PDA

View Full Version : How to read "K" value of light???



stradibarrius
29-Apr-2015, 07:22
Would any of the common "light" meters be able to give the "K" value of the light in a given scene?
I have just finished a remodel of my violin shop and the "T5" florescent bulbs give off a strange color cast. I have tried adjusting the WB values of my digital camera to see if I could at least get close.
I was trying to come up with a way of measuring the temperature of the light.

DrTang
29-Apr-2015, 07:56
Minolta used to make a color meter

spectra and lunasix made one years ago..but they seemed almost impossible to use

Bob Salomon
29-Apr-2015, 08:11
LuxMeter Pro on my iPhone reads it. Don't know ho precisely though with a discontinuous spectrum source like yours.

Peter De Smidt
29-Apr-2015, 08:19
These days florescent bulbs are all over the place. You can use something like a Whibal, if you're shooting raw, and white balance in post. Or you can use something like this: http://www.bhphotovideo.com/c/product/535317-REG/Dot_Line_DL_2577_77mm_White_Balance_Lens.html , and do a custom white balance in camera. I do the latter when I have to video judo tournaments with my dslr.

StoneNYC
29-Apr-2015, 11:04
They do make an "FL" filter, specifically for Florescent lighting, in most cases that should do it, unless you are trying to balance that with tungsten bulbs on the wall or sunlight coming through an uncoated window.

Peter De Smidt
29-Apr-2015, 11:31
The "FL" filter isn't very effective anymore, due to the bewildering array of different fluorescent lights. With some commercial interiors I shoot, the color values of the light fixtures will vary all over the place, as the maintenance people apparently buy whatever is cheapest, or whatever they just happen to grab that day.

Bob Salomon
29-Apr-2015, 11:36
They do make an "FL" filter, specifically for Florescent lighting, in most cases that should do it, unless you are trying to balance that with tungsten bulbs on the wall or sunlight coming through an uncoated window.

That will work for some daylight fluorescents but not with all. Each filter manufacturer that makes or made FL filters made them to correct what they felt was the "average" FL correction. But if you have different FL tubes then they made the filter for or if you have a mix of different daylight tubes then the filter may not help you at all. But another manufactures FL OR F Day might be better. BTW FL is a Tiffen trademark and when we were the B+W distributor Tiffen sent us a cease and desist letter and we had to have our designation changed to F Day. Also with Heliopan's version. Ran into the same problem with Heliopan's High Transmission polarizer reps. Turned out the High Transmission is a Trademark of Tiffen's also. So we had to have the factory change to HT on the filters.

StoneNYC
29-Apr-2015, 11:59
That will work for some daylight fluorescents but not with all. Each filter manufacturer that makes or made FL filters made them to correct what they felt was the "average" FL correction. But if you have different FL tubes then they made the filter for or if you have a mix of different daylight tubes then the filter may not help you at all. But another manufactures FL OR F Day might be better. BTW FL is a Tiffen trademark and when we were the B+W distributor Tiffen sent us a cease and desist letter and we had to have our designation changed to F Day. Also with Heliopan's version. Ran into the same problem with Heliopan's High Transmission polarizer reps. Turned out the High Transmission is a Trademark of Tiffen's also. So we had to have the factory change to HT on the filters.

I suppose that makes sense, and using the color meter to read the scene obviously would be best, I know that in the movie industry they just bring their own bulbs and if they ever shoot in at normal location that has fluorescent lighting, they will swap out every single bulb in the entire room with their special bulbs that are the right color tone, and then when they're done for the day they will remove those and replace them with the bulbs that existed before hand, but I figured for all intents and purposes the FL (or F-Day) filter was better than nothing, and a lot cheaper than getting tons of color accurate bulbs or buying an expensive color meter and tons of various temperature filters, the OP doesn't sound like they are looking to go crazy here, but I could be wrong.

Drew Wiley
29-Apr-2015, 12:07
Even the best color temp meters only work so-so with fluorescent lamps because these have a discontinuous spectrum. High CRI bulbs are less obnoxious than the usual el cheapo commercial varieties. Meters just read peak sensitivities corresponding to the specific three in the film itself. You can make advance film and printing tests, but don't expect good results. Those silly various purple-amber fluorescent filters do work OK for old-school cool white bulbs, just OK. Any quality
shot requires you either bringing in your own lighting or resorting to black and white film.

Luis-F-S
29-Apr-2015, 12:13
Minolta used to make a color meter

+1 and they work very well even after all these years. L

bob carnie
29-Apr-2015, 12:25
I have posted this before but what the hell.. I would be very interested in a handheld meter that can read LAB... not the paint chip device btw.

Drew Wiley
29-Apr-2015, 12:34
Anything can be had for a price, Bob. It's not the hardware but the software that's the issue. My wife used a small spectrophotometer in biotech that cost six
million dollars. It was stored behind a true timed bank vault door. The software was so proprietary that noboby in the entire corporation was allowed access to
the complete script. You should really get ahold of Joe Holmes in my neighborhood and try to get your heads together. Or maybe somebody like XRite is already
onto this. But don't expect any machine to be a substitute for the human eye. These things just speed workflow and save eye fatigue, unless of course they're
dealing in something molecular and otherwise invisible to human vision, like the expensive device I just described.

bob carnie
29-Apr-2015, 13:11
Hey they built the six dollar man... they should be able to make a LAB meter for much less.

Drew Wiley
29-Apr-2015, 13:21
Yeah, when I started fooling with paint spectrophotometers they cost nearly a hundred grand and were about ten feet wide and eight feet tall. IBM made them.
Now they're a couple thousand and a small as a darkroom densitometer. The funny thing is, the early ones were better because they plotted full spectrum and
didn't just interpolate points. The truly portable ones using battery and not tethered to a serious power supply aren't good for much at all; and the cell phone apps are just a marketing ploy and otherwise relatively useless. I knew about them before the public ever did, and the real reason behind them. To push enough light into any deep hue and get a reflective reading you need a pretty powerful xenon flash burst. But none of them are even remotely a substitute for the human eye, particularly at the extremes either direction. They just save eye fatigue with the approximate early stages of matching. When our full-time matcher hits a bind he always looks me up for an actual visual opinion.

StoneNYC
29-Apr-2015, 14:38
Sekonic makes a color meter FYI (as in currently) since Minolta doesn't obviously.

Nothing wrong with used, just thought I would mention it.

Drew Wiley
29-Apr-2015, 15:09
Gosh, Stone, I thought you had more class than that! Evvvverybody knows the ONLY color meter that ever existed is the Minolta. They can still be had. But no,
I'm not selling mine.

Taija71A
29-Apr-2015, 17:11
... Evvvverybody knows the ONLY color meter that ever existed is the Minolta...

'Houston, we have a problem'.

http://www.bhphotovideo.com/c/search?atclk=Brand_Sekonic&ci=900&N=4077634543+4289304159+4291226600

Kirk Gittings
29-Apr-2015, 18:58
I have a Minolta Color Meter III I've been meaning to sell for awhile......

StoneNYC
30-Apr-2015, 08:47
Gosh, Stone, I thought you had more class than that! Evvvverybody knows the ONLY color meter that ever existed is the Minolta. They can still be had. But no,
I'm not selling mine.

Yea yea.

I was on set with the DP on ... I think it was "Unfinished Business" ... But I worked on a lot of sets with a lot of DP's so it's sort of a blur, but anyone some talented and successful DP that had used his Minolta for years was trying out the new Sekonic light meter and color meter and said it seemed to do the job accurately and the only reason he wasn't switching was because his wasn't dead yet, but he wasn't worried about if it died and he had to switch.

So I took that as a sign not to worry. Sekonic's are very good also, I would have gotten an older Minolta if my pocket wizards worked with them like my Sekonic does. I've been considering finding a smaller spot/incident meter than the Sekonic 7xxDR (x's are numbers I can't remember but there's only one 700 series that I know of anyway) as its bulky and heavy but I don't think they make a dual meter that smaller with a spot that also goes lower in light reading. I could be wrong there's a lot out there. I know there's a spot that goes much lower in light trading I just don't think it also has an incident meter AND is smaller. But again, I'm open to being corrected, then I'll go buy one! :)

Drew Wiley
30-Apr-2015, 08:51
"Houston we have a problem. My super-expensive new Sekonic meter even tells me the correct Wratten filter to use, and now you're telling me that I can't get
those anymore even down there on earth?" Thanks. But I'm happy with my Minolta and my set of REAL Wrattens, many of which went out of manufacture long
ago. Good luck with those phony poly gels cut out from half-melted soda bottles.

Bob Salomon
30-Apr-2015, 08:55
"Houston we have a problem. My super-expensive new Sekonic meter even tells me the correct Wratten filter to use, and now you're telling me that I can't get
those anymore even down there on earth?" Thanks. But I'm happy with my Minolta and my set of REAL Wrattens, many of which went out of manufacture long
ago. Good luck with those phony poly gels cut out from half-melted soda bottles.

The problem with Kodak gels, other then being fragile, is that they were made from dyes that changes over time and exposure to light. So are all of yours still the original values?

StoneNYC
30-Apr-2015, 08:56
The problem with Kodak gels, other then being fragile, is that they were made from dyes that changes over time and exposure to light. So are all of yours still the original values?

Of course they are, he never actually takes them out of the boxes into the light ;)

Drew Wiley
30-Apr-2015, 09:29
Don't worry. All the filters I actually use for any serious color shooting are multicoated glass. I'd never use gels in the field. Too many things can go wrong; and
real gels aren't cheap by any means. What's nice about the Wrattens are all the specialty types they once made, some of which are invaluable in technical darkroom applications (typically contact work, so the light is not image-forming). And yeah, I do have a full set of Lee poly "gels" for similar applications or trial lighting shots (back when I did some architectural shooting) - a nice way to test the effect before investing in expensive glass filters. The Kodak filter handbook was wonderful. It gave not only the spectrograms of each but the fading characteristics relative to various light sources. And yeah, Bob, I do test them for fading from time to time. But all kinds of filters are made from dyes. Try to find a cyan one that isn't fugitive (though I've never found a need for one of those myself). Various flavors of magenta filters are highly useful in certain color reproduction techniques. These were commonly used back in graphics arts days with highly specific characteristics. You can't just substitute magenta colorhead light and get the same effect. But again, strictly a lab or darkroom application - nothing you'd use in the field. If Stone wants a 600C Fujinon bad enough, that should collapse his filter budget for a couple of decades anyway.

Greg Davis
30-Apr-2015, 14:07
I had a Minolta II color meter and sent it to Quality Light Metric for service, but parts are no longer available, so it was scrapped. I now have the Sekonic and find it to work very well. It has separate settings for film and digital, which doesn't make logical sense if it is supposed to measure the color of the light source, but you do get correct color of you use the film setting with film and digital with digital. It also reads the brightness of light in lux.

Darko Pozar
30-Apr-2015, 16:03
The problem with Kodak gels, other then being fragile, is that they were made from dyes that changes over time and exposure to light. So are all of yours still the original values?

and heat...

StoneNYC
30-Apr-2015, 16:04
If Stone wants a 600C Fujinon bad enough, that should collapse his filter budget for a couple of decades anyway.

I've got all the filters I need, I do need that lens ;)

David Lobato
30-Apr-2015, 16:24
I don't see this mentioned, but fluorescent light color correction is not in the realm of degrees Kelvin (K) light temperature correction which ranges from blue to amber.

Drew Wiley
30-Apr-2015, 16:27
Lots of glass filters share the same pitfalls with heat and light as gels. Traditional Tiffen filters are dyed tissue cemented between pieces of glass. Certain colors of
filters can only be made with fugitive dyes, regardless of how those dyes are dispersed, like into optical acrylic. So regardless, ya wanna keep your CC and LB
filter set in a dry cool environment. Some of our more commonly used taking filters and skylights etc are more robust; and some even rely completely on the coatings. Speaking of coatings; did you know, Stone, that those Fuji C coating are really really nice?... heh! heh! (as I slowly torture you)

Jac@stafford.net
30-Apr-2015, 16:47
Certain colors of
filters can only be made with fugitive dyes

Fugitive is not the word I would use. Fugitive dyes typically last less than 24 hours.

Drew Wiley
1-May-2015, 09:50
Fugitive was the word routinely used for most non-pigment color printing processes containing dyes for decades. It's a standard term even when comparing various options for categories of artist's pigments themselves, and has been a common term for this far longer than I've been alive. It's a term entrenched in the relevant literature as well as art technique textbooks. But it's relative, just like the opposite term, "lightfast". Very few hues are truly lightfast. If you want o know which ones, look at the surface of Mars. And those of us who do have reason to discuss color dyes a lot (not on this forum) use the term "fugitive" frequently. We even used it in an industrial paint and pigment context. But in what manner something is fugitive depends. That's why Kodak published three distinct categories of susceptibility for every single Wratten filter they offered. Fugitive could mean hours, days, years, whatever. All depends. But it does mean
ya gotta be aware of specific susceptibilities.

Oren Grad
2-May-2015, 08:51
FWIW, Kenko took over the Minolta meter business. The Color Meter IIIF still lives, as the KCM-3100:

http://www.bhphotovideo.com/c/product/514616-REG/Kenko_KCM_3100_KCM_3100_Professional_Color_Temperature.html

Greg Davis
2-May-2015, 12:40
That would is a really good meter, too. I chose Sekonic because I have their exposure meters and like them, and they gave an educator discount.

8x10 user
2-May-2015, 18:03
Gosh, Stone, I thought you had more class than that! Evvvverybody knows the ONLY color meter that ever existed is the Minolta. They can still be had. But no,
I'm not selling mine.

Well there is the Broncolor FCC.

Bruce Watson
3-May-2015, 05:40
I was trying to come up with a way of measuring the temperature of the light.

In general, you can't. "Color temperature" is about comparing the hue of an ideal black body radiator to the light source you are trying to characterize. In this case, it's trying to compare apples to concrete blocks. Fluorescent light is discontinuous, and the spectrum emitted typically has huge peaks and valleys, not found in the spectrum of an ideal black body radiator. The quality of light between the two sources is incomparable.

In practice, a color meter basically looks only at the orange-blue axis, and tells you where the light ends up on that axis. Most color meters tell you nothing about the magenta-green axis. Unfortunately, that's what matters most with cheap building fluorescent lights, because they tend to produce the majority of their light via a huge green spike. This is the reason many video cameras show you a sickly green color cast even when you set their "color temperature" to the correct value (say, 6500K for a "daylight" tube).

For this reason most modern video cameras include a manual white balance function that makes many measurements, including the dreaded magenta-green axis. A good manual white balance made against a calibrated gray target (think 18% Kodak gray) will generally give you fairly accurate color without a noticeable color cast in all but the most challenging lighting.

As to human clients visiting your shop... showing your wares under color corrected light will make a better impression. There are good fluorescent tubes out there -- KinoFlo True Match (http://www.kinoflo.com/Products%20Button/Lamps/True_Match/True_Match.html). Westcott. Some others. Very little in T5 except for biax bulbs (2G11 base) which are common in cinematography lighting, but not in building lighting.

AFSmithphoto
4-Sep-2015, 10:18
In general, you can't. "Color temperature" is about comparing the hue of an ideal black body radiator to the light source you are trying to characterize. In this case, it's trying to compare apples to concrete blocks. Fluorescent light is discontinuous, and the spectrum emitted typically has huge peaks and valleys, not found in the spectrum of an ideal black body radiator. The quality of light between the two sources is incomparable.

In practice, a color meter basically looks only at the orange-blue axis, and tells you where the light ends up on that axis. Most color meters tell you nothing about the magenta-green axis. Unfortunately, that's what matters most with cheap building fluorescent lights, because they tend to produce the majority of their light via a huge green spike. This is the reason many video cameras show you a sickly green color cast even when you set their "color temperature" to the correct value (say, 6500K for a "daylight" tube).

For this reason most modern video cameras include a manual white balance function that makes many measurements, including the dreaded magenta-green axis. A good manual white balance made against a calibrated gray target (think 18% Kodak gray) will generally give you fairly accurate color without a noticeable color cast in all but the most challenging lighting.

As to human clients visiting your shop... showing your wares under color corrected light will make a better impression. There are good fluorescent tubes out there -- KinoFlo True Match (http://www.kinoflo.com/Products%20Button/Lamps/True_Match/True_Match.html). Westcott. Some others. Very little in T5 except for biax bulbs (2G11 base) which are common in cinematography lighting, but not in building lighting.

Bruce is dead on. The Kelvin scale doesn't even really reflect shifts on the magenta/green axis. There are other methods of quantifying, but to get a meter that can read a fluorescent accurately AND communicate better than a simple Kelvin number typically cost a fortune.

If all you're after is a rough idea, "Pocket Light Meter" for the iPhone includes a Kelvin scale. Not perfectly accurate even for easier to read sources.

Your best bet is to stick a piece of white poster board under one of these lights in such a way that it is not also lit by any extraneous sources and take a picture with it exposed around zone 6 or 7.

Bring that shot into lightroom (or whatever you use) and use your eyedropper tool to balance the shot so that the card is true white. Whatever lightroom tells you your balance is your Kelvin rating.

Mind the handy green magenta slider just below the color temp. (Assuming you are using lightroom) If its strong toward magenta, you lights are green and vice versa.

Kirk Gittings
4-Sep-2015, 11:06
The "FL" filter isn't very effective anymore, due to the bewildering array of different fluorescent lights. With some commercial interiors I shoot, the color values of the light fixtures will vary all over the place, as the maintenance people apparently buy whatever is cheapest, or whatever they just happen to grab that day.

Exactly. we used to have cool white and warm white. Now there are dozens of fluorescents. On a DSLR I put it on fluorescent, but include a grey card in the scene which I use to neutralize the color more finally in Lightroom.

Kirk Gittings
4-Sep-2015, 11:17
The Minolta color meter was never very accurate for fluorescents-I used one for decades for architectural interiors. Every film had a different responce-particularly like from Kodak to Fuji. One had to take what it said and "use your experience" with a particular film then also adjust for long exposure color shift etc. The Minolta color meter took none of that into account nor did it have any memory function. It was mainly about testing and experience. It was tedious. I do not miss it. Digital is a dream compared to transparencies with flourescents.

Michael Mutmansky
4-Sep-2015, 13:46
Would any of the common "light" meters be able to give the "K" value of the light in a given scene?
I have just finished a remodel of my violin shop and the "T5" florescent bulbs give off a strange color cast. I have tried adjusting the WB values of my digital camera to see if I could at least get close.
I was trying to come up with a way of measuring the temperature of the light.

All but the cheap off-shore lamps will print the color temperature on the end of the lamp (you probably call them 'light bulbs'). Go up into the luminaire and pull one down. It probably requires a twist of the lamp to get it out. You should be able to see some markings on the lamp near the socket at one end. The color temperature will be printed there.

However, as has been said, this is only somewhat useful because the green/magenta level will be undetermined by this value, and since the light is a discontinuous spectral distribution, there will still be shortcomings even if you perfectly match up the CT value.


---Michael

VictoriaPerelet
8-Sep-2015, 21:58
Pretty much any digital camera with manual control has K white balance control where you can preset light temp.
Set your exposure. If you are combining flash and constant light it is important.
Guestimate color temp, set it, take pic of white piece of paper. Paper color on preview will be either bluish or redish, based on that select another temp.

Its longer to write - in few iterations you'll get color temp.

Works fine with flash (if you can connect sync cable), thungsten, hmi etc combined all together.


But then again some people have allergy on digital.... even when it helps to take decent film pic

luggeroflinhofs
27-Nov-2015, 05:00
Back to many years ago, one of my hats was as an industrial photographer. I used the Minolta II as a starting point but needed experience to get it right. In talking with the plant maintenance people we concluded that all the overhead florescence bulbs were of the GE "cool white" brand, both in the plant and the office area. I asked that they only replace with this same bulb, which they did. This allowed me to use a magenta 40cc filter on camera with daylight transparency film any where in the building. Worked with strobe flash also. For exec portraits, kept them away from windows.

Tom