πŸ’‘ Measuring and Metering – ColorMeter

We’ve discussed how cameras use a built-in meter to analyze the exposure of the scene. This can be used to auto-adjust the exposure or as an informational tool to assist your exposure decisions. As advanced as these metering systems are, there’s a lot of merit to using a purpose-built meter for both photo and video work.

A dedicated meter can be positioned to read the light from outside the camera’s position. We call this an ‘incident’ reading. When the light is metered from the camera position we are measuring the light reflected from the subject and scene and into the meter. That’s termed a ‘reflective’ reading and is similar to the in-camera meter you’re used to using. But a dedicated meter does more than allow you to measure light away from the camera. It will read the intensity and color quality of a light. Quantifying these properties lets us do lots of useful stuff. We could compare two lights, a key and a fill, against each other to calculate our “lighting ratio.” We could compare two different lights from two different manufacturers to compare their output and color quality. We could even use the meter to determine “ambient” light on a set or to help us “set a stop” on the camera that correlates with the amount of light in a scene.

Those working in film are generally stronger advocates of light meter use. While there are major benefits to using an external meter, you need to know those benefits benefit you before buying one. In the days when you couldn’t see what you were shooting it was a necessity. But remember, the reflective meter in your camera is telling you the ultimate exposure of your scene, which is arguably the thing you care about most. And you can review your shot at will during setup, shooting and playback. So while a light meter can be very useful for a lighting department to set up a scene to a certain base stop of light, an individual run-and-gun shooter will likely derive less use from it.

Uses For A Light Meter

Lighting background and foreground relative to each other

Lighting a green screen is easier with a meter because, as mentioned, you can measure the light falling on foreground and background separately. It’s often helpful when shooting on a white background or a green screen to expose the background one stop brighter than the subject. This relative reflection is hard to gauge with a reflective meter, but still somewhat possible with a waveform monitor.

Measuring light output

Light’s intensity is commonly measured in lux which is equal to 1 lux = 1 lumen / mΒ², though we won’t get into that quite yet. As a reference, you’d want around 2000 lux at 1 meter for any light you intend to use as a main source of light (the key) in anything but a darker interior. 

Measuring light quality

Now that tungsten sources are less frequently used, the “color quality” of lights varies greatly. Old-school sources like HMI have always been pretty terrible color-quality-wise, but they weren’t as accessible to consumers. Now that you have a variety of inexpensive LED options to choose from, it’s important to be able to measure their color quality.

What meter do I buy?

Sekonic is a common choice, but their basic meters only measure light intensity, not color information. I don’t find these as useful, and once you get the color information the meter gets pricey. A cheaper way to get into this is with an old Android phone, an inexpensive ColorMunki probe, and an app called “Color Meter” by Graeme Gille. The ColorMunki is one of the few “affordable” spectrophotometers/spectroradiometers. These big words simply mean that the meter is able to sense light being reflected into its sensor and break apart the wavelengths that comprise that light for analysis. This feature is useful for display calibration as well as calculating light quality.

At NAB 2017 I compared the performance of this combination to a much pricier Sekonic and was quite blown away by the performance of this application.

As discussed in the lighting section, “discrete spectra” lights contain narrow bands of color as opposed to “continuous spectra” lights like an old tungsten bulb or the sun which have a smooth, predictable spectral output. This can affect your picture rather dramatically since if the light source illuminating your subject doesn’t contain all the colors required for accurate depiction you’ll see a color shift. A few standards exist to help gauge the expected outcome of your light.

CRI: A color rendering index. Basically 15 patches of color. We know what color they should be as illuminated by the sun and reflected into our eyes so we can take a reading and see how our light compares. 100 is a perfect score, and generally speaking, lights about 90 or so are usable. “Extended” CRI simply refers to use of an average of all the patches rather than the first eight (which you’ll hear called “general CRI” or “Ra” for “average”). CRI is an older standard with some flaws, but it’s a good starting point.

TLCI: The “Television Lighting Consistency Index” improved upon CRI by referring the result to an actual video capture device rather than a human’s perception of vision. Aside from that, it’s a similar metric, measured from 0–100, with ostensibly superior accuracy to a traditional CRI measurement.

SSI: But that still wasn’t good enough. CRI used human color perception as the target, TLCI used a video system. But “video systems” vary greatly, and the 3-chip cameras TLCI was designed around are the minority in today’s single chip world. The “Spectral Similarity Index” tries to remove this problem by simply comparing our light to the spectral output of a known standard (again, the sun) and representing the closeness of the match with a single number.

TM-30-15: This metric one-ups CRI by using a whopping 99 patches, but also seeks to convey saturation information as well.

ColorMeter

This Quick Guide Video explains most of the basic features.

  • Remember to read the “Y” value for measuring LUX (cd/m2) which will be in the middle if the readout is XYZ and the first value if it’s Yu’v’
  • Remember to use “Daylight Locus” instead of “Black Body Locus” when measuring something higher than 5000ΒΊK.
  • Use β€œlighting” preset with ambient mode to measure lights. I like to add two more CRI value readouts to this preset mode.
  • Long presses do different things as do two-finger taps.
  • Swipe left then drag up and down to re-order.
  • Swipe right to delete the results category.
  • You can use a Chromecast as a pattern source generator!
  • Take ambient light readings
  • Measure CRI and TLCI
  • Measure printed media
  • Measure flash (Xenon) intensity (in Lux) and spectrum
  • Measure contrast ratio! (below)
  • Because the CRI 1995 standard dictates that when the white point is further from the illuminant locus than 0.0054 Duv, then the CRI calculation is not valid. The CRI uses a two part illuminant locus as a reference, the Black Body locus for temperatures below 5000ΒΊK, and the daylight locus above 5000ΒΊK, so if the CCT Readout shows a number exceeding that threshold you’ll get the ‘invalid’ warning on your results. Make sure you’re using the right reference (daylight or tungsten). (Invalid) on CRI means it’s outside of what it would consider a reasonable delta but the reading should still be pretty accurate.
  • With Argyll CMS remember the MIDDLE value is Lux! Y=Lightness
  • xy is just another way to represent XYZ
  • U’v’ is simply a more perceptually uniform space than XYZ
  • Set to β€œAmbient” measurement position. Calibration is independent of light in the environment from my experimentation. It also seems the meter can be an inch away from the screen without drastically different results.
  • Compared i1D3 colorimeters with Colormunki Spectro (only the latter has ability to measure CRI/TLCI spectrum). Differences were negligible–ColorMunki averages 80 lux lower reading.
  • An easy way to measure the contrast ratio of an LCD panel with the app: Add a Visual Contrast Readout, measure Black & save to Reference, then measure White. This can be useful for comparing panels at trade shows.
  • You can gain an idea of gamut using ColorMeter by measuring R, G & B primaries, saving each to the appropriate Register, and then switch to Chromaticity plot, and turn on “Show Primaries” in the Options menu.
  • You can add Pantone color palettes as .cxf files to the app and then scan a color in real life and the app tells you the closest Pantone match.