You think the red in your photo looks wrong because the white balance is off. It usually isn't. The red looks wrong because your sensor's spectral sensitivity does not match the light hitting the subject. The camera records a number. That number represents the energy the sensor absorbed, not the hue you saw with your eyes.
Your eyes use three cone types, each with its own response curve. A digital camera uses a Bayer filter mosaic with different curves again. They overlap differently. When the light source is missing the wavelengths your sensor cares about most, the color shifts in a predictable, mathematical way — before white balance even enters the picture.
Gamut and Spectral Sensitivity Are Two Different Things
A color gamut is the triangle of primaries a display, printer, or color space can reproduce, anchored to a white point. That's a fixed geometric object — you can draw it, compare it to another gamut, measure how much of it overlaps.
Spectral sensitivity is upstream of all that. It's the curve describing how strongly a sensor (or a film emulsion layer) responds to each wavelength of light. The sensor doesn't have a gamut of its own in the display sense. What it has is a set of three overlapping response curves that determine what raw data even reaches the pipeline that will later map it into some gamut.
This distinction matters because it explains why two cameras can target the identical output gamut and still disagree on the same scene. The gamut is the destination. Spectral sensitivity decides what evidence you're allowed to bring with you.
Where the Data Actually Gets Lost
Consider a deep blue object under tungsten light. Tungsten is low in blue energy to begin with. The sensor's blue channel starts underfed. Red and green stay relatively full. Later, the processor tries to balance the white point, but it can't reconstruct blue data that was never captured. The result is a blue that reads as purple or gray — not a processing mistake, just a spectral shortfall no algorithm can undo.
Film behaves the same way, for a different reason. Color negative stock stacks three emulsion layers, each sensitive to a different band, and uses a yellow filter layer to keep blue from bleeding into the green and red layers below it. That filter is doing a specific, narrow job — protecting layer separation, not defining a gamut. If the light source lacks the wavelengths that layer needs, the whole color balance drifts, because one layer received less relative exposure than the film's design assumes.
Why Muted Magentas and Clipped Blues Happen
Magenta is a mix of red and blue. If the sensor's blue response is narrow, or the light source is deficient in blue, magenta loses its cool half and collapses toward a dull, brownish pink. This is common under incandescent or cheap LED sources with a weak blue spectrum. The camera can't invent blue data — it can only scale what little it got, and low gain on a starved channel just compresses it further.
Clipped blues sit at the other end. Point a sensor at a scene rich in blue — an overcast sky, certain LED panels — and the blue channel can hit its ceiling before red or green get close to theirs. Once it clips, the ratio between channels breaks, and neighboring colors shift too. Reds can look artificially warm simply because blue stopped participating in the balance.
Matching the Light to the Sensor
You can predict most of this. Know the spectral power distribution of your light source, and you can predict how a given sensor will respond before you shoot. A 3200K tungsten source is blue-starved; expect a quiet blue channel. A 5600K daylight source is closer to balanced across all three. An LED with a ragged spectrum — spikes here, gaps there — will show artifacts specifically in the colors that fall into its gaps.
That's also why two cameras see the same scene differently: one sensor might carry a broader blue response than another. Under a light source with a weak blue peak, the broader sensor pulls in more of what little blue exists, and the image reads cooler and more saturated. The narrower sensor comes back warmer and flatter. Neither is wrong. Both are accurately reporting the collision between a specific light spectrum and specific hardware — you can see the same effect by feeding different stock profiles the same frame in the Cineon studio.
Color isn't a property of the object in front of the lens. It's a property of the light, the sensor that caught it, and the gamut it eventually gets mapped into — three separate stages, easy to blur into one. When a red looks wrong, the white balance slider only touches the last stage. If the deficit happened upstream, at the sensor, no amount of sliding fixes it. You'd need different light, or a different sensor.