A.T. said:
Maybe you know some movies from other labs. I saw some videos of tunnels stating up quickly to supersonic, but there is to much noise (shock waves from the walls?) at Mach 1. What one would need is a slow transition. If there was a shock wave going back, it would move at Mach 2 relative to the camera. This is fast, but hi-speed cameras capture shock waves propagating at Mach 1 relative to the camera, so Mach 2 should be possible to.
It wouldn't necessarily be moving at Mach 2 relative to the camera. When you start a supersonic tunnel, the air is initially at rest or approximately at rest relative to the camera. The starting shock then travels through the test section, and does so at the design Mach number as long as it remains a normal shock. If it doesn't (it often doesn't) then it's a little more complicated but it won't be a drastically different speed. This means it is propagating into the stagnant air at the design Mach number, meaning Mach 6 for the tunnel I use, for example, not Mach 12.
Capturing shock motion on video is possible for lower Mach numbers if you have the right equipment. You need a very bright light source to illuminate your image sensor on your camera over such a short duration. You need a very fast camera (generally 10's of kfps), and you need a relatively low Mach number. For example,
this paper have a pretty neat schlieren technique detailed in it. You'll notice it was published in 2005, so it wasn't until recently that this was even possible, and even then it was novel and uncommon enough as to be publishable in a relatively well-known journal.
I've also heard of it being done with lasers, but that doesn't generate a movie; it generates a point measurement or a series of time-resolved point measurements. We also did some work here using a focusing schlieren system and measured at 2 MHz, but those were done with photodiodes and, again, were point measurements, not movies.
So yes, it's possible, but not trivial to procure the equipment or cheap, so unless you have a good reason to do it (we don't), then generally you don't do it.