Predictive color balance model

msf245

I’m a student at Cornell University working on a design project aimed at creating a formulaic model to predict the color temperature of sunlight at any time and geographic location in the world. Our algorithms are well on their way and we need empirical data to verify our work. Thus, we are turning to the online photography communities to help us out!

We are asking people from all around the world to capture the color temperature of direct sunlight on a, preferably, sunny day by photographing (in RAW) a white balance card pointed toward the sun. From that image, we'll determine the color temperature and add it to our growing dataset. While we want pictures taken at any time of day and sun position, the white balance card musts be placed in direct sunlight.

Please send all photos (in any RAW format) to msf245@cornell.edu and include the following information in the email:
-Location (lat/long, if possible)
-Time
-Weather description (brief: cloudy, clear, overcast, a picture would work as well)

JAG

I have moved this post from challenges to the appropriate forum.

ziggy53

Sadly, unless you specify, or supply, a specific white target, and very specific exposure protocol (to prevent color channel clipping), the file samples you get will have limited and uncertain value.

pathfinder

I was concerned about the specifics of the white balance target also Ziggy, I would think it would need to be uniform in all of the potential locations, and with a consistent exposure as you mentioned.

There are commercial non-specular grey white balance targets that should solve most of Ziggy's concerns.

One I have used for years is from BalanceSmarter, but I think the same device is available from Lastolite as the EzyBalance card (The Xrite Color Checker Passport has a uniform grey card in it for color balancing that might be suitable. White paper is just so variable I would think.)

The target from EzyBalance is a non-specular medium grey target that should photograph with all three color channels ( R, G, B ) as a single spike centered in the camera's histogram. The real beauty of the device is that it also gives one an accurate exposure, as the three spikes should be dead center in the histogram as well when exposed correctly. Mine puts the three channel spikes dead center in my histogram on my camera body when exposed according to Sunny 16 or via Av mode, which is fairly convincing to me that the exposure is correct.

One way to check exposure is to shoot a 32 step grey scale and see if all the 32 steps are distinguishable in your image, before image editing.

Shooting a white card in Av mode is likely to come out grey as that is what the incident meter in one's camera is designed to do, just like shooting snow fields in the sun without + exposure compensation.

The OP is probably quite aware of these concerns, and may have plans to deal with them that they did not share in the opening post of this thread. Perhaps, they might share these details with us.

Angelo

ziggy53 wrote: »

Sadly, unless you specify, or supply, a specific white target, and very specific exposure protocol (to prevent color channel clipping), the file samples you get will have limited and uncertain value.

aren't pro WB cards universally uniform?

ziggy53

Angelo wrote: »

aren't pro WB cards universally uniform?

Here is Pathfinder's test:

http://www.dgrin.com/showpost.php?p=800869&postcount=1

pathfinder

Angelo wrote: »

aren't pro WB cards universally uniform?

Let's assume they are, Angelo. A pure white card, shot in Av mode, as a full frame filling image, will photograph as a mid tone grey, right? That is what reflected light meters do.

An incident meter would give the correct exposure to capture the white card as white. But cameras have reflected light meters

ziggy53

FWIW, I sent an e-mail to the address in the first post on July 20, expressing my concerns. Now, 10 days later, there is still no response.

Without some scientific controls over the WB target and exposure procedure, it's hard to put much faith in their results.