Does sound bounce in all directions, or as a ray, like light? When light is shined on a perfectly smooth surface it reflects from it as a ray, going in only one direction.
However, from my reading of Rayleigh's Theory of Sound he describes sound waves as emanating in "spheres". Does this mean that when sound bounces off a smooth surface, the energy scatters equally in all directions?
 A: Light only reflects as a "ray" if the incoming light itself is collimated a ray, like a laser pointer. Otherwise a more correct statement would be that the reflected angle is the same as the incident angle as opposed to lambertian reflection. 
Sound i am not so sure about, since pressure waves actually affect the medium in which it travels. Light doesn't do this in the same way as pressure waves so i'd say that the spherical approach is the correct one. Of course, the "beam" formed by an accoustic element could surely be altered with the help of some form of acoustic antenna. I'm sure some form of plate right behind acoustic element surely could disturb the isotropical behaviour.
A: Both sound and light are wave phenomena and tend to spread spherically. The crucial difference is the wavelength. Typical sound wavelengths range between  8 meters (40 Hz)  and 8 cm (4000 Hz). Visible light wavelengths range from 400 nm to 700 nm. The phenomenon of diffraction, essentially the tendency to spread in all directions, plays a much larger role for sound than for light. Also most surfaces are reflective for sound, while they are diffusive for light. As a consequence vision requires a line of sight while sound does not. 
A: I think your asking about specular reflection:
https://www.azooptics.com/Article.aspx?ArticleID=822
"Specular reflection is a type of surface reflectance often described as a mirror-like reflection of light from the surface. In specular reflection, the incident light is reflected into a single outgoing direction. The name specular is derived from the Latin word speculum, meaning mirror."
For more, google specular reflection.
So when the surface roughness is much smaller that a wavelength specular reflection will occur, otherwise there will be diffuse reflection.
https://en.wikipedia.org/wiki/Diffuse_reflection
"Diffuse reflection is the reflection of light or other waves or particles from a surface such that a ray incident on the surface is scattered at many angles rather than at just one angle as in the case of specular reflection."
Again for more, google diffuse reflection.
Note-this answer applies to sound or light.
Re. "When light is shined on a perfectly smooth surface it reflects from it as a ray, going in only one direction."  The same can happen to sound--it is called specular reflection.
So to answer your question:
"Does this mean that when sound bounces off a smooth surface, the energy scatters equally in all directions?".
The answer is no because of specular reflection.
A: Sound bounces as a superposition of rays, accounting for the relative phases between the rays.
You've probably seen something like this, after tossing a stone into a still pond:

You can obtain the shape of the reflected wavefront by thinking of rays emanating perpendicularly from an incoming wavefront which all get reflected like light rays in a mirror.
To get the amplitude correct, hence the energy flux as a function of angle, you can space these incoming rays so that the density of sources on the incoming wavefront approximates its amplitude (we do this at a peak of the wave). If this is true at one time, it will be true at all times because of the wave equation. Therefore, the amplitude of the reflected wave will also be the density of rays, times an order 1 factor $\sin \phi$ where $\phi$ accounts for the phase.
So the energy does not have to dissipate equally in all directions, because the different parts of the wave front collide at the wall with different amplitudes. The Huygens-Fresnel principle says that we can think of this distribution of incident amplitudes on the wall as a distribution of point sources whose loudness is precisely the incident amplitude. Each of these radiate equally in all directions, but they may be distributed in an interesting way. For a point source, they will always bounce with most energy back at the point source.
