Inception cinematic explained

the cinematic is done with 2 recordings, one with greenscreen sky.
it is very simple, but time consuming. Requires simple math.

record one facing up and recording one facing down, first recording movement has to match the other movement except for one thing
it has to be mirrored, and the z position has to be higher.

export campath with cam_export in iw5mvm
original path:

0_7.28818e+06/80/-0.142083/0.309782/-0.046876/-0.938962/-125.07/-1418.12/242.749;7.28871e+06/80/-0.448162/0.0924428/-0.0466581/-0.887935/-424.755/-1291.92/67.7724;7.28932e+06/80/0.969426/0.0209061/-0.227716/0.0890008/-589.636/-794.224/122.207;7.28985e+06/80/0.93524/0.0606329/-0.287798/0.197035/-562.962/-303.354/113.727;7.29088e+06/80/0.722695/-0.0351666/-0.689286/-0.0368712/-564.978/527.438/216.252;


organised:

0_7.28818e+06/80/-0.142083/0.309782/-0.046876/-0.938962/-125.07/-1418.12/242.749;
7.28871e+06/80/-0.448162/0.0924428/-0.0466581/-0.887935/-424.755/-1291.92/67.7724;
7.28932e+06/80/0.969426/0.0209061/-0.227716/0.0890008/-589.636/-794.224/122.207;
7.28985e+06/80/0.93524/0.0606329/-0.287798/0.197035/-562.962/-303.354/113.727;
7.29088e+06/80/0.722695/-0.0351666/-0.689286/-0.0368712/-564.978/527.438/216.252;


We can display it as the following:

demotick/fov/aw/ax/ay/az/x/y/z;

or you can call it like this
Point in timeline/fov/viewport W angle/viewport X angle/viewport Y angle/viewport Z angle/x position/y position/z position; (; separates this point from the next point)

Apply the following sums (to mirror):
ay * -1
az * -1
 for example, if ay is 1, it becomes -1, if az is -0.3 is becomes 0.3

brings you to the following:

0_7.28818e+06/80/-0.142083/0.309782/0.046876/0.938962/-125.07/-1418.12/242.749;
7.28871e+06/80/-0.448162/0.0924428/0.0466581/0.887935/-424.755/-1291.92/67.7724;
7.28932e+06/80/0.969426/0.0209061/0.227716/-0.0890008/-589.636/-794.224/122.207;
7.28985e+06/80/0.93524/0.0606329/0.287798/-0.197035/-562.962/-303.354/113.727;
7.29088e+06/80/0.722695/-0.0351666/0.689286/0.0368712/-564.978/527.438/216.252;


Get the difference between points (in xyz)
(second point minus first point, third point minus second point, and so on..)

x:
-125.07 + -299.685 = the next point => -424.755
-424.755 + -164.881
-589.636 + 26.674
-562.962 + -2.016
-564.978


y:
-1418.12 + 126.2
-1291.92 + 497.696
-794.224 + 490.87
-303.354 + 830.792
527.438


z:
242.749 + -174.9766
67.7724 + 54.4346
122.207 + -8.48
113.727 + 102.525
216.252


now we have to invert (*-1) these differences and add them to the new xyz starting point (preferably somewhere up high)

new starting point (xyz): 400/900/1400

x:
400 + (-299.685*-1) = 699.685
699.685 + (-164.881*-1) = 864.566
864.566 + (26.674*-1) = 837.892
837.892 + (-2.016*-1) = 839.908
839.908


y:
900.899 + (126.2*-1) = 774.699
774.699 + (497.696*-1) = 277.003
277.003 + (490.87*-1) = -213.867
-213.867 + (-830.792*-1) = -1044.659
-1044.659


z:
1400 + (-174.9766*-1) = 1574.9766
1574.9766 + (54.4346*-1) = 1520.542
1520.542 + (-8.48*-1) = 1529.022
1529.022 + (102.525*-1) = 1426.497
1426.497


add to the campath:

0_7.28818e+06/80/-0.142083/0.309782/0.046876/0.938962/400/900.899/1400;
7.28871e+06/80/-0.448162/0.0924428/0.0466581/0.887935/699.685/774.699/1574.9766;
7.28932e+06/80/0.969426/0.0209061/0.227716/-0.0890008/864.566/277.003/1520.542;
7.28985e+06/80/0.93524/0.0606329/0.287798/-0.197035/837.892/-213.867/1529.022;
7.29088e+06/80/0.722695/-0.0351666/0.689286/0.0368712/839.908/-1044.659/1426.497;

de-organise and put it in a .cam file (go to your favorite text editor, paste the new campath and save as .cam)

0_7.28818e+06/80/-0.142083/0.309782/0.046876/0.938962/400/900.899/1400;7.28871e+06/80/-0.448162/0.0924428/0.0466581/0.887935/699.685/774.699/1574.9766;7.28932e+06/80/0.969426/0.0209061/0.227716/-0.0890008/864.566/277.003/1520.542;7.28985e+06/80/0.93524/0.0606329/0.287798/-0.197035/837.892/-213.867/1529.022;7.29088e+06/80/0.722695/-0.0351666/0.689286/0.0368712/839.908/-1044.659/1426.497;


go ingame anosud use cam_import to load the new camera movement

in post, you're gonna have to mirror the footage both horizontally and vertically (scale*-1)

PS i recommend you use the frozen dollycam function, works better and is easier to use when combing maps :)
ALSO, i included an excel file to ease the process of calculating the mirrorred coords.