3

Let's say we have some incandescent, flourescent, luminescent lamps and LEDs, and there are photos made with different cameras in auto mode with each of these lamps on, and with a white sheet of office paper lit by it. Is there a way to tell the color temperature of each lamp from the photo's EXIF data?

  • The question is unclear to me. Is only one light on in each photo? Or are all the light on in all the photos? – user50888 Nov 15 '17 at 14:29
  • Only one of each light bulbs is on, since we don't wanna mix the colors. You can also interpret it as only one light bulb exists with unknown color temperature, and we shoot it with different cameras on auto. I just wanted to give a slightly broader example. A real life scenario would be trying to find out color temperatures of light bulbs in other people's photos on the internet where the photo contains metadata. Flickr shows lots of EXIF data, and it's primarily where I intend to try this. – user1306322 Nov 15 '17 at 14:56
5

Some cameras have a section in the 'maker notes' section of the EXIF info that includes such information, but far from all cameras do. Even with the ones that do, not very many applications that show EXIF data will display that information in the EXIF data. Some raw convertors will open raw files using that value in the EXIF info as the default value if the selected options for opening raw files specifies such (as opposed to an option selected to open all files with a preset value such as 5200K).

Among the tools I routinely use, it isn't displayed. But if I upload an image taken with my Canon EOS 5D Mark III to Flickr and allow the EXIF info to be displayed, anyone can click on the "Show EXIF" link and then scroll way down to see it.

Here's a piece of a screen shot from a portion near the end of the EXIF info for this image at flickr:

enter image description here

Notice that the "Color Temperature Auto" and "Color Temperature Measured" values are identical. That indicates that when the camera is set to "Auto WB" the measured color temperature/white balance is used.

But keep in mind, color temperature is only a single axis across the entire color wheel we call White Balance. Artificial light sources are often well off the color temperature axis that is basically defined by the color of black body radiators at different temperatures. For instance, in addition to having a color temperature of about 3700K, traditional fluorescent bulbs also emit a green tint along the green←→magenta axis and need correction in the magenta direction. On the other hand, many of the popular LED stage lights found in small clubs are also at about 3700K but also have a decidedly magenta tint that requires compensation in the green direction along the green←→magenta axis. Both types of light are the same basic color temperature but look very different without compensation on the green←→magenta axis that is approximately perpendicular to the blue←→amber color temperature axis.

1

The color temperature should be printed somewhere on the bulb itself. Try looking at the small print around the base. You may also consider using a colorimeter that is able to measure the ambient temperature.

While you can take a raw photo and see an estimated color temperature in raw processing software, the values can vary. Bulbs with temperatures that vary by only a few hundred K may not be distinguishable when "measured" by cameras not designed to be used for colorimetry.

  • Camera auto, custom white balance, and software auto rarely match. Which, if any, is the "correct" temperature?

  • Ambient and mixed lighting in the environment will influence results.

  • Display calibration will affect attempts to refine "measurements" visually.

  • Different lenses can alter results. Consider the following images taken with the same lamp and white card, but different lenses. Custom white balance was set for each lens immediately prior to taking the photo. Values obtained were 2805K, 3164K, 6490K.

    2805 3164 6490

    Which "measurement" reflects the temperature of the bulb? This is not just a calibration issue. As long as photos look right, it doesn't matter what the camera "thinks" the color temperature is.

  • Other factors may influence results: Metering, shutter speed, aperture, distance to subject, etc. Using the second lens from earlier, temperatures ranged from about 3000K to 3200K when setting custom white balance with the same lamp and white card.

  • My intention was to figure out the real color temperature achievable in practice by the lamp, because whatever the manufacturer prints on the lamps as intended color temperature might be miscalibrated or distorted by chemical compound mixture proportions of the coloring. Also, incandescent and other gaseous lamps will just change color based on supplied electric power parameters. So, the question, rephrased, was "Is it possible to determine the real color temperature of light at the time the photo was taken, using the EXIF data provided by the camera" assuming it was calibrated well enough. – user1306322 Sep 30 at 22:36
  • 1
    If the answer is that we cannot with a sufficient degree of precision determine the real color temperature (even assuming perfect calibration of the bulb and exact match to its advertised color temp on the label) just from EXIF data, then I guess that's my answer. Gotta use the tools designed specifically for that purpose then until consumer cameras get better (or if they ever become usable for that purpose). – user1306322 Sep 30 at 22:39
  • Cameras are not designed to get accurate color temps. As long as photos look right, it doesn't matter if the camera reports 2805K, 3164K, or 6490K. It's not just a calibration issue. What to calibrate to change 6490K to 3000K? Some colorimeters, used to calibrate monitors, can measure the ambient color temperature. – xiota Oct 1 at 1:25
0

When I shoot raw with my camera ( Nik3300) and upload to Lightroom in the develop module de temp slide will show degrees Kelvin

  • can you add a screenshot as an example? – user1306322 Nov 15 '17 at 14:57
0

Issues

By default, fields relating to color temperature, white point, etc. in the EXIF data contain values that relate to the user experience goals of the camera manufacturer. They are not based on an established industry standard and are not scientific in the sense that two products may produce different color temperature values for the same scene due to differences in the way data is recorded or differences in user experience goals.

EXIF data can contain arbitrary values. The EXIF values in a file containing EXIF data are editable by anyone or piece of software with write access to the file.

Conculusion

EXIF data can be unreliable as a source of truth. Whether or not that unreliability matters depends on the particular application for which the EXIF data is a decision making input.

Better LoL Catz are one thing. Launching the missiles? Another.

  • I know it's not perfect, but it would be still neat if there was some level of certainty like within ±300 kelvins – user1306322 Nov 15 '17 at 16:50
0

I have tried to do that to find out which filter I should use on my flash, but it does not seem possible with a Nikon D7000. Now I use the app 'Light Spectrum Pro', it works great. I also found an alternative, 'Cine meter 2', which is very expensive compared to the app I use. It is very cheap compared to a professional light meter (800 - 1500 dollar), and seems to do pretty much the same, so I may very well buy it later.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.