Some clever things going on in this Exercise 2 (filming a scenario in a single take, from a moving camera). It starts with a shot through a window into a cafe; a character in a blue shirt with his back to us; we’re not being encouraged to notice anything in the shot: it’s played as a generalised wide shot setting a scene. Then another character comes into frame, and the camera sets off, walking behind her into the cafe, snaking round the tables like a daylit Goodfellas, until she stops at a low table, another female character, and the guy in the blue shirt, facing us now. The camera has brought us all the way round, 180 degrees, to the other side of the window. Very clever.
And when we get there, we have S, playing an immediately recognisable shady character. How so? It’s all in the eyes, furtive glances to either side, comedy winking, darting from one character to the other. Screen acting is all done with the eyes, and here you can actually see S switching his character on. And a note on Abd, in charge of the camera, stepping in a little, to mirror S’s conspiratorial beckoning in. The camera, and we, join in the scene at this point..
And what does it have to do with the long take? Performance, continuous and integral performance, is underwritten, guaranteed, by the long take. You can’t fake it if the camera is rolling.