Color Temperature Readings

[Paul Metcalf]

I'm seeking information from other photographers who have run color temperature tests in an effort to determine if my color meter is not performing properly.

My meter reads tungston lights around my house at 2500 - 3500 deg. K, which seems right. Also, my wife's Ott "daylight" sewing light reads around 5400 deg. K, which also seems right. Here's some outdoor data from today (10/31/2003) that I'm questioning (I'm in Seattle, WA., and its an absolutely beautifully clear, blue sky day with air temps in the low-40's):

7am, sun below horizon, sky reading is 8000 - 10000 K; 730am, sun is around 5 deg above horizon, direct reading of sun is 5000 - 6000 K; 8am, sun is about 10 deg above horizon, direct reading of sun is 8000 - 10000 K; 1130am, sun is about 28 deg above horizon (about as high as it gets this time of year in Seattle), direct reading of sun is 20000 K (or higher, meter pegs at this).

By direct reading, I'm pointing the opal disk of the meter at the sun, but it would also be gathering light from some of the sky as well.

If my meter is working properly, it's telling me that mid-day sunlight around Seattle in late October is very blue and is dominated by the large area of the blue sky, and the sunrise light is actually closer to what the "daylight balanced" film will register as normal (I would need a warming filter to actually make the lit scene look like sunrise light). I haven't tried sunset yet, as either I've been holed up in my office when the sun sets or it's the more typical rainy weather. Maybe this weekend we'll have clear skys to try this and see if the sunset readings are on par with literature (<3000 K).

Anyone with similar experience, and/or is my meter (and meter reader) wacko?

[Tim Curry]

I don't have a color meter, but I think it may not be a good idea to point a precision instrument at the sun. This may have fried some part of the circuitry and caused a problem for later readings.

[Ted Harris]

Paul,



Your indoor readings sound right on. "Sunlight" is so variable that it is difficult to tell if there is anythign wrong with those readings.



It would seem to me there are more precise and easier ways to check the calibration of your meter than what you are doing.



The most obvious to me is to use a MacBeth color chart.



One question, is your meter one of the older two color meters or a curretn three color meter?

[Paul Metcalf]

Tim- Appreciate your concern. Instructions for meter instruct that it be pointed at the light source, to include the sun (specifically indicates this with sketches showing such).

Ted- This is an older two-color incident meter, so I'm not able to read a Macbeth chart. I need calibrated source light (which is why I tested it with the Ott (sp?) light which is suppose to be "daylight" and indicates a reading around 5400 K).

thanks.

[Tim Curry]

Paul, Where I am (Tucson) we tend to have bright sunlight about as often as you have overcast. During the day in full sunlight, between 10am & 2pm, it is common to use a slight warming filter (I use an 81a) for transparency film because of the amount of light coming from the blue sky. This is most visible in shadows, they really are blue here. It doesn't seem possible that the full value sun we get could be affected by a blue sky, but it is normal here.

It sounds like in your area the sun is a novelty. Here we're just starting to get cooler weather, finally.

[David du Busc]

You might consider taking your meter to a pro camera store that sells color meters... compare yours and a new one under varied lighting conditions.

[Neal Shields]

On my meter (two color) if you play with it with a small flash light you find that there is a blue sensor and a red sensor left and right of each other under the round disk. They work against each other, one drives the needle one way and the other the other.

I haven't had mine long or taken many readings but I would always expect the sun to be much warmer than the sky. (my experience)

My north sky would read about 10,000 when the noon sun reads 5800.

Be sure you arn't covering part of one of the sensors with your fingers while taking a reading.

[Jim Rice]

I've always wanted a color temperature meter. But, i get the feeling this may be a fool's errand. Am I right?

[Ted Harris]

Jim,



,

Not a fools errand at all but also not something that most folk generally need. If you are doing work where absolute color accuracy is critical or you need to make subtle adjustments to compensate for mixed lighting conditions then a color meter can be very useful. Of course, today Photoshopcan do the job too. If you do decide to get one stay away from the old two color meters, they aren't worth the agony.

[Ted Harris]

Jim,



Not a fools errand at all but also not something that most folk generally need. If you are doing work where absolute color accuracy is critical or you need to make subtle adjustments to compensate for mixed lighting conditions then a color meter can be very useful. Of course, today Photoshopcan do the job too. If you do decide to get one stay away from the old two color meters, they aren't worth the agony.

[Jim Rice]

Thanks Ted. And on a related note, is there a good way to hold Wratten gels in front of a lens?

-j

[Ted Harris]

Jim,



There are a number of frame systems on the market (or used to be). The easiest to find is the Lee Filter System (www.leefiltersusa.com). They are also marketed in this country by Calumet under their own brand. Lee makes a number of frames that fit into their holders and lenshoods that are 'empty holders' that is to say into which you can insert your own gels. Similar systems may also be available from HiTech and used to be available from Sinar.

[David A. Goldfarb]

I have an old Gossen Sixticolor meter, which sounds like what you probably have, and I find it not terribly reliable with tungsten sources (perhaps because the levels are too low), but outdoors, it's useful for figuring out how much correction to apply in open shade, deep shade, overcast conditions, or some combination thereof (81A, B, or C?).

[Paul Metcalf]

Neal S- I tried your flashlight trick and I can see the side to side movement as you noted. One thing, though, the meter pegs at both ends of the meter when the light is close enough. Does yours do this? This makes me think that maybe there's something up with it pegging at 20000 K when pointed towards the sun (the sun this time of year in Seattle only gets up to 30 degs. max, so it's not that bright). Since the instructions mention pointing towards the light source, to include the sun if the case, I'm wondering if there is something with it getting too much light and oversaturating (???). On my way home friday, we had a clear sky sunset (in Seattle, no doubt!), and just as the sun was hitting the horizon (about 4:45) my direct reading was 3300 K, which seems reasonable. Earlier in the day (11:30am) the direct sun reading was 20000 K (basically pegged). The sky was clear and blue, but I would have thought there would be some impact from the sun. A search on the web does state that open shade under a clear blue sky can reach 20000 K, but it also states sunlight is 5500 K.

Confused in Seattle.

[Frank Petronio]

I "used to" do serious architectural photography and used a Minolta color meter. I never found that I could rely on a color meter to give me a perfectly accurate color temperature. There are just too many variables in the process to aim the CM at a scene and say, "yep, 5327.689 degrees - need a X and Y filter..." Instead, I thought of it as a tool that would tell the the difference of various light sources in my scene, so that I could filter "out" the mixed lighting. In other words, I want the lighting to match (often letting the incandescent and tungsten retain some of its warmth for a more natural look). But I would still choose the film to match the contrast and overall warmth/coolness that I wanted.

Then I would simply fold a few pieces of black photo tape into "L" shapes, attach one leg to the lens rim and the other to the gel filter... Or use one of the CHEAP spring loaded gel holders that Calumet sells.

Now I filter for the gross light and use Photoshop to fix the details. It isn't as elegant but much more time and cost effective. I really could have used Photoshop 15 years ago!

[Neal Shields]

Yes, my meter will peg both ways with a flashlight. A pretty dim one at that. As they work against each other, it sounds like one sensor is weaker than the other on your meter.

Neal