A spherical drop that strikes a flat surface square-on (90°) leaves a round stain. The shallower the strike, the more the stain elongates into an ellipse, because the drop smears out along its direction of travel while its width stays close to the drop's own diameter. The relationship between the two is exact trigonometry:
sin(θ) = width / length
θ = arcsin(width / length)
The simulator flies each droplet on a real ballistic arc — constant horizontal velocity, gravity pulling the vertical component down — and measures its velocity vector at the instant it crosses the wall. The angle between that vector and the wall plane is θ, so the ellipse drawn on the wall is generated directly from the physics, not faked from a slider.
- Launch elevation — the vertical angle the droplet leaves the source at; combined with gravity and travel distance it sets the true impact angle, which is usually shallower than the launch angle by the time the drop reaches the wall.
- Droplet speed — faster drops fly flatter arcs and lose less angle to gravity before impact.
- Droplets per shot — each shot fires a small randomized spray from one source point, producing several stains to analyze at once — more stains, tighter convergence estimate.
- Back-projection — for every stain, a ray is cast from its center, angled up at θ and pointed opposite its elongation direction (the "tail" always points away from the source). Where those rays best agree in 3D is the area of convergence — a least-squares closest point to all the rays, exactly the geometric idea behind real crime-scene stringing.
Limitation shown deliberately: real BPA stringing uses straight lines and ignores gravity's curvature on the drop's actual path, which is why the computed convergence point drifts from the true source — the same systematic bias forensic analysts correct for with the "tangent method".