Since I do have a rather extensive background in physics, shock wave production and supersonic fluid dynamics, I will try my best to explain how the whip crack occurs.

Summary of the production of the 'crack':

The initial velocity of the whip multiplied by its mass has initial momentum. As motion continues down the length of the whip momentum is conserved (minus some frictional losses). Near the tip the velocity is greatly increased to conserve momentum. The popper carries a packet of air with it due to the no-slip boundary condition. If the speed of the popper is sufficient it will create a 'whizz' sound which is the development of a Kármán vortex street. If the popper suddenly changes momentum, the packet of air that the popper carries with it will create a separation zone. If the change is fast enough, a separation bubble of low pressure occurs. When the air around this separation bubble pushes on the separation bubble, the bubble will collapse and the impact of air molecules will dissipate it's momentum in a wave.

If the separation velocity of the air packet, or inertial fluid drag pocket, and popper exceed the speed of sound, a cavitation bubble occurs between the two because the air molecules cannot move fast enough to fill in the space between them. When a cavitation bubble collapses a sharp 'boom' is heard because there will be a sharp increase in air pressure upon collapse (this pressure jump is a step function). That step function in air pressure dissipates out wards and is the sound of the crack when it hits the ear drum. When the popper rotates while creating the cavitation bubble, as in looping cracks, this generates vorticity around the bubble which aids in its growth, and produces and even louder crack. Think about the air around the cavitation bubble spinning like on a merry-go-round and the centrifugal reaction carries the cavitation boundary outwards.

The more fluffy the popper the bigger the cavitation bubble if the velocity is the same. A fluffier popper also causes more drag and slows down the end velocity.

Even if you do not break the sound barrier, a small separation bubble can collapse fast enough to sound really close to a crack.

The larger the relative separation velocity of the popper and the inertial fluid drag pocket the larger the cavitation bubble.

Why certain whip cracks are louder than others:

As we all know, certain cracks are louder than others. They can fit into 4 categories:

Termination of Linear Momentum - e.g. Flicks

Reversal of Linear Momentum - e.g. Towel cracks

Termination of Rotational Momentum - e.g. Cattlemans

Reversal of Rotation Momentum - e.g. Overheads

For cracks using the same whip, this is the order from quiet cracks to loud cracks. This is because when a the momentum of the whip terminates, the cavitation bubble it carries separates at the same velocity as the end of the whip prior to separation. When we reverse the momentum (pull back on the whip) the inertia of the cavitation bubble (more accurately the mass of the air around the cavitation bubble) carries it foward while the cracker moves backwards, so the relative escape velocity of the cracker and cavitation bubble is increased. The same is true for when we have a loop in the crack, the main difference is that the bubble developed is extend in a linear fashion or created an expanding development of vorticity with added angular momentum. Imagine the difference of a single gram of gun powder in a closed tube versus 2 grams (linear + angular) of gun powder not in a tube being set off.

Much more complicated mathematics is involved, but if you think about the cavition bubble moving away from the popper at with a certain amount of momentum, the following occurs:

Termination of Linear Momentum = > Linear Velocity of Cavitation Bubble

Reversal of Linear Momentum = > Linear Velocity of Cavitation Bubble + Linear Velocity of Retreating Popper

Termination of Rotational Momentum = > Linear Velocity of Cavitation Bubble + Angular Velocity of Cavitation Bubble

Reversal of Rotational Momentum = > Linear Velocity of Cavitation Bubble + Angular Velocity of Cavitation Bubble + Linear Velocity of Retreating Popper + Angular Velocity of Retreating Popper

Imagine if there are two spheres in water, if you pull them apart water will rush in between them. If fast enough, water can rush in fast enough and the liquid water will turn into water vapor. This is what happen if propellers in water spin too fast. If you only pull on one sphere the volume in between expands slower than if both moved. Now if you not only pull them apart, but you also increase their size and spin them, even more space develops between them as well as vorticity. This is what happens when a loop in the whip terminates, the inertial fluid drag pocket is now a torus (donut shaped) that is spinning. The torus is the motion envelope of the inertial fluid drag pocket and not just the pocket itself. Also the spinning action will cause a centrifugal reaction assisting the toroidal cavitaion bubble to increase in size. Imagine the air/water around it on a merry-go-round wanting to continue motion outward.

I hope this helps explain what happens when a whip cracks!

My basis: A Master's Degree in Multi-Phase Interaction in Computational Fluid Dynamics; Programmer for Institute of Hydraulic Research, Research with Eglin Air Force Base in High Speed Fluid Dynamics; Speaker at an American Institute of Aeronautics and Astronautics Conference on Multi-Phase, Multi-Scale, and Multi-Material Flow; and Discussion with Japanese Universities on Shockwave Production and Numerical Simulation of Shockwaves.

P.S. I have debated finding a University to acquire my Ph. D. in Fluid Dynamics with my dissertation on Whip Cracking Shock Wave Dynamics.