const float *vec, int width, int height,
int16_t us[4][4], int16_t vs[4][4], float *du, float *dv)
{
- const float x = vec[0] / (1.f - vec[1]) / s->iflat_range[0] * s->input_mirror_modifier[0];
- const float y = vec[2] / (1.f - vec[1]) / s->iflat_range[1] * s->input_mirror_modifier[1];
+ const float theta = acosf(-vec[2]);
+ const float r = tanf(theta * 0.5f);
+ const float c = r / hypotf(vec[0], vec[1]);
+ const float x = vec[0] * c / s->iflat_range[0] * s->input_mirror_modifier[0];
+ const float y = -vec[1] * c / s->iflat_range[1] * s->input_mirror_modifier[1];
const float uf = (x + 1.f) * width / 2.f;
const float vf = (y + 1.f) * height / 2.f;