5600K and 5500K lights creating a white balance disaster

[Natan Pakman]

I'm trying to figure out what's going on in a studio I work at, and in the process learn a bit about color temperature. The key light we are using on the host is a Coollights 655 fluorescent soft light, with tubes supposedly at 5600K. We are using a reflector for a fill light, and as a backlight we are using a Lowel Omni with a daylight gel. So, I am assuming the light exiting from the gel on the Omni is 5500K. First of all, is this correct, or could it be lower than 5500K?

The issue is that there is a noticeable color difference on the subject when turning the Omni on and off and it is creating a white balance issue with camera two (on a different talent), which doesn't use a daylight-gelled tungsten light (When I white balance the camera on the main subject with the omni on, the shot becomes colder than that from camera 2).

More than anything, I'm just wondering if a 100K difference can be so noticeable, or if the extent that this is noticeable indicates that it's a more than 100K difference. Also, could a green spike from the Coollights affect this or is that unlikely?

Excuse the newbieness.

[Natan Pakman]

Related to this, would buying multiple different CTB gels and playing with them to match the Coollight be logical?

[Charles Papert]

White balancing to a key shouldn't be affected by a backlight unless you are angling the card in such a way as to include it, for one.

It sounds like the Coollight isn't putting out as cool a color temp as the daylight gelled omni. A color temp meter would tell you if this is the case, and maybe your camera as well if it happens to have a kelvin readout. So yes, trying different flavors of CTB on the Omni would help--maybe 3/4 CTB would do it.

There is undoubtedly going to be some green/magenta shift going on as well, and that may be complicating the issue.

Screen grabs would also be helpful.

[Garrett Low]

What other lights do you have on in the studio? You say that you see a difference when turning no and off the back light. It sounds like the back light is spilling over onto your other subject which is only lit with tungsten. Why not gel the lights on subject 2 so that all of your lights will be around 5500K?

[Natan Pakman]

Charles, I am white balancing the camera on subject 1 with the backlight turned off. When I turn on the backlight, it WARMS the image up. So, I figure that the Coollight is COOLER than the Omni with the daylight gel. Since the Coollight has 5600K tubes, I figure the daylight-gelled Omni is at a lower K than that. I am thinking a full CTB in addition to an 1/8 or something would be closer to the ballpark, as I need to cool it even more. Is this type of gel-mixing done?

Garrett, subject two is lit only with daylight flos. There are no tungsten lights on him at all. The problem is that the white balance I get on camera 1 matches camera 2 when camera 1 has the backlight turned off (and we need it on!), and when I white balance camera 1 with the backlight on I do not match camera 2 (in this case, camera 1 is colder than camera 2).

I didn't want to complicated this discussion (at least not yet!), but there is a further problem here, which is that the flo on subject 2 is actually at 5200K with a white diffuser, not 5600K like the flo on subject 1. However, this is a minor point for this specific problem, given that without the backlight on subject one, camera 1 and 2 are pretty well-matched for color.

I'll post some pics of the situation later today.

[Charles Papert]

OK yes I get it--when you white balance with the omni included, you end up cooler because the omni is putting out a warmer color temp.

Again--backlights shouldn't be part of the white balance process. A card facing the camera should be only "seeing" frontal lighting. Not sure how your backlight is set but it would normally be blocked by the card, unless the card is tilted back enough for the backlight to spill on it.

More importantly: all that counts is what the camera sees. There's no rule (or they shouldn't be) that all of your lights need to be the same color temperature. I regularly use warmer backlights, especially on blondes. If it looks good on the subject, what's the issue?

How's Noho? Miss it!

[Natan Pakman]

Charles, in the end what I think is going on is that the backlight is blatantly spilling onto the subject's face. The solution then, of course, is to stop that from happening using the barn doors and better angling! I never white-balance with the backlight on, but the issue always occurs when we turn it on.

Separately from this there is a possible adjustment that needs to be made to the video switcher we use (Tricaster) to correct differences between the two cameras, but that is a whole other story.

Northampton, is, well, the same! Are you from here?

[Charles Papert]

Let's see stills then Natan. If it is intended to be a hairlight, sounds like it isn't mounted in the right place, or too strong.

I lived in Northampton from around '88 to '93. Played in bands, knocked around. Still have a bunch of friends there. I visit once every few years. Many great memories.

[Paul R Johnson]

100K from similar sources is not even noticeable, within the tolerances of the light sources. The issue I suspect is that flus, LEDs and Tungstens have different colour characteristics even when a meter tells you they are the same colour temp. The tungsten lamp has a gentle curve to the colour spectrum, and CT filter converts it as expected. The flu tubes don't have such a smooth curve, and LEDs seem very mountain range looking in their colour spectrum - that can be substantially different from an identical unit bought 6 months later! Lee Filter, here in the UK have brought out special colour filters to produce their popular colours from LED white light - because the tungsten versions don't produce anything like the right colour.

Only this week I have spent ages trying to match two discharge lamps - one very straw coloured, the other very blue. So we'd blue the straw and orangify the blue - having some half and full CT laying around. In the end using multiple layers of filters we got a match, BUT lost over half the light! On paper we had 5600 and 6000 colour temps - but the 6000 one already had in built filters to convert to 5600 and 3200, but when set to 5600 they were not a match - nowhere near!

[Chris Duczynski]

Can't imagine 100K would make any difference.
Youre not balancing on A or B with Auto WB on are you ? Because that will shift when you switch on the backlight.
Thats all I can think of.

[Daniel Epstein]

I think Paul is on the right track about this. Sounds like all your sources have different spectrums which are adding up to trouble white balancing. Also if you use different sources you end up with different balances on the cameras for each position. Using blue color correction gel on a tungsten light does not match a 5500k source over the full spectrum. It may be close enough but won't be perfect. Often ambient spill will cause problems in the shadows. Also color temperature numbers are a guide in this case but you can have different looks from AWB that have the same numbers depending on how the cameras are set up and the sources being used.. The cameras numbers are not always the same from camera to camera. As far as 100 degress being noticeable I think it can be. If you are in the tungsten end of the spectrum 100 degrees is significant. Above 5600K it is less of a problem. On a single camera shoot it is easier to mix things up. More problematic with multiple cameras.

[Bill Pryor]

CTB gel comes in 1/4, 1/2, 3/4 and full. Could it be you're not using full CTB on your backlight?

[Nino Giannotti]

One of the first thing I teach in my workshops is to NEVER mix light sources on the subject.

White balance each light individually and see what K reading you're getting.

All light sources, tungsten included, change color with age. If you use a dimmer it will change color.

Use different sources if you have to on the background, making sure of course that they do not spill into the subject.

[Natan Pakman]

The white balance situation has been resolved using a different-colored light that more closely matches the 5600K fluorescent coollight light. However, I have a question regarding bulbs. I have an interfit light (http://www.bhphotovideo.com/c/product/569816-REG/Interfit_INT216_Super_Cool_lite_9_Fluorescent.html), and might I add that you are free to bash its quality but it has served us nicely for interviews, and I was thinking about changing its 5200K bulbs to 5600K ones to match our fluorescent fixtures and use it as a fill light.

First, has anyone successfully changed CFL's to match the color temp of a fluorescent tube light fixture?

Second, I'm very uncertain as to the quality of the bulbs I am finding online. Here is one I found:

Amazon.com: ALZO 27W CFL Video-Lux® Photo Light Bulb 5600K - pack of 4 - matches color temperature of Kino Flo & Osram bulbs - daylight balanced - pure white light - 1300 Lumens: Camera & Photo

 

Amazon.com: ALZO 27W CFL Video-Lux® Photo Light Bulb 5600K - pack of 4 - matches color temperature of Kino Flo & Osram bulbs - daylight balanced - pure white light - 1300 Lumens: Camera & Photo

The price is excellent, but I'm unsure as to the quality. The CRI is not listed although another similar bulb from the same brand claims to have a CRI of 91. Does anyone have any recommendations for this or other brands of bulbs?

[Natan Pakman]

I've attached a still from today's show, comparing camera 1 and camera 2 after numerous adjustments to the lighting and the proc amp controls on our video switcher (Tricaster Pro). How does the balance look between the two cameras? Any other observations?

Just a note on Camera 2 on the right, we are waiting for an additional bulb to come in on his fill light, which will remove some of the shadow on his right cheek and shoulder.