Tungsten bulbs have to warm up to display correct temperature?

[Ronald Gerald Smith]

I am wondering if it takes tungsten bulbs (mogul bulbs and household bulbs) some time after they have been turned on to display accurate color temperature.

 

I am using a color meter to measure the Kelvin of these bulbs. I bought all of these bulbs at film tools. Here are the findings of my tests:

 

500 watts (USHIO JT120V-500WB Globe) test bulb A

Bare bulb - 2880-2810 (2850) Bare bulb is off by 350K.

 

I added some diffusion with the chimera video pro small size.

chimera single interior screen (interior baffle) - 2780-2750 (2760)

chimera single interior screen and outer screen screen - 2730-2690 (2700)

chimera single interior screen, outer screen with CTB - 4450-4300 (4400) (1G CC) some magenta

 

Ushio PH-212 150W Bulb (for china ball, household socket)

 

Bare bulb - 2850-2740 (2800) advertised as 3050. - bare bulb is off by 250 K.

 

Ushio JT120V 1000WB - supposed to be 3050L.

 

Bare bulb 2850 - 2790 (2830) - Off by 230K.

 

 

These bulbs have seen some moderate use - I shot a short film using these bulbs, but they should still have some life to them. Do these bulbs get warmer over use? And do they dramatically shift color temperature like that? Like 200-300 degrees K?

 

Also, maybe I am doing the test wrong and I should let the bulbs be on for like around 10 minutes and then meter it. What do you think? If I leave the bulb on for about 10-15 minutes will they have cooled down in degrees Kelvin? What is your expeerience?

Edited March 5, 2011 by Ronald Gerald Smith

[robert duke]

I am wondering if it takes tungsten bulbs (mogul bulbs and household bulbs) some time after they have been turned on to display accurate color temperature.

 

I am using a color meter to measure the Kelvin of these bulbs. I bought all of these bulbs at film tools. Here are the findings of my tests:

 

500 watts (USHIO JT120V-500WB Globe) test bulb A

Bare bulb - 2880-2810 (2850) Bare bulb is off by 350K.

 

I added some diffusion with the chimera video pro small size.

chimera single interior screen (interior baffle) - 2780-2750 (2760)

chimera single interior screen and outer screen screen - 2730-2690 (2700)

chimera single interior screen, outer screen with CTB - 4450-4300 (4400) (1G CC) some magenta

 

Ushio PH-212 150W Bulb (for china ball, household socket)

 

Bare bulb - 2850-2740 (2800) advertised as 3050. - bare bulb is off by 250 K.

 

Ushio JT120V 1000WB - supposed to be 3050L.

 

 

you might meter your power, you make be under powering the lamps.

 

Bare bulb 2850 - 2790 (2830) - Off by 230K.

 

 

These bulbs have seen some moderate use - I shot a short film using these bulbs, but they should still have some life to them. Do these bulbs get warmer over use? And do they dramatically shift color temperature like that? Like 200-300 degrees K?

 

Also, maybe I am doing the test wrong and I should let the bulbs be on for like around 10 minutes and then meter it. What do you think? If I leave the bulb on for about 10-15 minutes will they have cooled down in degrees Kelvin? What is your expeerience?

[M Joel W]

It might be the power. And if you're using thin gauge wires, particularly long ones, you can expect some big drops in color temperature. I've never heard of tungsten bulbs changing color temperature over time, though. I've also noticed tungsten lights seem very warm relative to, say 3200K kinoflos, so it may just be that.

[John Brawley]

I am wondering if it takes tungsten bulbs (mogul bulbs and household bulbs) some time after they have been turned on to display accurate color temperature.

 

 

You'll rarely get 3200K from a tungsten lamp. There are so many things that can affect the CT, but voltage would be the main one. Once you put the lamp into any kind of housing then you'll tend to get a drop as well as it takes on the colour of the reflector.

 

In the same way, no HMI is 5600 K either.

 

These numbers are just approximations but you'll get to know from experience that it's rare to get a lamp that will actually hit those numbers without correction.

 

jb

[Ronald Gerald Smith]

Hi John, you're explanation does make sense. It does kind make me wonder though why some bulbs are closer to the advertised temperature than others. For example, the ECA 250w 3200K Tungsten bulb is metered at 3100K, which is only 100 degrees K off, whereas the other lights are around 300K or more off.

[John Brawley]

Hi John, you're explanation does make sense. It does kind make me wonder though why some bulbs are closer to the advertised temperature than others. For example, the ECA 250w 3200K Tungsten bulb is metered at 3100K, which is only 100 degrees K off, whereas the other lights are around 300K or more off.

 

100K is barely noticeable in the scheme of things. You'll find that's the finest increment you can manually dust WB in most cameras.

 

I find most lamps are somewhere between 2800 and 3000K.

 

jb

[John Sprung]

Tungsten lamps come up to a stable color temperature in just a few hundredths to tenths of a second, depending on their size. The variations you're seeing depend on other factors, mainly the voltage at the lamp. Try measuring the voltage at the end of your stinger with the lamp off, and with it on. You should detect at least a few tenths of a volt drop when you turn on something of 500 Watts or more. You don't even need a movie light, a toaster will work as a test load. Of course you'll need a digital voltmeter, which you should own if you're going to be gaffing.

 

 

 

 

-- J.S.

[Ronald Gerald Smith]

Tungsten lamps come up to a stable color temperature in just a few hundredths to tenths of a second, depending on their size. The variations you're seeing depend on other factors, mainly the voltage at the lamp. Try measuring the voltage at the end of your stinger with the lamp off, and with it on. You should detect at least a few tenths of a volt drop when you turn on something of 500 Watts or more. You don't even need a movie light, a toaster will work as a test load. Of course you'll need a digital voltmeter, which you should own if you're going to be gaffing.

 

Thanks John, that makes a lot of sense and both comforting and unnerving to know that this is a common thing. Now, the question is, is there any way that I can go about getting to that 3200K (or marketed color temperature) with a lamp above 500 watts? Maybe with better stingers?

 

It's just been annoying recently because the bulbs that I bought (that are supposed to be 3200K but are more like 2800) are obviously warmer than the mole 1k, 650, and mini mole at the studio where I shoot. I believe that those mole lights are very very close to 3200K because I get consistent whites and good skin tone/saturation levels, whereas the bulbs that I bought (USHIO 500W-1000W) show a very very warm image and make my talent looks like orange people. Might it be something with the Mogul base bulbs? Are they prone to such drop in color temperature? Should I stay away from USHIO as a brand or mogul base bulbs in general?

 

Thank you all and forgive me if I am being redundant - I just want to get to the bottom of this.

[Hal Smith]

Thanks John, that makes a lot of sense and both comforting and unnerving to know that this is a common thing. Now, the question is, is there any way that I can go about getting to that 3200K (or marketed color temperature) with a lamp above 500 watts? Maybe with better stingers?

 

To add to John's thoughts: What voltage bulbs are you buying? Many bulbs come in slightly different voltages. For instance the HPL575/750's come in both 115 and 120 volts versions plus an "X" long life variety. If your power is close to 115 and your buying 120 volt bulbs you'll get a warmer color temperature. Worse yet, the "X" variety are really 130 volt (or so) bulbs, they get their long life by running the filaments cooler.

 

As John recommended, get a digital voltmeter and check the voltage at the lamp under load. When you run lamps off a dimmer, you MUST use a true RMS rated voltmeter to get an accurate reading due to the distortion of the power waveform caused by a dimmer. The only case where that's not true is variable transformer (Variac) or ancient resistor plate type dimmers.

[Adrian Sierkowski]

Also the easiest way to correct such bulbs which are too warm would be just to throw some CTB on them..

[Stuart Brereton]

Even 1/4 CTB corrects by about 600 degrees kelvin. For the variances that Ronald is talking about it's simply not worth bothering.

[Adrian Sierkowski]

I was referring to 1/8th though Stuart ;) I always have some on hand, for moments where something is too warm.

[Stuart Brereton]

it's a matter of taste, obviously, but I've never used 1/8 ctb. Differences of a few hundred kelvin are hardly worth worrying about.

[Adrian Sierkowski]

It really is a matter of taste. I only ever pull out the 1/8th when it doesn't feel "right," though personally I like "warm" tungsten lights -v- anything daylight balanced.

[K Borowski]

Just want to add that, IDK what the color temperature is, but theree is a recommendation with tungsten lamps, similar to earlier fluoros, that you want to wait say 15-20 minutes for the filament to warm up.

 

IDK what the temperature difference is, but, it is more of a fluctuation than anything else with a lamp. Also, as a lamp ages, carbon deposits will change its temperature. No two lamps have the same exact temperature, but here we are talking about single CC color corrections. Not sure how many mireds that is.

 

 

You need to quit throwing around 600K differences as if that number MEANS anything. 600K is a very big difference going from 2400K to 3200K. It is practically nothing going from 5600 to 6200K.

 

Please use proper terminology (MIREDS are micro reciprocal [1/1,000,000] degrees, and they were invented specifically for this use).

 

 

Also, depending on the gamma of the film, contrast of the lens, color balance of the lens, there are other variables in the system that need to be taken into account. I've heard advice to use a correction factor with reversal film as opposed to neg. stock. I assume that something similar will apply depending on digital versus film.

 

There are a lot of complicated variables to consider here. Remember there are color shifts from emulsion batch to emulsion batch, processing tolerances +/-1/6 to 1/4 of a stop with ECN-2, and monitor drifts, tolerance drift all up and down the quality control ladder. . .

 

Good luck dealing with CTB corrections in a 5- or 6 light setup, where each lamp has a different temperature. Can't change the gel on the camera for all of them!

 

 

 

Even with the tightest controls on set, and a perfectly balanced digital camera, or perfectly accounted for film emulsion laboratory/printer setup, you'll not be able to get a perfect 25-25-25 printer light dealing with several different tungsten bulbs.

 

Not an expert on fluoros, but I'm not sure if they are more accurate or not. I'm sure there are still rather large tolerances depending on how the mercury powder is coated on the tubes and the spectral transmittance of hte glass.

[Stuart Brereton]

Just want to add that, IDK what the color temperature is, but theree is a recommendation with tungsten lamps, similar to earlier fluoros, that you want to wait say 15-20 minutes for the filament to warm up.

 

You need to quit throwing around 600K differences as if that number MEANS anything. 600K is a very big difference going from 2400K to 3200K. It is practically nothing going from 5600 to 6200K.

 

Please use proper terminology (MIREDS are micro reciprocal [1/1,000,000] degrees, and they were invented specifically for this use).

 

 

Never heard of any recommendation to wait 15 - 20 minutes for tungsten to warm, not in nearly 20 years of using them. Sounds like internet BS to me.

 

As far as throwing around 600k differences goes, we all have our own ways of doing things. Yours is obviously to be as rude, condescending and aggravating as possible.

[Ronald Gerald Smith]

It's nice having warmer lights, but that's the whole reason why I buy 2800K bulbs for that particular reason. But if I am buying a 3200K bulb, I kind of expect it to match my other tungsten lights.

 

Also, going from 2800K to 3200k is -44 mired shift according to lee calculator. Even with a 1/4 ctb you still wouldn't get the preferred temperature - you'd still be -9 mired short. And plus 1/4 ctb kills half a stop! I use smaller lights most of the time for power situations so I can't afford to lose even half a stop. I know that I am nitpicking but I am sick of having to turn the saturation levels down just because of these bulbs!! Arg! Anyways it's nice to see varying viewpoints on this matter and I am learning a lot from all your comments.

[John Sprung]

but theree is a recommendation with tungsten lamps, similar to earlier fluoros, that you want to wait say 15-20 minutes for the filament to warm up.

 

IDK what the temperature difference is, but, it is more of a fluctuation than anything else with a lamp. Also, as a lamp ages, carbon deposits will change its temperature. .

 

Never heard of waiting for tungsten to warm up. Flip the switch, and it's as warm as it's gonna get before you can take two steps away from it. With ordinary tungsten argon borosilicate globes, you do eventually get darkening from tungsten, not carbon, depositing on the inside. The tungsten evaporates off of the filament. The reason for tungsten halogen quartz is that the globe can be run so hot that the tungsten evaporates back off the inside, keeping it clean. That's why you want to run halogens at full brightness for a couple minutes after having them dimmed way down for a while.

 

 

 

 

-- J.S.