III.5.2 Main differences to the image source method
Ray-tracing is fundamentally different than the image-source method when their abilities to find specular reflections paths are compared. The image-source method is guaranteed to find all the possible specular reflection paths whereas the ray-tracing performs a Monte Carlo sampling of the path space to find those paths. In practice this means that the results of these two methods will converge when number of rays is increased (as far as all the reflections are specular). If there are too few rays, it is highly likely that some of the specular paths will be missed by ray-tracing.

Another fundamental difference is that ray-tracing typically uses energies instead of the sound pressure. If the rays are used only for specular reflections, then even the ray-tracing can be conducted pressure-based, but the diffuse reflections make the difference here. The diffuse reflections don't have any relevant phase information and thus the rays need to be considered as carriers of energy.

Third main difference between these two approaches is related to the listener. In the image-source method, the listener can be thought to be a point, whereas in the ray-tracing the listener needs to have some volume. If it would be just a point, the probability of a random ray to intersect a point will be zero in practice and, thus, the listener would detect no rays. In addition, it is worth noting that the image-sources are invariant with respect to the listener location meaning that it is possible to pre-compute all the image-source locations and for each listener location perform only the path validation between the source and the listener. This is contrary to the typical practice in ray-tracing, where all rays need to be recomputed if the listener moves.