what is the normal shutter speed for feature movies?

[Gabriel Garcia Rosa]

i'm a young filmmaker learning my craft through trial and error and i am quickly becoming respected as a d.p. i have experiment with all speed's of the shutter and i'm enjoying the the affect of all of them but i must know what is the normal shutter speed for a feature movie? :001_smile:

 

[Peter Lundström]

Well if you mean on digital I'd guess 1/48...

 

[Charles Angus]

Normally film runs at 24fps, with a 180° shutter - which would indeed be 1/48th second.

 

[Ido Karilla]

50HZ or 60HZ????!!!!

50HZ or 60HZ????!!!!

All previous answers refer to ntsc land 60 HZ, if in pal land and running 24fps then 1\50 or 172.8 deg will be your base shutter!!

 

[David (dudi) Namir]

All previous answers refer to ntsc land 60 HZ, if in pal land and running 24fps then 1\50 or 172.8 deg will be your base shutter!!

and if shooting at 25fps 1/50 will be 180 deg.

 

[Yousuf Abbasi]

Dumb question here, but: Why exactly is 1/48 the base shutter speed (for ntsc), and how does it relate to 180 degrees?

 

[David Mullen ASC]

1/48th at 24 fps comes from film, not NTSC.

A film camera has a spinning disk that rotates in front of the film gate, with a pie slice hole cut-out to allow light to pass through and expose the film, when the solid portion of the disk rotates in front of the film, nothing gets exposed so the camera can advance the film to the next frame to be exposed when the open portion of the disk rotates again in front of it.

The standard movie camera shutter "angle" is 180 degrees, i.e. the rotating metal disk is half-open (a circle being 360 degrees) or you could describe it as a half-circle.

So at 24 frames per second, the film is in the gate for 1/24th of a second but only being exposed for half that time, thus 1/48th of a second.

So a 24P video camera is often set with an electronic shutter speed of 1/48th of a second to match the look of the 180 degree film shutter angle. The video camera itself does not have a real 180 degree shutter angle because it does not have a physical shutter moving in front of the sensor.

NTSC is 59.97 Hz because our power supply is 60 Hz. The ideal shutter speed would therefore be 1/60th of a second, which is a 144 degree shutter at 24 fps, but 180 degrees is within an acceptable window.

PAL countries have a 50 Hz power supply so the ideal shutter speed is 1/50th of a second, which is what you get with a 25 fps camera and a 180 degree shutter. Or at 24 fps, it would be a 172.8 degree shutter angle in order to get a 1/50th shutter time.

Ignoring the issue regarding safe shutter speeds and camera speeds to avoid flicker under AC discharge lights, as long as you are in the ballpark of the look of a 180 degree shutter, it's fine because motion blur -- which is what we are talking about -- is affected by how much movement there is. So for a shot of someone talking, it would be hard to see the difference between a 1/48th shutter time and a 1/60th shutter time, for example.

What is generally desirable is to avoid the classic look of 50i or 60i (interlaced-scan) video which is often shot with no shutter at all, i.e. shutterless or a metaphorical 360 degree shutter angle. There needs to be some dark-light-dark-light temporal gaps in the motion sampling that a shutter can create to get closer to how a film camera renders motion.

 

[Yousuf Abbasi]

1/48th at 24 fps comes from film, not NTSC.

A film camera has a spinning disk that rotates in front of the film gate, with a pie slice hole cut-out to allow light to pass through and expose the film, when the solid portion of the disk rotates in front of the film, nothing gets exposed so the camera can advance the film to the next frame to be exposed when the open portion of the disk rotates again in front of it.

The standard movie camera shutter "angle" is 180 degrees, i.e. the rotating metal disk is half-open (a circle being 360 degrees) or you could describe it as a half-circle.

So at 24 frames per second, the film is in the gate for 1/24th of a second but only being exposed for half that time, thus 1/48th of a second.

So a 24P video camera is often set with an electronic shutter speed of 1/48th of a second to match the look of the 180 degree film shutter angle. The video camera itself does not have a real 180 degree shutter angle because it does not have a physical shutter moving in front of the sensor.

NTSC is 59.97 Hz because our power supply is 60 Hz. The ideal shutter speed would therefore be 1/60th of a second, which is a 144 degree shutter at 24 fps, but 180 degrees is within an acceptable window.

PAL countries have a 50 Hz power supply so the ideal shutter speed is 1/50th of a second, which is what you get with a 25 fps camera and a 180 degree shutter. Or at 24 fps, it would be a 172.8 degree shutter angle in order to get a 1/50th shutter time.

Ignoring the issue regarding safe shutter speeds and camera speeds to avoid flicker under AC discharge lights, as long as you are in the ballpark of the look of a 180 degree shutter, it's fine because motion blur -- which is what we are talking about -- is affected by how much movement there is. So for a shot of someone talking, it would be hard to see the difference between a 1/48th shutter time and a 1/60th shutter time, for example.

What is generally desirable is to avoid the classic look of 50i or 60i (interlaced-scan) video which is often shot with no shutter at all, i.e. shutterless or a metaphorical 360 degree shutter angle. There needs to be some dark-light-dark-light temporal gaps in the motion sampling that a shutter can create to get closer to how a film camera renders motion.

Makes perfect sense (and pretty fascinating, too). Thanks David. :thumbup1:

 

[Gabriel Garcia Rosa]

1/48th at 24 fps comes from film, not NTSC.

A film camera has a spinning disk that rotates in front of the film gate, with a pie slice hole cut-out to allow light to pass through and expose the film, when the solid portion of the disk rotates in front of the film, nothing gets exposed so the camera can advance the film to the next frame to be exposed when the open portion of the disk rotates again in front of it.

The standard movie camera shutter "angle" is 180 degrees, i.e. the rotating metal disk is half-open (a circle being 360 degrees) or you could describe it as a half-circle.

So at 24 frames per second, the film is in the gate for 1/24th of a second but only being exposed for half that time, thus 1/48th of a second.

So a 24P video camera is often set with an electronic shutter speed of 1/48th of a second to match the look of the 180 degree film shutter angle. The video camera itself does not have a real 180 degree shutter angle because it does not have a physical shutter moving in front of the sensor.

NTSC is 59.97 Hz because our power supply is 60 Hz. The ideal shutter speed would therefore be 1/60th of a second, which is a 144 degree shutter at 24 fps, but 180 degrees is within an acceptable window.

PAL countries have a 50 Hz power supply so the ideal shutter speed is 1/50th of a second, which is what you get with a 25 fps camera and a 180 degree shutter. Or at 24 fps, it would be a 172.8 degree shutter angle in order to get a 1/50th shutter time.

Ignoring the issue regarding safe shutter speeds and camera speeds to avoid flicker under AC discharge lights, as long as you are in the ballpark of the look of a 180 degree shutter, it's fine because motion blur -- which is what we are talking about -- is affected by how much movement there is. So for a shot of someone talking, it would be hard to see the difference between a 1/48th shutter time and a 1/60th shutter time, for example.

What is generally desirable is to avoid the classic look of 50i or 60i (interlaced-scan) video which is often shot with no shutter at all, i.e. shutterless or a metaphorical 360 degree shutter angle. There needs to be some dark-light-dark-light temporal gaps in the motion sampling that a shutter can create to get closer to how a film camera renders motion.

makes perfect sense to me thanks david for explaining that.

 

[D Fuller]

1/48th at 24 fps comes from film, not NTSC.

NTSC is 59.97 Hz because our power supply is 60 Hz. The ideal shutter speed would therefore be 1/60th of a second, which is a 144 degree shutter at 24 fps, but 180 degrees is within an acceptable window.

David, I am constantly amazed by your generosity and patience in sharing your knowledge with the users of this forum. You have a knack for explaining things with remarkable clarity. Thank you.

But you lost me on this point. Wouldn't the ideal shutter speed be 1/60th only if the frame rate were 30 fps (or 29.97)? (It seems to me I remember cameras like the Beaulieu R16 that did have a 1/60th shutter speed at 24fps being very prone to stutter in a panning shot.) I've always thought that longer shutter speeds were better (esp for reducing stutter) but perhaps I need to re-think that.

 

[David Mullen ASC]

It's complicated to understand, even more me, but there are basically safe frame rates and shutter angle combinations. If you are shooting at 24 fps with a 144 degree shutter angle (1/60th of a second), you basically get two pulses per frame from a 60 Hz light -- same for a 25 fps camera with a 180 degree (1/50th) shutter shooting a 50 Hz light.

So a good starting point when setting the shutter on a video camera is to pick a shutter time that is 1/60th or 1/30th for a 60 Hz light (even 1/120th works I believe, but that's only one pulse captured per frame.) But that's only a starting point, you may need to play around a bit using the ECS.

The formula is any frame rate that divides evenly into twice the power frequency is safe. 1, 2, 3, 4, 5, 6, 8, 10, 15, 20, 24, 30, 40, 60, 120 fps for 60 Hz.

Anyway, here's a chart:
http://www.panavision.com.au/Information/Flickerfree60Hz.htm

Someone how understands math and electricity can explain it better than me...

But the principle is that flicker occurs when each frame exposed is getting a slightly different exposure than the proceeding or following frame, and this exposure variation appears as flicker. So there are certain combinations where the amount of AC pulsing captured is consistent frame to frame, and there are combinations where it isn't.

 

[Yousuf Abbasi]

I've been giving this some more thought, and am curious:

For a given a shutter angle and frame-rate, it is possible to calculate the exposure time (or shutter speed); BUT since this relationship does not take into account the area of the actual frame being exposed (e.g. 77-sqmm versus 346-sqmm), does that mean that a 16mm film frame and a 35mm film frame would be identically exposed under a given shutter angle in the same rotary disc? In other words, if the shutter angle is large enough to expose an ENTIRE 35mm frame, but you replace that frame with the smaller 16mm frame, will the exposures be identical (i.e. will the "extra" light just be ignored)?

More of a theoretical question, but I'm still curious. Hopefully I was clear, but in the likely event I was not, let me know and I will clarify.

 

[Stefan Christou]

If you were shooting film then no there would be no difference between exposures as long as you were using the same film stock, albeit cut into a different sized perforated ribon, to fit s16mm or s35mm.

This is the exact same thing as shooting cropping part of your sensor eg. 2k on Red One.

It doesn't matter if you are using different focal length lenses, as long as the T stop is the same in each case.

But if we are talking digital cameras, even using the same class of sensor, such as the forthcoming Scarlet s35mm and Scarlet 2/3", then the sensors would likely have different natural levels of sensitivity to light, and therefore will have a different level of exposure, just like if you used different ASA rated film stocks. (The sensors will probably have different levels of sensitivity because the smaller sensor will have more, smaller photo-sites packed on to the sensor, which have historically been less efficient at capturing light than their bigger brothers.)

I suppose the short answer is yes for film, as long as you are at the same T stop, but no if you are shooting digital.

 

[Yousuf Abbasi]

If you were shooting film then no there would be no difference between exposures as long as you were using the same film stock, albeit cut into a different sized perforated ribon, to fit s16mm or s35mm.

This is the exact same thing as shooting cropping part of your sensor eg. 2k on Red One.

It doesn't matter if you are using different focal length lenses, as long as the T stop is the same in each case.

But if we are talking digital cameras, even using the same class of sensor, such as the forthcoming Scarlet s35mm and Scarlet 2/3", then the sensors would likely have different natural levels of sensitivity to light, and therefore will have a different level of exposure, just like if you used different ASA rated film stocks. (The sensors will probably have different levels of sensitivity because the smaller sensor will have more, smaller photo-sites packed on to the sensor, which have historically been less efficient at capturing light than their bigger brothers.)

I suppose the short answer is yes for film, as long as you are at the same T stop, but no if you are shooting digital.

That's makes a lot of sense, and it's nice to get the digital-analogy too. Great job guys. It's so cool how much one can learn in this place!

 

[David Mullen ASC]

Other than the Arri D21, digital cameras have electronic shutters, so the size of the shutter blade relative to the sensor is a non-issue -- you set shutter speed by time (even when inputting a shutter angle).

There was only one movie camera that was switchable between 35mm and 16mm, the Eclair Cameflex, and switching didn't affect the exposure, the shutter exposed the film for the same length of time regardless of whether the gate was 16mm or 35mm in size.

 

[C. Kim Miles csc]

Another thing that David alluded to but maybe didn't say specifically is that a 1/48" exposure most accurately replicates the human eye's perception of the phenomenon known as "persistence of vision" and therefore provides the most natural-looking reproduction of normal movement.

If you look at (god forbid) a strip of film with motion in it, you'll see that the moving object in the frame is in fact blurred in each frame, yet when played back at normal speed, appears perfectly sharp. This is how the eye interprets movement, and is most accurately reproduced with a 1/48" exposure time.

Just my 2¢...