5-10s: 1ms gaps twice a second (0.2% slower)- small pops audible
10-15s: 10ms gaps twice a second (2% slower)- very audible
20-27s: 1 second of -3% speed, 1 second normal, alternating - sounds noticeably odd

The purpose of this demonstration is to show that any "non-intentional" tempo slowdown WILL demonstrate undesirable and noticeable effects. The only conceivable way around this is a purpose-built time stretching algorithm, but I have yet (personally) to hear one that isn't noticeable on music.

Track used for demonstration purposes is "Shots" by Imagine Dragons

Mikael Olsson 19