Anyone good at calculating color temps?

[Dansk]

I have some very odd ambient lighting that I have to balance some strobes up with and I'm scratching my head a little over it as I have never quite seen this temp before. I run Phase systems and profoto strobes with all sorts of modifiers and my default wb is 5400K with +4 magenta. I did a site survey at the shoot location a week back and its a huge warehouse full of the equipment and people I have to shoot so I have to soak up some ambient just due to the sheer size of this place I'd be adding ten zillion lights to cover this place if I wanted to use all my own lighting.

  Anyways I did take a few grabs of the place with a gray card on the survey and ran them through Phase just now and the temp comes up at 2900K with +42 magenta. I would add a full CTO and call it done but this +42 magenta is confusing me so what gel/s would you recommend to start with here? Main problems I have are 1. I will need to buy more gels than I currently have and would rather avoid buying 200 gels that i will never use again and 2. I dont have the time on set to play around too much with this so I'd like to be as prepped as I can be on the day.

 For what its worth I'll need enough to  cover 8 strobes. The ambient lighting in question is new and it is rectangular and appears to be some sort of the new Metal Halide or sodium but this is one type I have yet to see before. I have asked the client if they can find out the exact brand and model of lights they are in hopes to dig up tech specs on them but as of yet I have no info on them.


I should add I'm not about to start printing my own gels off i have read about that but i just dont have the means nor time to do so right now. Anyone have thoughts I'd appreciate it. Thanks guys

[Craig Murphy]

You think these lights are pushing out magenta?  If so it sounds like you need a full and a half minus green (magenta) gel on each flash unit.  They will also lower color temp so you probably will not need the full CTO.  Talk about sucking up light power.

[Dansk]

Well thats the confusing part It doesnt look like magenta but the software is telling me otherwise. Have a look for yourselves heres a couple of snaps of my assistant. The off color one was shot using wb of 5400K on a 1Ds which matches up with my Phase backs exactly I can swap systems mid shoot without having to pull out the gray card using that wb pre set when using my strobes.

So in this case when I load the captures into C1 Phase and set the gray balance as shown in proof 2 which is the same shot just balanced it gives me a temp reading of 2900K and +42 magenta which is way off my strobes. I wish I could just play around with some gels there and see what looks best. I guess I'll grab up a bunch of minusgreen and hope that'll do it. You dont think a combination of the CTO and minusgreen would be better? I have to confess Im no expert at balancing color temps i've always gritted my teeth at this process so thanks for the help

[Craig Murphy]

Looks to me like they are pushing out green which would be more likely.  In that case you would need plus 45 green on your strobes.  I only meant that the green gels will lower your color temp some also so you probably would not need a full cto.  Once you do a white balance with the gels on the strobes why would you put CTO's on unless you wanted to put them on to warm up a subject some?

[Dansk]

Makes sense. I dont have 45 plusgreen in my swatch book ( Roscoe ) I have the same rating as CTO so they are 1/4, 1/2 plusgreen etc. so a shot in the dark here at full plusgreen? I'll probably take a few variations and some CTO's as well in case I feel I need em. I really dont use gels much at all and typically steer way cleer of them and add desired effects in post but on this gig I have no choice.

 I really have no idea why this is always so damn tricky for my mind to work out these gels haha. I guess everyone has their weaknesses maybe if I used them more I'd be alright but its experiences like this that shy me away  

[Dansk]

I just spoke to the maintenance guy from the building and the lights are fluorescents with chrome fixtures... They are odd size though I am unfamiliar with this style and I wonder what the tint to the tubes were but the guy had no idea as they have yet to replace one because the building is brand new. Anyways I suppose I'll start with window green gels and pack the CTO's as a safety and hope for the best.

Thanks again Craig

[Tim Lookingbill]

Not sure if this would help. I'm not a high end photographer as you seem to be. Just a hobbyist who likes to tinker and experiment with issues like this and digital imaging in general.

I started playing around with my Pentax K100D's manual white balance with pretty much the same fluorescents as yours but mine were compacts in my apartment. The 3 default fluorescent settings in the menu system for shooting jpegs were pretty much giving the same heavy yellow renderings as in your first image. The camera just couldn't balance them properly. Going manual it kept over compensating injecting too much blue. I was going for a natural realistic rendering that I saw with my eyes.

I studied the color renderings of how the camera was compensating as a clue as to how it saw things, so I found a combination of apposing colors in colored objects around me and set the WB metering to spot within the camera's menu system and sampled these colors by framing them within this metering area. It required a green object, the white wall and a baby blue object which tricked the camera into giving the results I wanted. See the sample below.

I realize these aren't colored gels but I thought the color combo might gives some insight as to what you could rig in your situation. Your second shot looks fine but I'm not sure if that's the result you want from what you've posted. I'm not familiar with strobes, scanbacks, Phase systems and gels.

[attachment=6620:attachment]

[Tim Lookingbill]

Here's a more detailed demo of the subtleties that can be achieved by shifting the color ratios within the spot metered sampling point.[attachment=6622:attachment]

[Dansk]

Thanks for the help tlooknbill you certainly put effort into that post and its appreciated.

 The issues I'm having is mixing two different light temperature sources not white balance. I can balance either one as you see in the second pic ( which is actually the exact same pic ) thats not my isssue. The problem is that the house lighting is not very flattering for my needs so I have to add a bunch of artificial lights to the scenes I'll be shooting. So if I balance out to the house lights as shown my lights would look purple and if I balance out to my lights the house lights will look yellow ( as shown in pic 1 ) so its the mixing of the two sources with different color temps is the problem not balancing the one. There are solutions to solve this in post but that doesnt go over very well when you have clients and art directors peering over your shoulder expecting to see shots show up on the monitor that look as expected.

  More than anything I was confused by these particular lights and the readings I received and needed a sounding board hence this post. I am still flying in the dark somewhat and will have to perform some trial and error on the day to get my strobes gel'd to a point that I am happy with.

[Tim Lookingbill]

Now I understand.

Well I hope you get it worked out.

I learned something new in this thread. What a hassle it is being a pro photographer. I never knew what ya'll have to go through.

[Dansk]

Haha yeah well... every job has its challenges and being a pro photographer certainly keeps you on your toes. Creating your own vision is enough of a task but creating someone elses? fun stuff  

[fnagy]

Now I understand.

Well I hope you get it worked out.

I learned something new in this thread. What a hassle it is being a pro photographer. I never knew what ya'll have to go through.
[a href=\"index.php?act=findpost&pid=196157\"][{POST_SNAPBACK}][/a]

Man, no offense meant, but balancing lights is a second year course, or first year of assisting, choose one.

I assume from your post that your lights were the mains and the house lights were accent or was it the other way around?  What source were you using,  simply balance your lights with a gel to match the existing lights What colour temp.? Ever heard of a colour meter, it's a must for interior shooting, digital or not,  but be careful that sunlight does not get in (gell the windows?)  or you gel existing lights to match sunlight and strobes and voila!

That's why a good shooter gets the big bucks?  Has it figured out before they shoot and not try to make up for things after,  yes it can be done post, but only after many hrs. in front of a computer as opposed to shooting or alternatively pay someone to do that, big coin!  Get it right in front og the lens not the computer screen.

[AlanG]

The OP seems to know what he is doing and is on the right path.  So my post might be more for others who need to do this.

I don't know of any way to look at the picture or measure it on the computer to get "exactly" the filtration you'll need to balance the strobes to the daylight.  But with a little testing and experience, one can come close.  So in this case it is clear that the strobes will have to be balanced close to a warm white fluorescent and possibly would need additional green to actually match.  But it isn't always possible or practical to "perfectly" filter the strobes to the ambient.  In my experience if the main area lit with the strobes is balanced accurately, and the overall area is even close, everything will be fine. It isn't likely that someone cares about extreme color accuracy on the walls in the back of a warehouse.  

Keep in mind that fluorescent lights have a discontinuous spectrum.  Often these lights have very strong spikes at specific frequencies.  So whatever readings you get for color temperature may not be accurate or even very meaningful.   And color meters have problems with this type of lighting as they can't do a spectral analysis.

A good way is to test it as you have done, come up with your best shot at filtration, and then try it.  So keep in mind that if the light has very strong spikes in green, when correcting for this by white balancing the camera, the light still might not put out enough of a full spectrum to make all colors look as good as they might look under a continuous spectrum of light. (e.g. Daylight, tungsten, flash.)  This is why you will never get a perfect correction of the existing lighting to match whatever filters you use on your lights.  But you can come close.

By the way, I find it is very easy to retouch a photo and selectively change the color in various regions.  You can do this by overlaying and retouching through  two layers each with a different color correction, or with the Photoshop color replacement brush which is a very powerful tool.

[AlanG]

I looked at your two test shots of the man holding the grey target.

Why not put them up on your computer and hold filters over the "corrected image" until the grey target (not the skin tones) matches the uncorrected one. This would be a good start in trying your filtration as it already would have taken in the entire color response of the lights-camera-conversion-display system.  I'm not 100% sure you can work backwards like this but I also can't see why it wouldn't get you close.

This is a modern day way of how I used to judge the correction needed in color printing by holding up various viewing filters over the print.  I think it could work.

[dwdallam]

Why not dial up your cameras temperature (Or do it in RAW) until you get a neutral image. Write that temp down, and that's what the lights are pushing. Then buy your gels for that temperature for your strobes?

On the other hand, if you can't dial up a custom WB that works, then the lights are a mixed combination. If that's the case, you'd have to turn off of block some of the lights, which preferably would be the lowest number with a different color spectrum.

Am I missing something here?

[AlanG]

Why not dial up your cameras temperature (Or do it in RAW) until you get a neutral image. Write that temp down, and that's what the lights are pushing. Then buy your gels for that temperature for your strobes?

On the other hand, if you can't dial up a custom WB that works, then the lights are a mixed combination. If that's the case, you'd have to turn off of block some of the lights, which preferably would be the lowest number with a different color spectrum.

Am I missing something here?
[{POST_SNAPBACK}][/a]

If only color correction were always so simple.

How do you buy filters for a specific color temperature?  Unfortunately they are not calculated that way.

And as for color meters, this will show you how complicated it is to use a color meter on even one brand of lights. Note: Minolta disclaims its meters’ Color Compensating (CC) data when reading fluorescent lamps and recommends film tests to determine correct filtration.  And that is when you need accurate measurement the most!

[a href=\"http://www.temple.edu/fma/referencelibrary/manuals/Kino_Flo_Lamp_Data.pdf]http://www.temple.edu/fma/referencelibrary...o_Lamp_Data.pdf[/url]

Part of the problem is that we have several different types of measurement and filtration systems going on:

Color temperature - this is a logarithmic scale. So a 500 degree change from 2500K to 3000K is a much greater change than from 5500K to 6000K. So there is no such thing as a filter that will correct "2500 degrees Kelvin" other than from one specific temperature to another.  Also, as I said above, color temperature values are not accurate when applied to discontinuous spectral sources such as fluorescent, sodium vapor, mercuray vapor, etc.  Then in addition to the color temperature (red/blue - warm to cool), these lights also require a green/magenta correction. That green/magenta correction is calculated using CC values which are linear not logarithmic.  

Color meters give values of Kelvin and some will show what conversion filters are required.  In addition three color meters give red/green values expressed as CC values. These readings are accurate in measuring continuous spectrum light - such as how much green thick window glass add to the daylight coming into a room.  But I've found them to be "accurate" only by chance when measuring discontinuous light.

So this brings us to Roscoe and Lee filters that are typically used over lights.  Guess what, they use their own system for marking them. (Not color temperature which wouldn't work, and not CC which is from the Kodak Wrattan filter system.) That is why this gets very confusing.

In this case the strobe light might need to be corrected to around 3000 degrees (close to tungsten) plus they need around a 40CC green  So maybe a Roscoe CTO will get it warm enough (specifically 5500K to 2900K), but which Roscoe filter is around 40CC green?  

http://www.rosco.com/us/filters/roscolux.asp

So what is the solution?  I used to do this type of work all of the time when I shot architecture in mixed and varied types of lights. I did a lot of testing and recorded that information.

Mired values:

There is one universal system for translating all of this and working in linear values.  It is called the Mired system. What I did is I took all of my lens filters and I wrote the Mired values on each one.  This is for color temperature conversion only - e.g. from daylight to tungsten or some warming or and cooling.  I also figured out the equivalent Mired values for the filters that I placed over the lights.  So while Roscoe marks their  Roscolux #3407: Roscosun CTO as something that converts 5500K to 2900K I had to mark its Mired value based on my comparison with my Kodak color conversion gels that I used over the camera lens.

As for green/magenta, there are some limitations.
While one can get quite a tight to broad range of filters that can go over a camera lens (mostly needed only when shooting film) there is not such a fine range of filters to go over the lights.  Roscoe has green filters called 1/8, 1/4, 1/2, Plus green. But these are not CC values so you have to compare them to CC filters and mark them.  The first link I posted above has this info.  You can stack them on the lights of course (whereas stacking filters on a camera lens will degrade the image.)

So you see, in the end it is really simple...

This is why I started avoiding shooting commercial interiors. All of the unusual energy efficient lighting made shooting more of a science project than what I got into photography for in the first place. (Although I am good at the science part.)

Digital photography has made all of this much easier, as there are a number of simple retouching solution that don't require the lighting to perfectly balance. Plus you can shoot tethered and adjust your results on site to make sure things are to your liking.

[Dansk]

Yeah complicated isnt the word. Trial and error works but the biggest problems I face in this case are 1. Lack of time to mess around and 2. having to buy the "correct" gels without buying 200 more for safety that I will leave on the shelf for ten years before throwing out likely. Its wasteful and an expense not billable to this job so I wanted to figure this out as best I could before hand.

 The tip for viewing through a filter over the screen is solid and I will try that. But even still to be safe I will be buying more CTO as well as plusgreens. Looks like there is no choice in the matter.

  FWIW I have noticed the new lighting systems installed at a lot of stores and facilities are ever changing and not always for the better especially in terms of photography. Flat and weird color temps sure keeps us on our toes.

Thanks for the help Alan

[dwdallam]

Yeah I know all that generally speaking, but if it is truly the case that you have that many different lights in one room, you simply cannot balance it--period. So you just take the most dominate light and use it, or shut off the dominate light and use the better light--tunstun for example, and use tungstun gels over your lights.

I mean what else can you do given a time line? Designers know about color temp and can't possibly want you to do the impossible--to balance three or more different color lights with strange spectrums perfectly in 10 minutes!

But what I meant regarding the color problem of strange lights was to find the spectrum of the light by dialing up the WB, and then comparing that spectrum to known gels that you already have and know the temp for--have marked. Then just use the next higher or lower or whatever compensating gel. You now what I mean, a very calculated process of elimination.

Dansk

Why can't you simply balance the house lights by gelling them to get the color you want? Also, if you do a little searching, you can find how VERY quickly change very specific color tones in PS. I did some images once where the client needed the colors of the cloths to be different or shades to be different--very subtle.  After you find the color and effect you need, you can batch apply it. Also, yellow ambient light can work well depending on what you are after.

[dwdallam]

I just spoke to the maintenance guy from the building and the lights are fluorescents with chrome fixtures... They are odd size though I am unfamiliar with this style and I wonder what the tint to the tubes were but the guy had no idea as they have yet to replace one because the building is brand new. Anyways I suppose I'll start with window green gels and pack the CTO's as a safety and hope for the best.

Thanks again Craig
[a href=\"index.php?act=findpost&pid=196092\"][{POST_SNAPBACK}][/a]

Man I think you're over thinking this. You obviously know what to do, just do it. Pull one of the bulbs out if you want and look for the manufacturer stamp. Call them and ask them what temp they are. Take a few test shots using different gels. Done?

[Dansk]

dwdallum thanks for the input its appreciated.

   The place is just under a million square feet and I would guess there is over 5000 house fixtures not to mention the ceiling height is 60' or thereabouts so its not practical at all to gel the house lights. As for "dialing up the white balance"? Thats exactly what I did and I posted the numbers of correction in post 1. The color temperature from these lights sources is one I have yet to see myself and I hoped that someone here may have seen it and had a gel # or combo to go with which is why I posted the pics and Kelvin temps. I have had some solid input from Craig and Alan and at least now can whittle down my new gel purchase to about 50 instead of 200 so there was merit to this post for me.

  As for over thinking? Haha man thats the best compliment you could ever give me thank you!


edit* I have done some research about fluorescents on Google etc. and it seems there are many new types out now that have these spikes of color that vary across the board and can even be warmer than tungsten. Alan your tech input is bang on