What Is a Take? Why Directors Shoot Take 1, Take 2 and More

[satyaos]

What is the difference between take1, take2 etc.? Is it correct there is only one perfect take and rest are imperfect? And does director or editor save the imperfect takes for future purpose?

 

[Jared Isham]

A take is each time the shot is rolled on a piece of action. So, let's say the actor says "Hi, how are you?" - for each time the actor performs "Hi, how are you?" with the camera in the same position it is considered a "take". So take 1, take 2, take 3, etc. is the number of times that scene is performed.

When you move the camera or change the lens then it would be considered a new shot and the take count would start over at take 1.

An example of what a "scene 1" might look like.

Scene - Take:
1-1
1-2
1-3
new setup for the scene
1A-1
1A-2
1A-3
new setup for the scene
1B-1
1B-2
1B-3

Hope that helps.

As far as imperfect takes vs perfect takes. I don't focus on finding the perfect take. I will often piece parts of multiple takes together into a scene. The first half might be better than the second half in one take and the opposite in another. This is a very good reason to let your actors work through an entire scene without interrupting them.

I watch through all the takes before I start editing a scene, so never think a take is unusable because there might be something in it that will save the edit.

 

[satyaos]

Assume for scene-take (1-1) you choose 2-shot for the dialog "Hi, how are you".
Then for scene-take (1A-1) means you choose new setup as OTS for the same dialog "Hi, how are you". Is that right?

 

[Jared Isham]

Yeahp, exactly.
Scene 1= 2 shot
Scene 1A = OTS
Scene 1B = ECU (Extreme Close Up)

When you run out of letters you double them up
1AA
1BA
1CA
or a variation of that...but hopefully you don't have that many shots for your scene... that would be super, super complicated.

 

[satyaos]

Clapperboard mentions 3 things (scene, shot, take), then what is the shot number in the above example?
Eg: A conversation scene
{
X : hi, How r u?
y: i m fine
X : "where r u going"
}

And take increases until the imperfection happens in scene.

Scene, shot, take (1 : A :1) => "hi how r u", "i m fine", "where r u going"
Scene, shot, take (1 : A :2) => "hi how r u", "i m fine", "where r u going"
Scene, shot, take (1 : A :3) => "hi how r u", "i m fine", "where r u going"

Scene, shot, take (1 : B :1) => "hi how r u", "i m fine", "where r u going"
Scene, shot, take (1 : B :2) => "hi how r u", "i m fine", "where r u going"
Scene, shot, take (1 : B :3) => "hi how r u", "i m fine", "where r u going"

Scene, shot, take (1 : C :1) => "hi how r u", "i m fine", "where r u going"
Scene, shot, take (1 : C :2) => "hi how r u", "i m fine", "where r u going"
Scene, shot, take (1 : C :3) => "hi how r u", "i m fine", "where r u going"

Where A= 2-shot, B=OTS, C=Close up are 3-setups

Is the above correct?

 

[Jared Isham]

I never understood the 3 items - that is more of a European approach. In the US we usually it listed as "Roll, Scene, Take." Roll is what camera roll you are on. Each time you reload the camera you change the Roll #. Scene is the shot you are doing (1A, 1B, 1C, etc.) and take is how many times you shoot that shot.

I'm gonna guess that you have it correct in the way you set it up. That appears to be correct.

I think I'll try to put together a simple video in the next few weeks explaining what I know about how it works...but it does appear to be correct in your interpretation.