I'm wondering why, in films, we usually see and hear lightning and thunder at the same time?
Is there a theoretical reason for this? Or is this just because it would be weird to see the lightning, and then hear the thunder with a time offset?
Would it not be possible to make the thunder's sound time-shifted, depending on its distance, in order to enhance realism (within a shot that is long enough, of course) ?
Most film is art, not life/reality. Sound designers have to match the visual art on screen with the sonic art of their mix, and that usually includes a certain amount of "realer than real", suspending some of what we "know" about physics in order to tell the story in a way that translates to this 2-dimensional, 2-sense media.
Case in point, when we see some big impact on screen, such as a lightning bolt, we as audience members expect to hear it, much as we expect to hear the starship Enterprise whooshing by at warp speed, even though there's no air (and thus no air movement to create the "whoosh") in space. We also expect to see it match up with the visual effect on the screen; this both increases the overall impact of whatever just happened, and keeps us suspending disbelief that we're not watching a movie, we're looking through a window into another universe. Hearing a delayed sound effect might jerk us out of the director's world and wonder what's going on with the media system, and once we've left, it's hard to get us back.
The speed of sound delay would also, even in the ideal, make us feel further from the action. You see a lightning bolt hit a distant mountainside. 3 seconds later, you hear an ear-splitting thunderclap. While that might be totally accurate to the scope of the shot, in the ultimate ideal of the audience totally suspending reality and believing they're physically in this world you've created, they'll be thinking "damn, I'm glad I'm not over there", when in cinema the location of the camera can jump from place to place as needed, and the very next jump cut could put the audience right in the thick of the aftermath of that thunderbolt.
That also introduces another problem; from the previous perspective, the thunderbolt happened three seconds ago. So, when you jump-cut to the site of the impact, do you show it at the moment the thunderbolt hits, effectively going back in time 3 seconds, or do you show it in "real-time", as if the camera instantaneously moved in space while time progresses "normally"?
As a different example, let's say that, in the climactic scene of Return of the King, Peter Jackson decided that the sound produced by Sauron's eye imploding and then exploding should be accurately delayed based on the camera's distance, because the event produced a big visual shockwave and so it would be believable to have the sound hit you when the shock reached the camera. First off, at its full height that tower is almost a mile high. To have the entirety of it in frame, even as it's collapsing, without any vertical compression of perspective, the camera has to be even further away from it. So, the shot of Sauron exploding and the tower disintegrating from the shock would be unfolding in front of you, sound-effect-less, for over 4 seconds before the shock wave (and thus the sound) hits the observer at the camera. While PJ might have been able to make it work, that's a long time to be expecting the biggest explosion ever heard in Middle Earth and not get it.
Despite the practical difficulties in making these types of sound delays believable in film, such delays are sometimes used, but in "enhanced" situations; the visual effects team might speed up the speed of sound, and thus the shockwave, to lessen the time the shot has to go on before the sonic payoff. As an audience, we're not sitting there with a stopwatch and laser rangefinder to figure out whether the speed of sound is accurate in the shot; we see a big shockwave, then a short moment later we see and feel it hit us and go "whoa, this is pretty realistic".
You'll also see sound designers architecting a "double explosion". They'll engineer a big, powerful explosion sound right at the moment you see it, then the visual effects guys will put in a massive shock wave, and as that passes by the camera the sound designers will add an extra "whoosh". Or, with a slight delay (and a really fast shockwave), the designers will engineer a small crescendo to the big boom, so you know it's coming. These little audio tricks aren't what you hear in comparable "real life" events, but then again, neither is the sound of a movie punch or a movie explosion compared to the real thing.
Film isn't documentary (unless it is a documentary) and it constantly alters/plays with the flow of time - every picture edit is effectively interrupting the flow of time... Do conversations always occur tightly edited or perfectly/dramatically paced in real life? So film isn't necessarily about realism, which is effectively what you are asking. And the same applies to many aspects - do voices through telephones actually sound like they do in films? of course not, its only evoking the effect so we 'read' it correctly, it is not documenting the actual effect. But in terms of physics, this issue could also apply to any distant sound FX - early in my career I remember sitting on a wharf watching a poledriver hammer huge poles in, but it was quite a way away from me & there was a very noticeable delay in the audio reaching me. I thought about how if I presented sound FX for that moment the way they occur in reality, would likely be deemed out of sync...
(it also could be about trying to accentuate drama - if thunder & lightning occur at the same time to the viewer/listener, it means the lightning strike is very close to you)
I think its because as you say; the time offset would be odd in a narrative. In films the point of the thunder/lightning is to possibly give information about the environment for plot reasons or to enhance the mood of a scene. In mot cases there is no story reason to have "accurate" time offset of the thunder.
DISCLAIMER: I am veeeerry tired and maybe the following text is just plain stupid. In that case: Excuse me and ignore the following :)
I think thunderclaps can be used in different ways. If thunder is used only as a dramatic effect, there is no reason to be realistic when it comes to time delay. In those cases the thunder could be understood as something that isn't even part of the universe of the movie. It's more like a title used to introduce a new chapter in the film. Therefore it would actually be 'wrong' to use a time-offset, because you'd make it a part of the reality of the movie. Furthermore it speeds up the time between the approaching storm and the actual stoem.
If the thunder/approaching storm IS presented as a part of the reality of the movie, then a time-offset would totally make sense.
Of course it could also be just an aesthetic decision :)
Sound travels 330m/sec so if the thing is 2-3 km away, which is very close, there would be a 6-9 sec delay between visuals and sound.
The average picture in a move is 3-4 seconds so a realistic sound would interfere with the typical cuts in a film.
-> So a realistic sound design makes only sense if the scene is planned for a really thunder :)
Last night I was playing The Last of Us, not a film, a video game. There was a cut scene where the characters look off into the distance and a storm can be seen brewing on the horizon. Lighting fills the clouds but the thunder breaks a second later as it would in real life. It worked really well and gave the game a sense of reality. It also worked as a cue to tell me, the player, that the storm, which was over the place I had to get to, was still a fair distance away.
I think in this way a delay between the lighting strike and the thunder crash is something you could think about using.
