What is a spectrometer?
A spectrometer separates or measures light by wavelength. In an optical spectrometer, dispersive components such as a prism or diffraction grating can help distinguish spectral components before a detector records them. A useful spectrum reports more than a color name: it relates wavelength to a measured signal.
The tool on this page is a simplified spectral readout for Prism Riddle's light model. An ideal absorbing detector records the wavelength and power of each simulated ray that reaches its active line. It does not simulate an instrument's internal grating, slit function, calibration, noise or detector sensitivity. Read about real diffraction gratings and spectroscopy at OpenStax.
How to read the simulated spectrum
- Choose White as the source and No filter. All 27 wavelength samples should be present when the beam reaches the full detector.
- Select the red filter. Red-band samples remain strong; other sampled bands are greatly reduced.
- Change the source to Blue while leaving the red filter in place. The detector cannot discover red wavelengths that the source never emitted.
- Open View wavelength data to inspect the numerical values behind the bars.
- Open the setup in Free lab to move the absorbing detector, change the light path or compare several sources.
One power unit is the configured output of one standard lamp. The total readout is the sum of the received sample powers. Bar height uses the power scale written above the chart; the chart can rescale when the signal changes. Use the numbers when comparing two different settings, not just the apparent bar heights.
27 samples are not a continuous spectrum
The current game traces wavelengths from 430 to 670 nanometers in three sampled bands. The horizontal axis uses the actual wavelength positions. Empty spaces between samples are unmeasured positions, not a claim that a real lamp emits nothing there. The graph deliberately avoids drawing a smooth line through those gaps.
The White source gives equal power to the model's wavelength samples. It is not a calibrated daylight spectrum or a temperature-defined blackbody source. Red, green and blue are broad bands in this model, rather than individual single-wavelength lasers. A display's RGB colors are also only an approximation of what a spectral wavelength looks like.
What does a color filter remove?
Our idealized red filter transmits 92% of the incident red-band power and 0.8% of each other incident band. With an unfiltered one-unit White source, each of the three sampled bands initially carries one third of the power. The detector after a red filter therefore receives approximately 0.312 power units: 0.92/3 + 0.008 × 2/3.
Now use a one-unit Blue source with that same red filter. The detector receives 0.008 units in blue samples and zero in red samples. Filtering has removed most of the light; it has not converted blue photons into red ones. The table makes this difference visible even when a faint rendered beam is difficult to judge.
Why does the light stop at the detector?
This detector absorbs the rays crossing its finite active line. A receiver behind it does not receive those same rays again. Two detectors on separate splitter outputs can each read a share of the original power, but their combined readings cannot exceed the supplied power after losses.
To compare a before-and-after measurement, move the detector or use separate controlled runs. Inserting an absorbing screen into the input path of a curved mirror can prevent light from reaching the mirror at all. This is part of the experiment, not a failure of the light source.
Spectrum, visible color and puzzle goals
Different wavelength combinations can look similar on a monitor. The histogram is therefore more useful than a glowing sprite when checking whether a mixed-color goal is missing a band. A brightness condition, a color-balance condition and a maximum-energy limit are different requirements; the same beam can satisfy one while failing another.
Explore the concave mirror and convex mirror experiments to see how geometry changes which rays reach a particular area. Free lab supports multiple lamps, absorbing receivers and saved setups, so you can build your own comparisons without spending walkthrough credits.
A learning tool, not a calibrated instrument
Real spectroscopy involves wavelength resolution, detector response, noise, calibration and often complex dispersive optics. None of those instrument characteristics can be inferred from this idealized game readout. The purpose here is to make the simulated spectral power explicit and auditable, including zero-signal and partially calculated states.