The Nature of White Light

"The greatest deception in the realm of color photography is the belief that you can faithfully capture the colors of the physical world simply by recording the intensities of red, green, and blue."

This prevailing method—essentially the only one used in modern mainstream photography—is only truly effective under two highly specific (and rare) conditions:

1. The camera's spectral sensitivity curves are a perfect mirror of the human eye's perception.
2. The light source is perfectly smooth in its spectrum and remains identical for every photo taken.

Obviously, the first requirement is incredibly difficult to achieve, and the second is practically absolutely never true. This is why photographers are locked in a perpetual, losing battle to achieve "true" color.

Real-World Color Confusion

This isn't just a problem for photographers; it affects various professions and daily chores:

---

Building a DIY Spectrograph

To better understand these optical frustrations, I recently constructed my own spectrograph. I've documented the process here to organize my findings and share the knowledge.

The Bill of Materials

• Two-part plastic tubing (custom machined)
• Diffraction grating (1000 lines/mm foil from eBay)
• Two 50mm lenses
• PVC water pipe (42mm length)
• Hobby knife blades (for the slit)
• Hot glue

Technical Construction

Using a lathe, I crafted plastic tubing joined at an angle of $\theta = 146.6^\circ$. At the vertex of this angle, I installed the diffraction grating.

The Optical Path:

The camera lens is set to focus at infinity. The second lens acts as a collimator, with a narrow slit positioned at the center of its focal plane. I created this slit by hot-gluing two hobby knife blades over a hole in the lens cap.

To ensure a precise fit, I used a PVC pipe (machined to 42mm) to press-fit the lens bayonet. This setup allows me to use the lens's own focusing ring to align the slit perfectly, thereby focusing the entire instrument.

Note: Some alternative designs omit the second lens, opting for a very long tube to place the slit far from the grating.

---

Analyzing the Spectrum

The "Smooth" Baseline

I captured a spectrum using a blue-tinted "daylight" incandescent bulb (typically used as a dim indicator light).

• Left Edge: $\approx 700\text{nm}$
• Right Edge: $\approx 400\text{nm}$ (dipping into the UV range)

The LED Evolution

Comparing old and new LEDs reveals a massive difference in "color purity."

Old LEDs: These had a very broad bandwidth. For example, an old "green" LED actually spans from red, through orange and yellow, and into the green. It only looks green because:

1. It falls in the lower half of the green range.
2. It is often housed in a green-tinted plastic case.

Modern LEDs: These have a much narrower bandwidth, providing a far more precise and "pure" color definition.

Case Studies in LEDs

{
  "Old_Blue_LED": {
    "Cost": "Expensive (at the time)",
    "Brightness": "Dim",
    "Actual_Color": "Cyan-centered",
    "Observation": "Broadbanded; bought as a novelty when blue LEDs were thought impossible."
  },
  "UV_LED": {
    "Claim": "Ultraviolet",
    "Observation": "Strong signal in the far violet; required extending the image crop to capture it."
  }
}

The Goal of "White" Light

While indicator LEDs prioritize saturated colors, actual lamps are designed to mimic daylight. The gold standard for color reproduction is natural light under an open sky, which occurs in two primary forms:

• Sunny Day: A blend of yellowish sunlight and blue atmospheric scatter.
• Overcast Day: A thick layer of white clouds that diffuses and mixes the light.

Interestingly, clouds are not perfectly neutral; they absorb certain wavelengths more than others, affecting the final color of the light...