually required a matrix like this in [0,1]^3
1 0 0 0
0 -1 0 0
0 0 1 0
0 -2 0 1
Which is
1. not consistent with our multiplication convention and would require
transposing matrices or changing our multiplication order (from what is
generally used on opengl matrix guides/tutorials).
2. Produces incorrect values when input with actual vertices accounting for
the difference in multiplication order. e.g. some vertices multiplied by
the yflip matrix using vertex * yflip(== transpose(yflip) * vertex):
vertex: -> result: expected:
vec4(1,0,1,1) -> vec4(1,-2,1,1) vec4(1,1,1,1)
vec4(1,1,1,1) -> vec4(1,-3,1,1) vec4(1,0,1,1)
With the updated values, we now get the expected values.
Includes a test for this behaviour and the example above
2018-05-23 16:49:54 +00:00
|
|
|
GST_DEBUG ("Set matrix on the affine transformation meta:");
|
|
|
|
debug_matrix (n);
|
|
|
|
|
2017-08-17 03:46:04 +00:00
|
|
|
gst_gl_set_affine_transformation_meta_from_ndc (aff_meta, n);
|
|
|
|
gst_gl_get_affine_transformation_meta_as_ndc (aff_meta, res);
|
|
|
|
|
gl/utils: Fix NDC conversion matrices for column-majorness
The matrices were converting the wrong values with non-diagonal-only matrices.
e.g. a typical yflip matrix in [-1,1]^3 such as
1 0 0 0
0 -1 0 0
0 0 1 0
0 0 0 1
Would have actually required a matrix like this in [0,1]^3
1 0 0 0
0 -1 0 0
0 0 1 0
0 -2 0 1
Which is
1. not consistent with our multiplication convention and would require
transposing matrices or changing our multiplication order (from what is
generally used on opengl matrix guides/tutorials).
2. Produces incorrect values when input with actual vertices accounting for
the difference in multiplication order. e.g. some vertices multiplied by
the yflip matrix using vertex * yflip(== transpose(yflip) * vertex):
vertex: -> result: expected:
vec4(1,0,1,1) -> vec4(1,-2,1,1) vec4(1,1,1,1)
vec4(1,1,1,1) -> vec4(1,-3,1,1) vec4(1,0,1,1)
With the updated values, we now get the expected values.
Includes a test for this behaviour and the example above
2018-05-23 16:49:54 +00:00
|
|
|
GST_DEBUG ("Retrieve the matrix set on the affine meta:");
|
2017-08-17 03:46:04 +00:00
|
|
|
debug_matrix (res);
|
|
|
|
|
|
|
|
for (i = 0; i < G_N_ELEMENTS (res); i++) {
|
|
|
|
fail_unless (FEQ (res[i], n[i]), "value %f at index %u does not match "
|
|
|
|
"expected value %f", res[i], i, n[i]);
|
|
|
|
}
|
|
|
|
|
|
|
|
gst_buffer_unref (buffer);
|
|
|
|
}
|
|
|
|
|
|
|
|
GST_END_TEST;
|
2018-07-10 10:03:12 +00:00
|
|
|
|
2017-08-17 03:46:04 +00:00
|
|
|
static void
|
|
|
|
transpose_matrix4 (float *m, float *res)
|
|
|
|
{
|
|
|
|
int i, j;
|
|
|
|
|
|
|
|
for (i = 0; i < 4; i++) {
|
|
|
|
for (j = 0; j < 4; j++) {
|
|
|
|
int idx = i + (j * 4);
|
|
|
|
int swapped_idx = j + (i * 4);
|
|
|
|
|
2018-07-10 10:03:12 +00:00
|
|
|
GST_ERROR ("swapping %i %f into %i", idx, m[idx], swapped_idx);
|
|
|
|
|
2017-08-17 03:46:04 +00:00
|
|
|
if (i == j)
|
|
|
|
fail_unless (idx == swapped_idx);
|
|
|
|
|
|
|
|
res[swapped_idx] = m[idx];
|
|
|
|
}
|
|
|
|
}
|
|
|
|
}
|
2018-07-10 10:03:12 +00:00
|
|
|
|
2017-08-17 03:46:04 +00:00
|
|
|
static float
|
|
|
|
dot4 (float *v1, float *v2)
|
|
|
|
{
|
|
|
|
GST_TRACE ("%.4f * %.4f + %.4f * %.4f + %.4f * %.4f + %.4f * %.4f",
|
|
|
|
v1[0], v2[0], v1[1], v2[1], v1[2], v2[2], v1[3], v2[3]);
|
|
|
|
return v1[0] * v2[0] + v1[1] * v2[1] + v1[2] * v2[2] + v1[3] * v2[3];
|
|
|
|
}
|
|
|
|
|
|
|
|
/* m * v */
|
|
|
|
static void
|
|
|
|
_matrix_mult_vertex4 (float *m, float *v, float *res)
|
|
|
|
{
|
|
|
|
res[0] = dot4 (&m[0], v);
|
|
|
|
res[1] = dot4 (&m[4], v);
|
|
|
|
res[2] = dot4 (&m[8], v);
|
|
|
|
res[3] = dot4 (&m[12], v);
|
|
|
|
}
|
|
|
|
|
|
|
|
/* v * m */
|
2018-07-10 10:03:12 +00:00
|
|
|
/* Used because the default is for OpenGL to read matrices transposed on
|
|
|
|
* uploading */
|
2017-08-17 03:46:04 +00:00
|
|
|
static void
|
2018-07-10 10:03:12 +00:00
|
|
|
_vertex_mult_matrix4 (float *m, float *v, float *res)
|
2017-08-17 03:46:04 +00:00
|
|
|
{
|
|
|
|
float tmp[16] = { 0., };
|
|
|
|
|
2018-07-10 10:03:12 +00:00
|
|
|
GST_TRACE ("original matrix");
|
|
|
|
debug_matrix (m);
|
2017-08-17 03:46:04 +00:00
|
|
|
transpose_matrix4 (m, tmp);
|
gl/utils: Fix NDC conversion matrices for column-majorness
The matrices were converting the wrong values with non-diagonal-only matrices.
e.g. a typical yflip matrix in [-1,1]^3 such as
1 0 0 0
0 -1 0 0
0 0 1 0
0 0 0 1
Would have actually required a matrix like this in [0,1]^3
1 0 0 0
0 -1 0 0
0 0 1 0
0 -2 0 1
Which is
1. not consistent with our multiplication convention and would require
transposing matrices or changing our multiplication order (from what is
generally used on opengl matrix guides/tutorials).
2. Produces incorrect values when input with actual vertices accounting for
the difference in multiplication order. e.g. some vertices multiplied by
the yflip matrix using vertex * yflip(== transpose(yflip) * vertex):
vertex: -> result: expected:
vec4(1,0,1,1) -> vec4(1,-2,1,1) vec4(1,1,1,1)
vec4(1,1,1,1) -> vec4(1,-3,1,1) vec4(1,0,1,1)
With the updated values, we now get the expected values.
Includes a test for this behaviour and the example above
2018-05-23 16:49:54 +00:00
|
|
|
GST_TRACE ("transposed matrix");
|
|
|
|
debug_matrix (tmp);
|
2017-08-17 03:46:04 +00:00
|
|
|
_matrix_mult_vertex4 (tmp, v, res);
|
|
|
|
}
|
2018-07-10 10:03:12 +00:00
|
|
|
|
2017-08-17 03:46:04 +00:00
|
|
|
GST_START_TEST (test_matrix_vertex_identity)
|
|
|
|
{
|
|
|
|
float identity[] = {
|
|
|
|
1., 0., 0., 0.,
|
|
|
|
0., 1., 0., 0.,
|
|
|
|
0., 0., 1., 0.,
|
|
|
|
0., 0., 0., 1.,
|
|
|
|
};
|
|
|
|
|
|
|
|
float v[] = { 1., 1., 1., 1. };
|
|
|
|
float res[4] = { 0., };
|
|
|
|
int i;
|
|
|
|
|
2018-07-10 10:03:12 +00:00
|
|
|
_vertex_mult_matrix4 (identity, v, res);
|
gl/utils: Fix NDC conversion matrices for column-majorness
The matrices were converting the wrong values with non-diagonal-only matrices.
e.g. a typical yflip matrix in [-1,1]^3 such as
1 0 0 0
0 -1 0 0
0 0 1 0
0 0 0 1
Would have actually required a matrix like this in [0,1]^3
1 0 0 0
0 -1 0 0
0 0 1 0
0 -2 0 1
Which is
1. not consistent with our multiplication convention and would require
transposing matrices or changing our multiplication order (from what is
generally used on opengl matrix guides/tutorials).
2. Produces incorrect values when input with actual vertices accounting for
the difference in multiplication order. e.g. some vertices multiplied by
the yflip matrix using vertex * yflip(== transpose(yflip) * vertex):
vertex: -> result: expected:
vec4(1,0,1,1) -> vec4(1,-2,1,1) vec4(1,1,1,1)
vec4(1,1,1,1) -> vec4(1,-3,1,1) vec4(1,0,1,1)
With the updated values, we now get the expected values.
Includes a test for this behaviour and the example above
2018-05-23 16:49:54 +00:00
|
|
|
GST_DEBUG ("vertex: %" VEC4_FORMAT, VEC4_ARGS (v));
|
|
|
|
GST_DEBUG ("result: %" VEC4_FORMAT, VEC4_ARGS (res));
|
2017-08-17 03:46:04 +00:00
|
|
|
|
|
|
|
for (i = 0; i < 4; i++) {
|
|
|
|
fail_unless (FEQ (res[i], v[i]), "value %f at index %u does not match "
|
|
|
|
"expected value %f", res[i], i, v[i]);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
GST_END_TEST;
|
|
|
|
|
|
|
|
GST_START_TEST (test_matrix_vertex_scale)
|
|
|
|
{
|
|
|
|
float scale[] = {
|
|
|
|
1.5, 0., 0., 0.,
|
|
|
|
0., 2.5, 0., 0.,
|
|
|
|
0., 0., 3., 0.,
|
|
|
|
0., 0., 0., 1.,
|
|
|
|
};
|
|
|
|
|
|
|
|
float v[] = { 1., 1., 1., 1. };
|
|
|
|
float expected[] = { 1.5, 2.5, 3., 1. };
|
|
|
|
float res[4] = { 0., };
|
|
|
|
int i;
|
|
|
|
|
2018-07-10 10:03:12 +00:00
|
|
|
_vertex_mult_matrix4 (scale, v, res);
|
gl/utils: Fix NDC conversion matrices for column-majorness
The matrices were converting the wrong values with non-diagonal-only matrices.
e.g. a typical yflip matrix in [-1,1]^3 such as
1 0 0 0
0 -1 0 0
0 0 1 0
0 0 0 1
Would have actually required a matrix like this in [0,1]^3
1 0 0 0
0 -1 0 0
0 0 1 0
0 -2 0 1
Which is
1. not consistent with our multiplication convention and would require
transposing matrices or changing our multiplication order (from what is
generally used on opengl matrix guides/tutorials).
2. Produces incorrect values when input with actual vertices accounting for
the difference in multiplication order. e.g. some vertices multiplied by
the yflip matrix using vertex * yflip(== transpose(yflip) * vertex):
vertex: -> result: expected:
vec4(1,0,1,1) -> vec4(1,-2,1,1) vec4(1,1,1,1)
vec4(1,1,1,1) -> vec4(1,-3,1,1) vec4(1,0,1,1)
With the updated values, we now get the expected values.
Includes a test for this behaviour and the example above
2018-05-23 16:49:54 +00:00
|
|
|
GST_DEBUG ("vertex: %" VEC4_FORMAT, VEC4_ARGS (v));
|
|
|
|
GST_DEBUG ("result: %" VEC4_FORMAT, VEC4_ARGS (res));
|
2017-08-17 03:46:04 +00:00
|
|
|
|
|
|
|
for (i = 0; i < 4; i++) {
|
|
|
|
fail_unless (FEQ (res[i], expected[i]),
|
|
|
|
"value %f at index %u does not match " "expected value %f", res[i], i,
|
|
|
|
expected[i]);
|
|
|
|
}
|
gl/utils: Fix NDC conversion matrices for column-majorness
The matrices were converting the wrong values with non-diagonal-only matrices.
e.g. a typical yflip matrix in [-1,1]^3 such as
1 0 0 0
0 -1 0 0
0 0 1 0
0 0 0 1
Would have actually required a matrix like this in [0,1]^3
1 0 0 0
0 -1 0 0
0 0 1 0
0 -2 0 1
Which is
1. not consistent with our multiplication convention and would require
transposing matrices or changing our multiplication order (from what is
generally used on opengl matrix guides/tutorials).
2. Produces incorrect values when input with actual vertices accounting for
the difference in multiplication order. e.g. some vertices multiplied by
the yflip matrix using vertex * yflip(== transpose(yflip) * vertex):
vertex: -> result: expected:
vec4(1,0,1,1) -> vec4(1,-2,1,1) vec4(1,1,1,1)
vec4(1,1,1,1) -> vec4(1,-3,1,1) vec4(1,0,1,1)
With the updated values, we now get the expected values.
Includes a test for this behaviour and the example above
2018-05-23 16:49:54 +00:00
|
|
|
}
|
2017-08-17 03:46:04 +00:00
|
|
|
|
gl/utils: Fix NDC conversion matrices for column-majorness
The matrices were converting the wrong values with non-diagonal-only matrices.
e.g. a typical yflip matrix in [-1,1]^3 such as
1 0 0 0
0 -1 0 0
0 0 1 0
0 0 0 1
Would have actually required a matrix like this in [0,1]^3
1 0 0 0
0 -1 0 0
0 0 1 0
0 -2 0 1
Which is
1. not consistent with our multiplication convention and would require
transposing matrices or changing our multiplication order (from what is
generally used on opengl matrix guides/tutorials).
2. Produces incorrect values when input with actual vertices accounting for
the difference in multiplication order. e.g. some vertices multiplied by
the yflip matrix using vertex * yflip(== transpose(yflip) * vertex):
vertex: -> result: expected:
vec4(1,0,1,1) -> vec4(1,-2,1,1) vec4(1,1,1,1)
vec4(1,1,1,1) -> vec4(1,-3,1,1) vec4(1,0,1,1)
With the updated values, we now get the expected values.
Includes a test for this behaviour and the example above
2018-05-23 16:49:54 +00:00
|
|
|
GST_END_TEST;
|
|
|
|
|
2018-07-10 10:03:12 +00:00
|
|
|
GST_START_TEST (test_matrix_vertex_translate)
|
|
|
|
{
|
|
|
|
float translate_1[] = {
|
|
|
|
1., 0., 0., 0.,
|
|
|
|
0., 1., 0., 0.,
|
|
|
|
0., 0., 1., 0.,
|
|
|
|
1., 2., 3., 1.,
|
|
|
|
};
|
|
|
|
|
|
|
|
float v[] = { 1., 1., 1., 1. };
|
|
|
|
float expected[] = { 2., 3., 4., 1. };
|
|
|
|
float res[4] = { 0., };
|
|
|
|
int i;
|
|
|
|
|
|
|
|
_vertex_mult_matrix4 (translate_1, v, res);
|
|
|
|
|
|
|
|
for (i = 0; i < 4; i++) {
|
|
|
|
fail_unless (FEQ (res[i], expected[i]),
|
|
|
|
"value %f at index %u does not match " "expected value %f", res[i], i,
|
|
|
|
expected[i]);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
GST_END_TEST;
|
|
|
|
|
gl/utils: Fix NDC conversion matrices for column-majorness
The matrices were converting the wrong values with non-diagonal-only matrices.
e.g. a typical yflip matrix in [-1,1]^3 such as
1 0 0 0
0 -1 0 0
0 0 1 0
0 0 0 1
Would have actually required a matrix like this in [0,1]^3
1 0 0 0
0 -1 0 0
0 0 1 0
0 -2 0 1
Which is
1. not consistent with our multiplication convention and would require
transposing matrices or changing our multiplication order (from what is
generally used on opengl matrix guides/tutorials).
2. Produces incorrect values when input with actual vertices accounting for
the difference in multiplication order. e.g. some vertices multiplied by
the yflip matrix using vertex * yflip(== transpose(yflip) * vertex):
vertex: -> result: expected:
vec4(1,0,1,1) -> vec4(1,-2,1,1) vec4(1,1,1,1)
vec4(1,1,1,1) -> vec4(1,-3,1,1) vec4(1,0,1,1)
With the updated values, we now get the expected values.
Includes a test for this behaviour and the example above
2018-05-23 16:49:54 +00:00
|
|
|
GST_START_TEST (test_matrix_vertex_y_invert)
|
|
|
|
{
|
2018-06-21 07:47:09 +00:00
|
|
|
GstBuffer *buffer = gst_buffer_new ();
|
|
|
|
GstVideoAffineTransformationMeta *aff_meta;
|
|
|
|
|
gl/utils: Fix NDC conversion matrices for column-majorness
The matrices were converting the wrong values with non-diagonal-only matrices.
e.g. a typical yflip matrix in [-1,1]^3 such as
1 0 0 0
0 -1 0 0
0 0 1 0
0 0 0 1
Would have actually required a matrix like this in [0,1]^3
1 0 0 0
0 -1 0 0
0 0 1 0
0 -2 0 1
Which is
1. not consistent with our multiplication convention and would require
transposing matrices or changing our multiplication order (from what is
generally used on opengl matrix guides/tutorials).
2. Produces incorrect values when input with actual vertices accounting for
the difference in multiplication order. e.g. some vertices multiplied by
the yflip matrix using vertex * yflip(== transpose(yflip) * vertex):
vertex: -> result: expected:
vec4(1,0,1,1) -> vec4(1,-2,1,1) vec4(1,1,1,1)
vec4(1,1,1,1) -> vec4(1,-3,1,1) vec4(1,0,1,1)
With the updated values, we now get the expected values.
Includes a test for this behaviour and the example above
2018-05-23 16:49:54 +00:00
|
|
|
float y_invert[] = {
|
|
|
|
1., 0., 0., 0.,
|
|
|
|
0., -1., 0., 0.,
|
|
|
|
0., 0., 1., 0.,
|
|
|
|
0., 0., 0., 1.,
|
|
|
|
};
|
|
|
|
|
|
|
|
float v[] = { 1., 1., 1., 1. };
|
|
|
|
float expected[] = { 1., -1., 1., 1. };
|
|
|
|
float res[4] = { 0., };
|
|
|
|
int i;
|
|
|
|
|
|
|
|
/* The y_invert matrix but with a coordinate space of [0, 1]^3 instead
|
|
|
|
* of [-1, 1]^3 */
|
2018-06-21 07:47:09 +00:00
|
|
|
|
|
|
|
aff_meta = gst_buffer_add_video_affine_transformation_meta (buffer);
|
gl/utils: Fix NDC conversion matrices for column-majorness
The matrices were converting the wrong values with non-diagonal-only matrices.
e.g. a typical yflip matrix in [-1,1]^3 such as
1 0 0 0
0 -1 0 0
0 0 1 0
0 0 0 1
Would have actually required a matrix like this in [0,1]^3
1 0 0 0
0 -1 0 0
0 0 1 0
0 -2 0 1
Which is
1. not consistent with our multiplication convention and would require
transposing matrices or changing our multiplication order (from what is
generally used on opengl matrix guides/tutorials).
2. Produces incorrect values when input with actual vertices accounting for
the difference in multiplication order. e.g. some vertices multiplied by
the yflip matrix using vertex * yflip(== transpose(yflip) * vertex):
vertex: -> result: expected:
vec4(1,0,1,1) -> vec4(1,-2,1,1) vec4(1,1,1,1)
vec4(1,1,1,1) -> vec4(1,-3,1,1) vec4(1,0,1,1)
With the updated values, we now get the expected values.
Includes a test for this behaviour and the example above
2018-05-23 16:49:54 +00:00
|
|
|
|
|
|
|
GST_DEBUG ("y-invert");
|
|
|
|
debug_matrix (y_invert);
|
|
|
|
|
2018-07-10 10:03:12 +00:00
|
|
|
_vertex_mult_matrix4 (y_invert, v, res);
|
gl/utils: Fix NDC conversion matrices for column-majorness
The matrices were converting the wrong values with non-diagonal-only matrices.
e.g. a typical yflip matrix in [-1,1]^3 such as
1 0 0 0
0 -1 0 0
0 0 1 0
0 0 0 1
Would have actually required a matrix like this in [0,1]^3
1 0 0 0
0 -1 0 0
0 0 1 0
0 -2 0 1
Which is
1. not consistent with our multiplication convention and would require
transposing matrices or changing our multiplication order (from what is
generally used on opengl matrix guides/tutorials).
2. Produces incorrect values when input with actual vertices accounting for
the difference in multiplication order. e.g. some vertices multiplied by
the yflip matrix using vertex * yflip(== transpose(yflip) * vertex):
vertex: -> result: expected:
vec4(1,0,1,1) -> vec4(1,-2,1,1) vec4(1,1,1,1)
vec4(1,1,1,1) -> vec4(1,-3,1,1) vec4(1,0,1,1)
With the updated values, we now get the expected values.
Includes a test for this behaviour and the example above
2018-05-23 16:49:54 +00:00
|
|
|
GST_DEBUG ("vertex: %" VEC4_FORMAT, VEC4_ARGS (v));
|
|
|
|
GST_DEBUG ("result: %" VEC4_FORMAT, VEC4_ARGS (res));
|
|
|
|
|
|
|
|
for (i = 0; i < 4; i++) {
|
|
|
|
fail_unless (FEQ (res[i], expected[i]),
|
|
|
|
"value %f at index %u does not match " "expected value %f", res[i], i,
|
|
|
|
expected[i]);
|
|
|
|
}
|
|
|
|
|
2019-08-29 17:42:39 +00:00
|
|
|
/* now test the [0, 1]^3 matrix and update the test values accordingly */
|
2018-06-21 07:47:09 +00:00
|
|
|
gst_gl_set_affine_transformation_meta_from_ndc (aff_meta, y_invert);
|
gl/utils: Fix NDC conversion matrices for column-majorness
The matrices were converting the wrong values with non-diagonal-only matrices.
e.g. a typical yflip matrix in [-1,1]^3 such as
1 0 0 0
0 -1 0 0
0 0 1 0
0 0 0 1
Would have actually required a matrix like this in [0,1]^3
1 0 0 0
0 -1 0 0
0 0 1 0
0 -2 0 1
Which is
1. not consistent with our multiplication convention and would require
transposing matrices or changing our multiplication order (from what is
generally used on opengl matrix guides/tutorials).
2. Produces incorrect values when input with actual vertices accounting for
the difference in multiplication order. e.g. some vertices multiplied by
the yflip matrix using vertex * yflip(== transpose(yflip) * vertex):
vertex: -> result: expected:
vec4(1,0,1,1) -> vec4(1,-2,1,1) vec4(1,1,1,1)
vec4(1,1,1,1) -> vec4(1,-3,1,1) vec4(1,0,1,1)
With the updated values, we now get the expected values.
Includes a test for this behaviour and the example above
2018-05-23 16:49:54 +00:00
|
|
|
expected[1] = 0.;
|
|
|
|
|
|
|
|
GST_DEBUG ("y-invert from ndc [-1,1]^3 to [0,1]^3");
|
2018-06-21 07:47:09 +00:00
|
|
|
debug_matrix (aff_meta->matrix);
|
gl/utils: Fix NDC conversion matrices for column-majorness
The matrices were converting the wrong values with non-diagonal-only matrices.
e.g. a typical yflip matrix in [-1,1]^3 such as
1 0 0 0
0 -1 0 0
0 0 1 0
0 0 0 1
Would have actually required a matrix like this in [0,1]^3
1 0 0 0
0 -1 0 0
0 0 1 0
0 -2 0 1
Which is
1. not consistent with our multiplication convention and would require
transposing matrices or changing our multiplication order (from what is
generally used on opengl matrix guides/tutorials).
2. Produces incorrect values when input with actual vertices accounting for
the difference in multiplication order. e.g. some vertices multiplied by
the yflip matrix using vertex * yflip(== transpose(yflip) * vertex):
vertex: -> result: expected:
vec4(1,0,1,1) -> vec4(1,-2,1,1) vec4(1,1,1,1)
vec4(1,1,1,1) -> vec4(1,-3,1,1) vec4(1,0,1,1)
With the updated values, we now get the expected values.
Includes a test for this behaviour and the example above
2018-05-23 16:49:54 +00:00
|
|
|
|
2018-07-10 10:03:12 +00:00
|
|
|
_vertex_mult_matrix4 (aff_meta->matrix, v, res);
|
gl/utils: Fix NDC conversion matrices for column-majorness
The matrices were converting the wrong values with non-diagonal-only matrices.
e.g. a typical yflip matrix in [-1,1]^3 such as
1 0 0 0
0 -1 0 0
0 0 1 0
0 0 0 1
Would have actually required a matrix like this in [0,1]^3
1 0 0 0
0 -1 0 0
0 0 1 0
0 -2 0 1
Which is
1. not consistent with our multiplication convention and would require
transposing matrices or changing our multiplication order (from what is
generally used on opengl matrix guides/tutorials).
2. Produces incorrect values when input with actual vertices accounting for
the difference in multiplication order. e.g. some vertices multiplied by
the yflip matrix using vertex * yflip(== transpose(yflip) * vertex):
vertex: -> result: expected:
vec4(1,0,1,1) -> vec4(1,-2,1,1) vec4(1,1,1,1)
vec4(1,1,1,1) -> vec4(1,-3,1,1) vec4(1,0,1,1)
With the updated values, we now get the expected values.
Includes a test for this behaviour and the example above
2018-05-23 16:49:54 +00:00
|
|
|
GST_DEBUG ("vertex: %" VEC4_FORMAT, VEC4_ARGS (v));
|
|
|
|
GST_DEBUG ("result: %" VEC4_FORMAT, VEC4_ARGS (res));
|
|
|
|
|
|
|
|
for (i = 0; i < 4; i++) {
|
|
|
|
fail_unless (FEQ (res[i], expected[i]),
|
|
|
|
"value %f at index %u does not match " "expected value %f", res[i], i,
|
|
|
|
expected[i]);
|
|
|
|
}
|
|
|
|
|
|
|
|
/* test vec4(1,0,1,1) -> vec4(1,1,1,1) */
|
|
|
|
v[1] = 0.;
|
|
|
|
expected[1] = 1.;
|
2018-07-10 10:03:12 +00:00
|
|
|
_vertex_mult_matrix4 (aff_meta->matrix, v, res);
|
gl/utils: Fix NDC conversion matrices for column-majorness
The matrices were converting the wrong values with non-diagonal-only matrices.
e.g. a typical yflip matrix in [-1,1]^3 such as
1 0 0 0
0 -1 0 0
0 0 1 0
0 0 0 1
Would have actually required a matrix like this in [0,1]^3
1 0 0 0
0 -1 0 0
0 0 1 0
0 -2 0 1
Which is
1. not consistent with our multiplication convention and would require
transposing matrices or changing our multiplication order (from what is
generally used on opengl matrix guides/tutorials).
2. Produces incorrect values when input with actual vertices accounting for
the difference in multiplication order. e.g. some vertices multiplied by
the yflip matrix using vertex * yflip(== transpose(yflip) * vertex):
vertex: -> result: expected:
vec4(1,0,1,1) -> vec4(1,-2,1,1) vec4(1,1,1,1)
vec4(1,1,1,1) -> vec4(1,-3,1,1) vec4(1,0,1,1)
With the updated values, we now get the expected values.
Includes a test for this behaviour and the example above
2018-05-23 16:49:54 +00:00
|
|
|
GST_DEBUG ("vertex: %" VEC4_FORMAT, VEC4_ARGS (v));
|
|
|
|
GST_DEBUG ("result: %" VEC4_FORMAT, VEC4_ARGS (res));
|
2017-08-17 03:46:04 +00:00
|
|
|
|
|
|
|
for (i = 0; i < 4; i++) {
|
|
|
|
fail_unless (FEQ (res[i], expected[i]),
|
|
|
|
"value %f at index %u does not match " "expected value %f", res[i], i,
|
|
|
|
expected[i]);
|
|
|
|
}
|
2019-03-06 09:23:22 +00:00
|
|
|
|
|
|
|
gst_buffer_unref (buffer);
|
2017-08-17 03:46:04 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
GST_END_TEST;
|
|
|
|
|
|
|
|
static Suite *
|
gl/utils: Fix NDC conversion matrices for column-majorness
The matrices were converting the wrong values with non-diagonal-only matrices.
e.g. a typical yflip matrix in [-1,1]^3 such as
1 0 0 0
0 -1 0 0
0 0 1 0
0 0 0 1
Would have actually required a matrix like this in [0,1]^3
1 0 0 0
0 -1 0 0
0 0 1 0
0 -2 0 1
Which is
1. not consistent with our multiplication convention and would require
transposing matrices or changing our multiplication order (from what is
generally used on opengl matrix guides/tutorials).
2. Produces incorrect values when input with actual vertices accounting for
the difference in multiplication order. e.g. some vertices multiplied by
the yflip matrix using vertex * yflip(== transpose(yflip) * vertex):
vertex: -> result: expected:
vec4(1,0,1,1) -> vec4(1,-2,1,1) vec4(1,1,1,1)
vec4(1,1,1,1) -> vec4(1,-3,1,1) vec4(1,0,1,1)
With the updated values, we now get the expected values.
Includes a test for this behaviour and the example above
2018-05-23 16:49:54 +00:00
|
|
|
gst_gl_matrix_suite (void)
|
2017-08-17 03:46:04 +00:00
|
|
|
{
|
|
|
|
Suite *s = suite_create ("GstGLMatrix");
|
|
|
|
TCase *tc_chain = tcase_create ("matrix");
|
|
|
|
|
|
|
|
suite_add_tcase (s, tc_chain);
|
|
|
|
tcase_add_test (tc_chain, test_matrix_multiply);
|
|
|
|
tcase_add_test (tc_chain, test_matrix_ndc);
|
gl/utils: Fix NDC conversion matrices for column-majorness
The matrices were converting the wrong values with non-diagonal-only matrices.
e.g. a typical yflip matrix in [-1,1]^3 such as
1 0 0 0
0 -1 0 0
0 0 1 0
0 0 0 1
Would have actually required a matrix like this in [0,1]^3
1 0 0 0
0 -1 0 0
0 0 1 0
0 -2 0 1
Which is
1. not consistent with our multiplication convention and would require
transposing matrices or changing our multiplication order (from what is
generally used on opengl matrix guides/tutorials).
2. Produces incorrect values when input with actual vertices accounting for
the difference in multiplication order. e.g. some vertices multiplied by
the yflip matrix using vertex * yflip(== transpose(yflip) * vertex):
vertex: -> result: expected:
vec4(1,0,1,1) -> vec4(1,-2,1,1) vec4(1,1,1,1)
vec4(1,1,1,1) -> vec4(1,-3,1,1) vec4(1,0,1,1)
With the updated values, we now get the expected values.
Includes a test for this behaviour and the example above
2018-05-23 16:49:54 +00:00
|
|
|
tcase_add_test (tc_chain, test_matrix_vertex_identity);
|
|
|
|
tcase_add_test (tc_chain, test_matrix_vertex_scale);
|
2018-07-10 10:03:12 +00:00
|
|
|
tcase_add_test (tc_chain, test_matrix_vertex_translate);
|
gl/utils: Fix NDC conversion matrices for column-majorness
The matrices were converting the wrong values with non-diagonal-only matrices.
e.g. a typical yflip matrix in [-1,1]^3 such as
1 0 0 0
0 -1 0 0
0 0 1 0
0 0 0 1
Would have actually required a matrix like this in [0,1]^3
1 0 0 0
0 -1 0 0
0 0 1 0
0 -2 0 1
Which is
1. not consistent with our multiplication convention and would require
transposing matrices or changing our multiplication order (from what is
generally used on opengl matrix guides/tutorials).
2. Produces incorrect values when input with actual vertices accounting for
the difference in multiplication order. e.g. some vertices multiplied by
the yflip matrix using vertex * yflip(== transpose(yflip) * vertex):
vertex: -> result: expected:
vec4(1,0,1,1) -> vec4(1,-2,1,1) vec4(1,1,1,1)
vec4(1,1,1,1) -> vec4(1,-3,1,1) vec4(1,0,1,1)
With the updated values, we now get the expected values.
Includes a test for this behaviour and the example above
2018-05-23 16:49:54 +00:00
|
|
|
tcase_add_test (tc_chain, test_matrix_vertex_y_invert);
|
2017-08-17 03:46:04 +00:00
|
|
|
|
|
|
|
return s;
|
|
|
|
}
|
|
|
|
|
gl/utils: Fix NDC conversion matrices for column-majorness
The matrices were converting the wrong values with non-diagonal-only matrices.
e.g. a typical yflip matrix in [-1,1]^3 such as
1 0 0 0
0 -1 0 0
0 0 1 0
0 0 0 1
Would have actually required a matrix like this in [0,1]^3
1 0 0 0
0 -1 0 0
0 0 1 0
0 -2 0 1
Which is
1. not consistent with our multiplication convention and would require
transposing matrices or changing our multiplication order (from what is
generally used on opengl matrix guides/tutorials).
2. Produces incorrect values when input with actual vertices accounting for
the difference in multiplication order. e.g. some vertices multiplied by
the yflip matrix using vertex * yflip(== transpose(yflip) * vertex):
vertex: -> result: expected:
vec4(1,0,1,1) -> vec4(1,-2,1,1) vec4(1,1,1,1)
vec4(1,1,1,1) -> vec4(1,-3,1,1) vec4(1,0,1,1)
With the updated values, we now get the expected values.
Includes a test for this behaviour and the example above
2018-05-23 16:49:54 +00:00
|
|
|
GST_CHECK_MAIN (gst_gl_matrix);
|