Flickering in the US

[Erik Nordlund]

Hi everyone.

 

I currently work as an AC/technical magician ;) on a documentary project called Press Pause Play. In short terms the film aims to explore how artists and musicians around the world is affected by ongoing changes in how we produce and consume media. (The internet)

 

Anyways. For the following three weeks we will be in the US talking to different people and I'm wondering; since we're shooting 25 fps on a 180 degree shutter. (The Swedish way of doing things) What will happen to for example fluorescent lights in the US where the electrical system is 60hz? We're shooting on RED.

 

I've been talking to a lot of people here and no one seems to have a really good answer.

Edited September 30, 2009 by Erik Nordlund

[John Sprung]

Most newer flourescents and CFL's have high frequency ballasts, and should be fine. They're more efficient than the old magnetic ballasts, most of which have been replaced by now.

 

But if you do find magnetic ballasts somewhere, you have a problem. Given 60 Hz. power, they produce a light pulse on every half cycle, so you have 120 pulses per second. Shutter angle doesn't matter here like it does with CRT TV's. The critical thing is that the light pulse rate must be an integer multiple of your frame rate. With a film camera, you'd be OK with 30 x 4 = 120, or 24 x 5 = 120. Shooting 24 fps and speeding everything up 4% would work for film.

 

You'd have to shoot 24 fps on the Red as well. But on the Red, you have yet another consideration. The rolling shutter does unfortunate things with pulses of light, such as strobes. The light output of a flourescent starts at zero where the AC sine wave crosses zero, and stays there until it reaches a high enough voltage to strike an arc. For the rest of the half cycle, it follows, IIRC, Sin^2 or something like that. So, if you get dark horizontal bands at 180 degrees, you may have to open up to 360 degrees. That way the entire frame is exposed to an integer number of light pulses -- five.

 

 

 

 

-- J.S.

[Erik Nordlund]

Most newer flourescents and CFL's have high frequency ballasts, and should be fine. They're more efficient than the old magnetic ballasts, most of which have been replaced by now.

 

But if you do find magnetic ballasts somewhere, you have a problem. Given 60 Hz. power, they produce a light pulse on every half cycle, so you have 120 pulses per second. Shutter angle doesn't matter here like it does with CRT TV's. The critical thing is that the light pulse rate must be an integer multiple of your frame rate. With a film camera, you'd be OK with 30 x 4 = 120, or 24 x 5 = 120. Shooting 24 fps and speeding everything up 4% would work for film.

 

You'd have to shoot 24 fps on the Red as well. But on the Red, you have yet another consideration. The rolling shutter does unfortunate things with pulses of light, such as strobes. The light output of a flourescent starts at zero where the AC sine wave crosses zero, and stays there until it reaches a high enough voltage to strike an arc. For the rest of the half cycle, it follows, IIRC, Sin^2 or something like that. So, if you get dark horizontal bands at 180 degrees, you may have to open up to 360 degrees. That way the entire frame is exposed to an integer number of light pulses -- five.

 

-- J.S.

 

Thank you John for your extensive answer!

It's hard to do reliable tests until i get to the U.S. That's the real tricky part with this. But I guess that's the way it's gotta be.

 

Thanks again!

[Phil Rhodes]

I've shot quite a bit of PAL video in the US, though I don't know how this will equate to your progressive scan situation. There is sometimes slightly visible flicker in fluorescents, but it isn't terribly bad.

 

P

[Guest zhuo chun bei]

Anybody knows the numbers of http://www.eiluxury.com louie vuitton in CA, I need to call them and get my bag repaired.

[Michael Nash]

I've shot quite a bit of PAL video in the US, though I don't know how this will equate to your progressive scan situation. There is sometimes slightly visible flicker in fluorescents, but it isn't terribly bad.

 

P

 

For 25fps/60Hz shooting, set your shutter to 172.8 degrees instead of 180 (1/60th of a second exposure instead of 1/50).

[John Sprung]

For 25fps/60Hz shooting, set your shutter to 172.8 degrees instead of 180 (1/60th of a second exposure instead of 1/50).

 

Actually, 172.8 degrees is the magic number for shooting PAL/SECAM CRT's at 24 fps instead of 25 fps.

 

25 fps at 180 degrees gets you an exposure of 1/50 second. Reducing the shutter angle to 172.8 cuts the duty cycle down from 50% to 48%, producing the same 1/50 second exposure at 24 fps. With CRT's, if your exposure exactly matches the field period, you get rid of the roll bar. It's still there, but its height is zero lines. Your shutter opens when the CRT is at some point in its scan, and closes exactly when it gets back to that same place on the TV screen.

 

With flourescent lights, the considerations are different. There's no scanning, just areas of the picture that are illuminated by a source that pulsates at twice the power frequency. To not flicker, you need to get the same amount of light and dark from the source in every exposure. How long the exposures are isn't as important as having the frame rate be a whole number multiple of the light pulse rate. So, here in the 60 Hz = 120 light pulses per second part of the world you could use:

 

1 x 120 = 2 x 60 = 3 x 40 = 4 x 30 = 5 x 24 = 6 x 20 = 7 x 17.143.... = 8 x 15 = 9 x 13.3333.... = 10 x 12, etc.

 

Shooting with a really small shutter angle runs the risk of some takes hitting the dark part of the flourescent cycle, between the zero crossing and reaching strike voltage again. But other than that, shutter angle isn't so important here. For shooting CRT's it's critical, but not for pulsating light sources.

 

 

 

 

-- J.S.

[Michael Nash]

Actually, 172.8 degrees is the magic number for shooting PAL/SECAM CRT's at 24 fps instead of 25 fps.

 

...

 

With flourescent lights, the considerations are different. There's no scanning, just areas of the picture that are illuminated by a source that pulsates at twice the power frequency. ... But other than that, shutter angle isn't so important here. For shooting CRT's it's critical, but not for pulsating light sources.

 

You're right, I had my "magic numbers" reversed. I should have said 150 degrees for 25fps/60 Hz. (172.8 is indeed for 24fps/50hz.).

 

You're also right that fluorescent lights usually don't cause an appreciable flicker when shot slightly off-speed (as in this case); however my experience with PAL shooting has always shown some slight flicker from many 60Hz. sources, such as discharge-type street lights, etc. Knocking the shutter down from 1/50 to 1/60 always clears it right up with no visible affect on motion blur. But that's 50 field PAL; never tried it with 25 fps progressive on the RED.

[Tim O'Connor]

Hey Michael, haven't seen you for a while. Good to see you!

[John Draus]

You're right, I had my "magic numbers" reversed. I should have said 150 degrees for 25fps/60 Hz. (172.8 is indeed for 24fps/50hz.).

 

You're also right that fluorescent lights usually don't cause an appreciable flicker when shot slightly off-speed (as in this case); however my experience with PAL shooting has always shown some slight flicker from many 60Hz. sources, such as discharge-type street lights, etc. Knocking the shutter down from 1/50 to 1/60 always clears it right up with no visible affect on motion blur. But that's 50 field PAL; never tried it with 25 fps progressive on the RED.

 

 

Would this 150deg shutter also work for Magnetic ballast HMI's when shooting 16mm at 25fps in a 60htz setting?

[John Sprung]

Shutter angle is only important for shooting film with CRT TV sets in the shot. For HMI's, frame rate is the only thing you have to constrain to an integer number of light pulses.

 

 

 

 

 

-- J.S.