Pumping colors in fluorescent light at DVinfo.net

Urban Skargren · July 5th, 2007, 01:06 PM

Hi!
Just had a strange experience with pumping colors.
I did a daytime interview in a persons home where I did a spontaneus arrangement with mixed lighting. The interview object was sitting in a sofa and I wanted the background a little out of focus so I had to place him so that the windows were in the background (looked nice with them overexposed). In the ceiling there was a weak incandescent chandelier. So I took the nearest lamp, a fluorescent reading light, and pointed it towards the person and white-balanced on a white paper that the person was asked to hold in front of him.
But, when I came home I noticed how the colors were pumping: changing from bluish to greenish to reddish and back all the time. A couple of seconds bluish, then a couple of seconds of greenish and so on. Very irritating.

Strange thing is that I made three interviews, but the other two were stable in terms of colors. But in these the fluoresent lamp was like 1 meter or more away from the object, where in the 'pumping' interview the lamp was only 0,5 m away from the face.

I'm pretty 100% sure I was using manual white balance, else I couldn't even had seen the white balance triangles on the screen (blinking until steady).
Anybody has experienced something alike? I think I have seen this somewhere else in my recordings before, so I am beginning to have great respect of fluorescent light.

John Miller · July 5th, 2007, 02:47 PM

Have a look at this thread - it's about the same thing:

http://www.dvinfo.net/conf/showpost....09&postcount=9

Daniel Epstein · July 6th, 2007, 07:14 AM

What shutter speed where you using on the camera? You could have been out of phase with the light coming from the flourescent. If you were using a shutter speed other than the electricity being supplied to the light (60 Hz or 50 Hz) you might see this effect.

Urban Skargren · July 6th, 2007, 01:12 PM

1/180. In the other two interviews (same lighting, just a couple of feet further away from the lamp) I used 1/90 and 1/100 and they didn't freak out.
Reason I used these shutter speeds were trying to aquire short DoF. We have 50 Hz electricity here.

Can anyone explain this fenomenon techically?

Daniel Epstein · July 6th, 2007, 08:46 PM

Well you pretty much answered it with the Shutter. The interviews with the shutter at 90 and 100 were close enough to the frequency or a multiple of the frequency of the light 50Hz so you don't see a color shift or strobing. The Faster 180 shutter is going in and out of phase with the lights frequency and you are seeing the color of the lights output changing as it pulses at 50 HZ. It is kind of like shooting an old CRT TV with a shutter speed higher than 50 HZ. You will see less and less of the picture on the CRT as you raise the shutter speed because the Electron beam in the tube is actually scanning from top to bottom every 50th of a second so the faster shutter is taking a picture faster than the scan.

Urban Skargren · July 7th, 2007, 01:33 AM

What about professional fluorescent lights - the same effect?

Marcus Marchesseault · July 7th, 2007, 03:36 AM

Modern, quality fluorescent lights don't have this problem because they pulse thousands of times a second instead of 50 or 60 times. The older fixtures pulse at the speed of the power grid and the new "electronic" ballasts create their own frequency. I'm not even sure the older "magnetic" ballasts are up to code in this country anymore as industrial lighting has had to switch to more energy efficient tubes that generally use electronic ballasts. FYI, compact fluorescents pulse thousands of times a second (electronic ballast) so they should be fine for most applications.

Daniel Epstein · July 7th, 2007, 07:43 AM

Professional flourescent fixtures should not have this problem but you would really have to test the particular fixture to see if the problem comes up with higher shutter speeds since you can't really measure the frequency of the light fixture so you still could end up with a phase issue which takes several seconds to resolve. I have been in situations where flourescent fixtures flickered a lot and others where there was no problem.

Richard Andrewski · July 7th, 2007, 08:30 AM

The best electronic ballasts output driving frequency is from 25,000 to 40,000 hz so its not a problem at that point unless you are doing some kind of super high FPS shooting. Most pro level fixtures are using these type ballasts now. You can still find plenty of magnetic ballasts around as legacy leftovers of the first few generations of office/warehouse lighting but its getting rarer now. There's a formula for figuring safe FPS and shutter speeds when you aren't working with a high frequency ballast--either fluorescent or HMI:

Safe camera speeds 60hz frequency:

Camera speed in fps = shutter angle / 3 / LPEP

Where LPEP = number of light peaks that you want to capture per exposure
period (typically 2 per cycle, or hz).

Safe shutter angles with 60hz frequency:

Shutter angle = speed in fps * 3 * LPEP

For 50hz power replace 3 with 3.6 in both formulas.

Hope this helps.

July 5th, 2007, 01:06 PM	#1
Urban Skargren Major Player Join Date: May 2007 Location: Medellin, Colombia Posts: 225	Pumping colors in fluorescent light Hi! Just had a strange experience with pumping colors. I did a daytime interview in a persons home where I did a spontaneus arrangement with mixed lighting. The interview object was sitting in a sofa and I wanted the background a little out of focus so I had to place him so that the windows were in the background (looked nice with them overexposed). In the ceiling there was a weak incandescent chandelier. So I took the nearest lamp, a fluorescent reading light, and pointed it towards the person and white-balanced on a white paper that the person was asked to hold in front of him. But, when I came home I noticed how the colors were pumping: changing from bluish to greenish to reddish and back all the time. A couple of seconds bluish, then a couple of seconds of greenish and so on. Very irritating. Strange thing is that I made three interviews, but the other two were stable in terms of colors. But in these the fluoresent lamp was like 1 meter or more away from the object, where in the 'pumping' interview the lamp was only 0,5 m away from the face. I'm pretty 100% sure I was using manual white balance, else I couldn't even had seen the white balance triangles on the screen (blinking until steady). Anybody has experienced something alike? I think I have seen this somewhere else in my recordings before, so I am beginning to have great respect of fluorescent light.

July 7th, 2007, 08:30 AM	#9
Richard Andrewski Major Player Join Date: Jun 2006 Location: Shenzhen, China Posts: 781	The best electronic ballasts output driving frequency is from 25,000 to 40,000 hz so its not a problem at that point unless you are doing some kind of super high FPS shooting. Most pro level fixtures are using these type ballasts now. You can still find plenty of magnetic ballasts around as legacy leftovers of the first few generations of office/warehouse lighting but its getting rarer now. There's a formula for figuring safe FPS and shutter speeds when you aren't working with a high frequency ballast--either fluorescent or HMI: Safe camera speeds 60hz frequency: Camera speed in fps = shutter angle / 3 / LPEP Where LPEP = number of light peaks that you want to capture per exposure period (typically 2 per cycle, or hz). Safe shutter angles with 60hz frequency: Shutter angle = speed in fps * 3 * LPEP For 50hz power replace 3 with 3.6 in both formulas. Hope this helps. __________________ Richard Andrewski - Cool Lights USA - RED #114 http://www.coollights.biz

July 5th, 2007, 02:47 PM	#2
John Miller Major Player Join Date: Oct 2005 Location: Hillsborough, NC, USA Posts: 968	Have a look at this thread - it's about the same thing: http://www.dvinfo.net/conf/showpost....09&postcount=9

July 6th, 2007, 07:14 AM	#3
Daniel Epstein Trustee Join Date: Jan 2006 Location: New York, NY Posts: 1,267	What shutter speed where you using on the camera? You could have been out of phase with the light coming from the flourescent. If you were using a shutter speed other than the electricity being supplied to the light (60 Hz or 50 Hz) you might see this effect.

July 6th, 2007, 01:12 PM	#4
Urban Skargren Major Player Join Date: May 2007 Location: Medellin, Colombia Posts: 225	1/180. In the other two interviews (same lighting, just a couple of feet further away from the lamp) I used 1/90 and 1/100 and they didn't freak out. Reason I used these shutter speeds were trying to aquire short DoF. We have 50 Hz electricity here. Can anyone explain this fenomenon techically?

July 6th, 2007, 08:46 PM	#5
Daniel Epstein Trustee Join Date: Jan 2006 Location: New York, NY Posts: 1,267	Well you pretty much answered it with the Shutter. The interviews with the shutter at 90 and 100 were close enough to the frequency or a multiple of the frequency of the light 50Hz so you don't see a color shift or strobing. The Faster 180 shutter is going in and out of phase with the lights frequency and you are seeing the color of the lights output changing as it pulses at 50 HZ. It is kind of like shooting an old CRT TV with a shutter speed higher than 50 HZ. You will see less and less of the picture on the CRT as you raise the shutter speed because the Electron beam in the tube is actually scanning from top to bottom every 50th of a second so the faster shutter is taking a picture faster than the scan.

July 7th, 2007, 01:33 AM	#6
Urban Skargren Major Player Join Date: May 2007 Location: Medellin, Colombia Posts: 225	What about professional fluorescent lights - the same effect?

July 7th, 2007, 03:36 AM	#7
Marcus Marchesseault Trustee Join Date: Nov 2005 Location: Honolulu, HI Posts: 1,961	Modern, quality fluorescent lights don't have this problem because they pulse thousands of times a second instead of 50 or 60 times. The older fixtures pulse at the speed of the power grid and the new "electronic" ballasts create their own frequency. I'm not even sure the older "magnetic" ballasts are up to code in this country anymore as industrial lighting has had to switch to more energy efficient tubes that generally use electronic ballasts. FYI, compact fluorescents pulse thousands of times a second (electronic ballast) so they should be fine for most applications.

July 7th, 2007, 07:43 AM	#8
Daniel Epstein Trustee Join Date: Jan 2006 Location: New York, NY Posts: 1,267	Professional flourescent fixtures should not have this problem but you would really have to test the particular fixture to see if the problem comes up with higher shutter speeds since you can't really measure the frequency of the light fixture so you still could end up with a phase issue which takes several seconds to resolve. I have been in situations where flourescent fixtures flickered a lot and others where there was no problem.