White Balance under Sodium gym lights?

[ampguy]

Have used a 1/2 dozen P&S and video cameras and can't quite get consistent natural lighting under the school's gym lights. Another parent was telling me last year the schools used Sodium lights, which are cheaper than flourescents.

I can get close and acceptable with measuring a white piece of paper with the preset modes, but it still does not match what the eyes see in color balance.

So for JPGs, does any one have tricks for getting accurate white balance, and accurate colors? This is not specific to SLRs, also had trouble with the RD-1.

It's not bad enough to want to do raw, but it seems like the custom modes don't quite work?? Not really SLR specific, but today was at an event with D40x and couldn't get WB quite right.

 

[Al Kaplan]

Forget it. It ain't gonna happen! Some types of lamps might be bright and economical but they'll never give you "correct color". They have a "discontinuous spectrum", with certain wave lengths of color missing completely. Not that many years ago flourescent lights were notorious for making everything look a bit green in photographs, and for the same reason. They've pretty much overcome the problem with current flourescent tubes. Maybe they'll get sodium lights to behave at some point. Be grateful that you're not trying to get perfect colors on color transparency film!

 

[rogue_designer]

Al's got it. Gas discharge and HID lights like Sodium and Mercury vapor put out alot of light, but they have wonky spiked spectrums that make color balance a major pain. Even worse when doing outdoor shots and mixing the light sources.

Setting a custom wb with a grey card, or an expo disc or similar helps, but you won't get all the way there. It's not even a matter of correction, as much as it is filling in information that isn't there due to the limited spectrum. Even shooting raw will only get you part of the way there.

If you can't mount strobes in the rafters like the pros, you'll have to live with less than optimal color. BW conversion works. 😀

Not terribly helpful, I know. Good luck.

 

[ampguy]

Thanks, I think I'm going to look into bracketing WB, the other thing I can do with my D40x is to find the best compromise of skin tone and surrounding colors, keep that image, and use it for WB reference when I'm in that environment (assuming they don't change the lights 😀)

 

[pachuco]

Out of curiosity, why are you opposed to shooting in RAW? It would solve your problem and give you additional stops if you are a bit off on exposure.

 

[mr. mohaupt]

The only solution I ever found was to shoot slow shutter speeds. It allowed more of those different light frequencies to be captured and get "closer" to what should be a normal white balance. I shot sports in said gyms so a lot of the time I would spend hours in photoshop trying to get it close.


I think 1/90 (maybe 1/125 i cant remember anymore) or slower should capture enough of those different frequencies to get "the right color"

~m

 

[uhligfd]

mr. mohaupt, how do you think a constant "wrong light", i.e., a disparate spectral output such as from a sodium light will average out as "normal" or balanced light over longer time spans if exposed or averaged over a longer time period?

Do these weird spectral balance lights fluctuate in their frequency contents, maybe ?

I would doubt that, though. So what do we attribute this to? Anyone know?

 

[Roger Hicks]

I am constantly amazed at what I can get with manual WB, even under near-monochromatic sodium vapour lights (I seem to recall that there are two wavelengths in street lighting). But I do better with manual WB (and Raw) than with trying to re-balance from any other WB option.

Tashi delek,

R.

 

[rxmd]

You can always shoot a reference picture of a colour reference chart for some later fine-correction. It won't help all the way with lamps with odd spectrums, but it's better than nothing already.

Regarding Roger's post, under monochromatic light you get monochromatic pictures, i.e. no colour information. The two sodium lines are too close to each other. While the brain does fill in some things so that they seem to be in colour, you need at least one other light source to get colour in the camera.

 

[Roger Hicks]

Regarding Roger's post, under monochromatic light you get monochromatic pictures, i.e. no colour information. The two sodium lines are too close to each other. While the brain does fill in some things so that they seem to be in colour, you need at least one other light source to get colour in the camera.

Absolutely true. I have long puzzled over this, and the only conclusion I can reach is that apparently trivial amounts of supplementary lighting must contribute a disproportionate amount of colour information.

Tashi delek,

R.

 

[mr. mohaupt]

mr. mohaupt, how do you think a constant "wrong light", i.e., a disparate spectral output such as from a sodium light will average out as "normal" or balanced light over longer time spans if exposed or averaged over a longer time period?

Do these weird spectral balance lights fluctuate in their frequency contents, maybe ?

I would doubt that, though. So what do we attribute this to? Anyone know?

*shrugs* I guess not but what I found in my experience is if you motor drive through some images at very high shutter speeds you would see 5 or so different hues in each image (reds, greens, yellows, blues, etc). The longer shutter speeds would capture all of these colors and make what seemed like a "truer white balance." I use to use Auto White Balance on my mkII

Try it out, I will look for some examples but I don't really shoot sports in gym's anymore.



Now as for the sodium lights like Street lights, thats another story. He mentioned Gym Lights.

~m

 

[ampguy]

At my location, I can't get consistent color, even at 1/30th.

I might try raw but will probably use elements, not NEF

I agree that the perceived colors under the lighting are going to be different than when the objects are under other lighting, for things that can be moved (people, clothing, etc.) however, even under the lighting, an object won't seem to match the LCD review, even after perfect presetting of a pure white paper or grey scale card under this kind of lighting.

Lastly, I think I will shoot b&w JPGs under this light 😉

 

[payasam]

Within the limits imposed by the nature of the light, you'll have the greatest flexibility with raw.

 

[uhligfd]

Maybe - and i am trying to find a scientific basis for mr. mohaupt's findings here - maybe, as lights constantly turn on and off (just move your arm/hand quickly under fluorescent lights as on a street corner with a light) and you will see the image (hand) jump in discrete steps as you move it, this in sync with the cycles of the power company/light.

And so maybe, longer exposures may give room for "switched off" lights (when using 1/250 sec in their off cycle phase) to contribute their "false spectrum", too. Since each fluorescent has a slightly different spectrum output due to age, heat etc, the overall light at 1/30 sec or 1/8 sec might be more easily balanced ... than at 1/500 sec 0r 1/200 sec???

Just surmising here. Any experts on this, the physics of gas vapor lights and power line cycling?

 

[Al Kaplan]

Sodium vapor gives off light when an electric current excites the molecules. The color is based on the color that sodium molecules give off. They're visually bright in relation to the amount of electricity they use but don't have to be accurate as far as color goes.

Flourescent lights contain phosphors that glow when the gas, mercury vapor, inside the tube gets excited by the electricity going through it. The light you see is mostly all from those glowing phosphors, not from the excited gas molecules. Over the years they've gotten better at developing phosphors, or combinations of phosphors, to create a pretty good continuous spectrum, and more recently even keeping the greenish appearance under control. This has nothing at all to do with the length of exposure. It's just physics. Which vapor and which phosphors.

 

[bmattock]

I use an 'Expo-Disc' to set white balance in such situations. Works pretty well for me. Other companies make similar products now, but that's the one I use.

http://www.flickr.com/photos/wigwam/sets/72157603927200458/

 

[ampguy]

Those look great

Those look great

JPGs, or post processed? Is the expo disc anything other than a white lens cap? would it have any advantages to shooting white paper 18" away?

I use an 'Expo-Disc' to set white balance in such situations. Works pretty well for me. Other companies make similar products now, but that's the one I use.

http://www.flickr.com/photos/wigwam/sets/72157603927200458/

 

[navilluspm]

Setting white balance with a grey card is a good starting point - if you don't have an expo disc. Canon actually recomments using an 18% grey card for optimum WB in their Rebel XTI manual (in a little note on the bottom of the page). It will give you a decent starting point, but not perfect exposure.

 

[ampguy]

yeah

yeah

But I have a Nikon 😉

Setting white balance with a grey card is a good starting point - if you don't have an expo disc. Canon actually recomments using an 18% grey card for optimum WB in their Rebel XTI manual (in a little note on the bottom of the page). It will give you a decent starting point, but not perfect exposure.

 

[bmattock]

JPGs, or post processed? Is the expo disc anything other than a white lens cap? would it have any advantages to shooting white paper 18" away?

You can use the expo disc or its clones with either RAW or JPG shooting.

If you shoot RAW, when you convert in post-processing, you choose 'Camera WB' as your starting point (I use UFRaw, not sure what the choice would be for other programs). If you use JPG, you should be good to go.

I do not know if there is any appreciable difference between shooting through a "white lens cap" and the expo disc - I have not tried shooting through a white lens cap, and indeed, have never seen such a thing. Someone here mentioned that they used a white coffee-can lid or something like that - hey whatever works, I guess. I know the expo disc is purpose made and works very well for me.

As to the advantage over shooting a white piece of paper, it has several.

First, you only have to take one photo - not one photo with and one without the paper.

Second, if you shoot JPG, then you do not have to WB in post-processing. Shooting a piece of paper, you do.

Third, I'm color-blind. I can't properly set WB from a white piece of paper, I end up with color casts that I can't see, but you normals can.

Fourth, it is a set-and-forget kind of thing. The cap is in my bag. I get into a situation with wonky lighting, like a school gym, and I do the 'set custom WB' thing on my camera, hold the expo disc over the lens, hit the shutter, and if it says 'OK' then I'm done. Unless the light changes (like from ambient light coming in windows as well as gym lighting), I'm good to go for the entire event. I might reset WB if I change lenses.

I mean, to me it is dead easy and works great. I've posted about it many times, and many are the people who have gone out of their way to comment on what a fraud it must be, how they can set WB brilliantly by performing fifteen steps instead of the expo disc, how WB is a crock anyway, and all manner of sour grapes thereunto pertaining. I have not met one person who has used an expo disc (or one of its clones) who doesn't like it and add it to their bag of tricks permanently. But hey, whatever floats your boat. I got mine, I love it, and if no one else likes it - groovy.