Night Shooting - The Steel Gel

[drcheese]

Greetings all,

I've been doing some more nighttime lighting recently, mostly interiors and mostly at actual night time. Still, for a lot of this I want to boost up that moonlight look and I end up using a steel gel, usually doubled up, over a tungsten light shot through a window.

I like the look and think it represents night pretty well. But it seems like a more popular gel to use is CTB, and I was wondering what people's thoughts are out there. Do you have a preference and why? Is the steel gel even part of your kit?

Additionally I just entered the world of HMI with the purchase of my first 1.2kw. It'll be my most powerful light, so I can see wanting to use it as a key, but it's already daylight balanced. So I was wondering what people do to give an HMI that nighttime look. Perhaps just a single layer of steel gel?

 

[Marc Villafan]

Greetings all,

I've been doing some more nighttime lighting recently, mostly interiors and mostly at actual night time. Still, for a lot of this I want to boost up that moonlight look and I end up using a steel gel, usually doubled up, over a tungsten light shot through a window.

I like the look and think it represents night pretty well. But it seems like a more popular gel to use is CTB, and I was wondering what people's thoughts are out there. Do you have a preference and why? Is the steel gel even part of your kit?

Additionally I just entered the world of HMI with the purchase of my first 1.2kw. It'll be my most powerful light, so I can see wanting to use it as a key, but it's already daylight balanced. So I was wondering what people do to give an HMI that nighttime look. Perhaps just a single layer of steel gel?

If you want the exact same look of your Tungsten moon light, you should use a color meter.

Calculate the mired value of your Tungsten light with the gels you use on it. To do so divide 1,000,000 by your kelvin reading.

Assume you you have a 5200 kelvin reading on your tungsten filtered light:

1,000,000
5200 Kelvin=~192.3mireds

Now lets say your HMI has a kelvin reading of 3600.
1,000,000
3600 Kelvin=~277.7mireds

Subtract your tungsten light with your HMI mired calculation.
192.3 - 277.7= -85mired shift.

Now go to your favorite gel filter company and look for a gel that has the same mired shift as the result of your calculation, which in this example, happens to be a mired shift of -85.

I believe the result should be extremely close, color temperature wise to your tungsten when you gel your HMI with the calculated mired shift gel. So now just go plug in your own numbers and tell us how it works out.

For more information, checkout Ryan's color temperature article, here.

 

[drcheese]

Thanks for the link the article, looking at things through mired values is a new interesting way to see things.

A couple questions/comments on the math though, in case I'm confused:
- My HMI would be ~5600K not 3600K right?
- Additionally, my tungsten light is 3600k base, then with a CTB it might be close to 5200 or like that, but a Steel gel is more akin to a party gel than for use with adjusting color temperature. So I don't think it has quite the same affect on color temperature? I've never really looked much at party gels for color balancing.
- But, in which case I would have to find the mired shift of a steel gel (which I cannot seem to find, but is likely less powerful that a full CTB), but then that leads me to believe it will be some version of CTO (Straw? Amber?) that would lead the same look? Hmm...

Additionally, going away from the technical, what do people think of the quality of Steel as a moonlight look?

 

[Ryan Patrick O'Hara]

^^ Tungsten is generally 3200k :tongue:

 

[drcheese]

Whoops, swear I knew that All these 6s and 2s flying around!

Ryan, since you wrote the article maybe you can shed some light, uh, so to speak... how would party gels, and the steel in particular, affect color temperature?

 

[Marc Villafan]

^^ Tungsten is generally 3200k :tongue:

Hey hey expert, I just picked any random Kelvin number to give the example, than went back and labeled tungsten, HMI so he can plug his own variables in. :happy:. Haha. I was just typing that maybe we can get you to weigh in. Do party gels have a different affect on color than CT gels?

To Dr.Cheese, I believe the easiest and most accurate way would be to use a color temperature meter.
Your HMI can vary greatly, specifically with the age of the bulb. But a bulb that has burned a few hours I believe should be around 5600k-6000k. Don't forget to allow them

 

[Ryan Patrick O'Hara]

Thanks for the link the article, looking at things through mired values is a new interesting way to see things.

A couple questions/comments on the math though, in case I'm confused:
- My HMI would be ~5600K not 3600K right?
- Additionally, my tungsten light is 3600k base, then with a CTB it might be close to 5200 or like that, but a Steel gel is more akin to a party gel than for use with adjusting color temperature. So I don't think it has quite the same affect on color temperature? I've never really looked much at party gels for color balancing.
- But, in which case I would have to find the mired shift of a steel gel (which I cannot seem to find, but is likely less powerful that a full CTB), but then that leads me to believe it will be some version of CTO (Straw? Amber?) that would lead the same look? Hmm...

Additionally, going away from the technical, what do people think of the quality of Steel as a moonlight look?

Mired shift calculations are only really applicable for color temperature gels... like CTO, CTB, CTS, etc. All gels (with the exception of ND gels) change color temperature in some way... but because they change more than simply the color temperature 'blue/orange' spectrum, the mired values aren't very applicable.

For instance your steel gel ( Roscoe or Lee) does not have a manufacturer rated mired value because it is not a strictly rated. In this case a color meter would yeild the best results and identify what the gel is doing to the color. You can of course make the choice based on eye or monitor... but it won't be as exact. If you have a good trained eye, you can do it pretty easily.

 

[drcheese]

"to allow them... " ack the anticipation! Although I think you were about to say "allow them to warm up for a while before taking the reading".

The post makes a lot more sense now, all of the tungsten 5200k, HMI 3600k was a bit eyebrow raising. I thought maybe you knew that the Steel gel would change it to 5200k.

Overall, I wonder if this falls more under the realm of theatrical lighting (as in lighting for an effect) as opposed to color correcting lighting. Without knowing how Steel affects color temperature I would think the rough color correcting way to to get the HMI to match the previous look would be to use a full CTO and then put a steel over that, though that seems silly. Or of course to buy a color meter

Alternatively does anyone have a nice HMI/gel combo they like for simulating moonlight?

 

[Ryan Patrick O'Hara]

Hey hey expert, I just picked any random Kelvin number to give the example, than went back and labeled tungsten, HMI so he can plug his own variables in. :happy:. Haha. I was just typing that maybe we can get you to weigh in. Do party gels have a different affect on color than CT gels?

To Dr.Cheese, I believe the easiest and most accurate way would be to use a color temperature meter.
Your HMI can vary greatly, specifically with the age of the bulb. But a bulb that has burned a few hours I believe should be around 5600k-6000k. Don't forget to allow them

Hahaha, I see what you did. I was just throwing that out there because it seemed that number was starting to get used as tungsten light ct. Ya know?

Overall, I wonder if this falls more under the realm of theatrical lighting (as in lighting for an effect) as opposed to color correcting lighting. Without knowing how Steel affects color temperature I would think the rough color correcting way to to get the HMI to match the previous look would be to use a full CTO and then put a steel over that, though that seems silly. Or of course to buy a color meter

Alternatively does anyone have a nice HMI/gel combo they like for simulating moonlight?

Not sure it needs to fall under one or the other. Both use it for effect. I imagine the difference would be theater would be more exaggerated while film would be more subtle. Then again, a film can have exaggerated color lighting and stage can be subtle... so hey.

What is also worth mentioning is what is the camera white balance set at? If the camera whitebalance is set to match the HMI color temp, then the blue steel gel on the HMI might be a great approach. Then you don't have to take the HMI down to tungsten level and back to steel blue gel. (Double geling like that is funny but not the right way to go about it. )

If the camera whitebalance is tungsten or about 3200*k you are simply making blue even 'bluer' with the steel gel. It depends on the environment. If you are lighting other parts of the scene with tungsten units, you generally won't want to CTB the units to match the HMI because you'll lose almost 2 stops of light! So in that case I'd find a CTO or party gel which would take the HMI down to the level of moonlight blue you want. It really depends on the situation.

 

[Marc Villafan]

Hahaha, I see what you did. I was just throwing that out there because it seemed that number was starting to get used as tungsten light ct. Ya know?



Not sure it needs to fall under one or the other. Both use it for effect. I imagine the difference would be theater would be more exaggerated while film would be more subtle. Then again, a film can have exaggerated color lighting and stage can be subtle... so hey.

What is also worth mentioning is what is the camera white balance set at? If the camera whitebalance is set to match the HMI color temp, then the blue steel gel on the HMI might be a great approach. Then you don't have to take the HMI down to tungsten level and back to steel blue gel. (Double geling like that is funny but not the right way to go about it. )

If the camera whitebalance is tungsten or about 3200*k you are simply making blue even 'bluer' with the steel gel. It depends on the environment. If you are lighting other parts of the scene with tungsten units, you generally won't want to CTB the units to match the HMI because you'll lose almost 2 stops of light! So in that case I'd find a CTO or party gel which would take the HMI down to the level of moonlight blue you want. It really depends on the situation.

So if you have a Color Temperature Meter and calculate the mireds of a party gelled light. Than you look for a Color Temperature gel that would match the mired shift. Would the color of the light end up the same? Regardless whether it was a party gel or not? I would think now that you say they don't alter the same blue/orange spectrum as CT gels that the result wouldn't end up being the same.
But just to be sure, can you clarify this?

 

[Ryan Patrick O'Hara]

I guess my point was that those gels are not made to change color temperature so they are not given a manufacturer mired value... but we know that all colored gels will effect the color of the light...

My point is that CT gels are made to only alter the color temperature of the light... nothing else. There are other types of blues (think cyan, teal, plum etc) and oranges (orange-red, yellow orange, etc) which can change the correlated color temperature of a light on the kelvin scale but also alter the color of the light by changing the green, magenta, and etc values.

So my point is you could probably find a blue-teal gel that would lower a color temperature of the kelvin scale down the same as a stregth of CTB, say 1/2, but if you were to balance your camera to that color temperature, one light would look 'white' while the other would look 'greenish'.

To answer your hypothetical question you can use a non CT gel to correct color temperature, but they are not made to do so, and the color can be changed in other ways as a result. I've used a thick blue party gel as CTB before. It worked, and the results were not horrible. That is why I said to use the steel with a light meter to see what it does. Some light meters can give magenta/green info as well, so you can get a more exact idea outside of mireds on how the color is being effected. For most non-CT effects, an experienced eye can really go far.

 

[drcheese]

Not sure it needs to fall under one or the other. Both use it for effect. I imagine the difference would be theater would be more exaggerated while film would be more subtle. Then again, a film can have exaggerated color lighting and stage can be subtle... so hey.

Oops, I didn't mean to imply that it's only used in theatre not film, but more that it's used more to achieve a stylistic (theatrical) look as opposed to match technical color temperatures. Although of course CTB is used stylistically as well, and you can use party gels to color correct in a pinch. So maybe the delineation still doesn't work.

What is also worth mentioning is what is the camera white balance set at? If the camera whitebalance is set to match the HMI color temp, then the blue steel gel on the HMI might be a great approach. Then you don't have to take the HMI down to tungsten level and back to steel blue gel. (Double geling like that is funny but not the right way to go about it. )

If the camera whitebalance is tungsten or about 3200*k you are simply making blue even 'bluer' with the steel gel. It depends on the environment. If you are lighting other parts of the scene with tungsten units, you generally won't want to CTB the units to match the HMI because you'll lose almost 2 stops of light! So in that case I'd find a CTO or party gel which would take the HMI down to the level of moonlight blue you want. It really depends on the situation.

I was thinking it as camera whitebalanced 3200k, primarily because of the 2 stop CTB punishment (as compared to 2/3 stop with CTO) and that most other lights I have are Tungsten. However, I do have some daylight bulbs in a cool light, so that's a great option to keep in mind.

It sounds like a fun experiment to see what gel + an HMI most approximates a Tungsten + Steel gel. Since we're talking about different spectrums as well, I wonder if tossing a Salmon on the HMI would orange it up enough while adjusting the other spectrums in the right direction. I'd think more so than an Amber, but then again...

 

[Ryan Patrick O'Hara]

It sounds like a fun experiment to see what gel + an HMI most approximates a Tungsten + Steel gel. Since we're talking about different spectrums as well, I wonder if tossing a Salmon on the HMI would orange it up enough while adjusting the other spectrums in the right direction. I'd think more so than an Amber, but then again...

I think you are completely on the right track.

 

[Samm_Parnell]

i use 1 1/2 ctb with opel or full white diffusion for my key. it is blue but if you are balanced for it it looks really good, of course you lose stops, i think its 3.5 - 4 with that much correction but with a good 12k tung or 9 lite maxi brute it looks fine and has the power to be placed back far enough for good angles on the light too.

with an hmi maybe 1/2 ctb correction with opel or full white, once again depending on the wattage of the lamp,

samm