Seriously, it is exceptionally sloppy. It is definately done by someone that doesn't understand accoustics or has worked very much with audio mixing.
I would imagine there are ratehr big differences in level from explosion to the rest, though I haven't pulled it in to see. It is a very canned effect, and sounds exactly the same each time (which is a physicall impossibility) as well as at times seems disproportionately levelled to the rest of what is going on in the picture.
That panasonic uses auto volume controls through compression and gain control. if when compression doesn't prevent clipping the input it lowers level, which means it should look alot like mastered audio with little change in level aside from slight peaks.
If there's serious changes, it was faked.
It's almost as if it was done by a kid who didn't know much about sound and accoustics, or mixing properly for foley work.
...almost? no, I take that back: it is exactly like that.
It has
some dynamic range. In the 100 ms preceding the second "explosion" the peak levels are -2.75 to -3 dBFS; in the 100 ms after the attack of the "explosion" they go up to -0.7 to -0.8 dBFS. The RMS values increase too, but only by a little more than 1 dB. It also doesn't have the "pushed through an L2 set on stun" look I'm familiar with from mastered musical material, nor does it appear heavily squashed in general.
It also lacks the artifacts I'd expect from the sort of limiters you're likely to find in a camcorder- tight slope above threshold, fast attack and relatively slow release times- when confronted by the pressure waves from explosions. The arrival of a loud bang should cause the recorder to reduce gain and exert a ducking effect on the background sounds, but this doesn't seem to happen in this clip. Just before the first "explosion" there's an odd falling tone that sounds like it starts at about 800 Hz and sweeps downwards. The "explosion" has no effect on its level. That strikes me as wrong.
Another observation. If you listen to the original video, it's stone mono, which is quite consistent with a camera mic.
Not so the audio clip from mediumrecords. It's genuine stereo. The really interesting thing about that is that the spatial characteristics of the two "explosions" are very different. The first one is quite narrow and up the center while the second one is very wide and hasn't much of a center component.
Now, if you had set up a stereo micing arrangement where the camera was and then recorded two explosions a couple of seconds apart high up on the tower, the acoustic path between the blasts and the mics
wouldn't change significantly between explosions. If local reflections produce an impression of space on the second explosion, they would do exactly the same for the first. For the two to take place in what are audibly different acoustic environments is just plain wrong and suggests the use of canned effects which weren't chosen carefully.
There's another thing. The "explosions"
reverberate. While the recording didn't take place in a perfectly free field, outdoors comes fairly close. I would expect large buildings nearby to produce discrete echoes, and reflections from one's immediate surroundings to be audible, but the kind of crashing, several-second-decay-time, gymnasium-like reverberation I hear on the "explosions" is just
absurd.
Phony as a Mexican shilling.
Of course, these are neither measurements nor calculations, nor have I got any degrees to wave. They're just the observations of an old sound man; take 'em for whatever you think they're worth.