The average shot length (ASL) of films in 2014 was 2.5 seconds. For better or for worse, this has been the average for the past decade or so, so we’re used to it as viewers. Anything longer or shorter, we notice. But what does it do to the dynamics of a scene when we notice?
The border crossing scene in Sicario demonstrates how varying shot lengths can expand or compress time, and thus dramatically increase the tension in a scene.
The scene itself is short on surprises. All exteriors are shot in wide angle with deep focus; whatever bad guys are out there are not only in frame and in focus the entire time, they’re even pointed out to us by the good guys. In other words, we’re not wondering IF something is going to happen, we’re wondering WHEN. And when is manipulated by shot length.
The ASL of the entire scene, minus its long establishing and closing shots, is 2.7 seconds, so right in the ballpark of today’s ASL. As the scene unfolds, the good guys go through various actions. They scan for threats (an ASL of 5.8 seconds), they prep for combat (4.8 seconds), and they aim their weapons (3.1 seconds). We notice how long these shots are because they’re so much longer than the average ASL. And because they’re so long, we assume that something is about to happen – otherwise, why hold our attention on one character, or one car, for so long? – and expect the next cut to throw us into the action. Each time it doesn’t, however, the tension grows.
And not only does the next cut NOT lead to action, it sometimes leads us into an even longer shot: a 2 second shot of del Toro aiming is followed by a 9 second shot of the bad guys getting their weapons ready. It also doesn’t help that most of these shots are static. The scanning shots consist merely of cars moving forward slowly, the camera panning slowly over them, or both. The prep shots are almost completely static, with just windows being rolled down. And the aiming shots are completely static, with Benicio del Toro’s character motionless.
When someone finally moves, the tension explodes. They move fast, too: as the good guys get out of their cars, the ASL is 1.4 seconds. And it feels fast, both because the cuts are slightly faster than the overall scene (1.4 vs 2.7 seconds) and SO MUCH faster than everything that’s come before it (1.4 vs 5.8 seconds). When violence erupts, it’s faster still, with an ASL of just .67 seconds or 15 frames. Thus it’s natural that Emily Blunt’s character is surprised when it erupts – so are we.
But Sicario doesn’t just decrease ASL for its action scenes, it expands and contracts ASL as the scene continues. After the good guys pile out of their cars (an ASL of 1.4 seconds), they end up in a stand off (2.3 seconds). This noticeable switch in tempo doesn’t give us a breather so much as put us back where we were, waiting for something to happen again. The increase in ASL has provided no relief. This push and pull continues, with the ASL relaxing after each burst of gunfire, from .67 to 2 seconds after the first and from .62 to 2.7 after the second.
After the second burst of gunfire, we should be able to relax. Since the scene so painstakingly kept us aware of all the bad guys, with eight separate shots of their cars getting into position, and now they’re all dead, we should be in the clear. With both groups of bad guys neutralized, Emily Blunt’s character does a long 4 second scan of the carnage. It’s over.
But this shot ends with our first jump scare of the scene, with the reveal of the crooked cop in the rearview mirror. Tension is now at its highest as not only does the ASL accelerate (to .87 seconds or 22 frames), but suddenly we’re back to wondering not just WHEN but also IF. Even as the ASL falls back to 2.7 seconds for the next few shots, it’s not immediately clear if the danger is past. Only the 11 second closing shot signals it’s over.
They say people in stressful situations experience time compression. Sicario achieves the cinematic equivalent with its ASL manipulation. While The AV Club called the border crossing scene one of the best action scenes of 2015, only 18.6 of its 251 seconds involve any action. Tensest scene may be more accurate.